2000s fashion essay

To revisit this article, visit My Profile, then View saved stories .

A 2000s Fashion History Lesson: Logos, Low-Rise, and It Bags

By Lilah Ramzi

If one moment sums up early-2000s fashion, it was February 2000: Jennifer Lopez wore a chiffon Versace dress to the 42nd Grammy Awards, and the world was never the same. With a neckline that plunged all the way below her belly button, the world was in a tizzy— but how did it stay on? And what did it look like from the side? Everyone wanted to see the dress ! So much so that it inspired the idea of Google Image Search. And so began a decade of digital fashion consumption.

Dubbed the aughts and the naughts (the former used in the US, the latter in the UK), the decade was ruled by celebrity: a proliferation of reality TV , blogs filled with paparazzi shots, and 2006’s Twitter, which gave fans a front-row seat into the lives and musings of their favorite celebrities—often without a publicist’s intervention.

Regarding mainstream fashion, it was a decade of taste so bad it was almost good— almost . Think Justin Timberlake and Britney Spears in matching denim getups , Juicy Couture sweatsuits galore, Von Dutch trucker hats, and Ed Hardy tattoo tees.

On the runways, logomania ruled, and designers sexed it up with Tom Ford, Donatella Versace, Roberto Cavalli, and Hervé Léger infusing heady doses of sultry into their collections. A whirlwind look back at the in-your-face trends of early-2000s fashion, below.

Women’s Trends of the 2000s

If eyes are the window to the soul, by the time the 2000s rolled around so too were handbags. Thanks in part to a certain Ms. Carrie Bradshaw , bags became the focal point of an outfit. After the launch of Fendi’s Baguette in 1997 (as fashion lore goes, the bag, designed by Silvia Venturini Fendi, was initially unpopular among the Fendi design team, who feared it had too much personality during a minimalist-fashion era), it skyrocketed to handbag stardom and is often considered fashion’s first It bag . Bradshaw, of course, famously carried multiple iterations of the Baguette throughout Sex and the City ’s fabulous six-season run.

In the 2000s each brand had a white-hot bag (or two) that fashion folk clamored to carry. As mentioned, Fendi had its carb-inspired Baguette; Balenciaga had its slouchy Motorcycle City Bag , introduced in 2001 by then creative director Nicolas Ghesquière ; Chloé had its boho Paddington bag , created by then creative director Phoebe Philo in 2005; Galliano debuted the Saddle bag at Dior’s spring-summer 2000 show. And over at Louis Vuitton, Marc Jacobs was cranking out It bag after It bag with clever collaborations. Artists were invited to make interventions with Vuitton’s storied monogram: In 2001, it was Stephen Sprouse’s graffiti collection; in 2003, it was Takashi Murakami’s smiley flower-and-cherry collections. These bags were not to go unnoticed: statement making, eye-catching, and the center of attention!

In 2002, Sarah Mower charted the rise of the It bag in Vogue ’s October issue : “In about 1993, fashionettes everywhere wanted the identical item at the same time (a Kate Spade shopper, a sporty Prada knapsack). Then came the great Baguette hunt for exotic hide and fabric varieties (a standard shape, but now in individualistic surface patterns). Before long, thanks to Gucci, Prada, Fendi, Dior, we were buying entire bag wardrobes, toting them as up-to- the-split-second status trophies.”

“Fashion is pushing its love affair with logos to the limit. If you’ve got it, flaunt it!” wrote Vogue in its March 2000 issue .

All the aforementioned It bags were typically brandished with a logo—but that didn’t stop there! Dior’s bikinis featured their Oblique monogram; Burberry checked just about every object that could be checked; Gucci’s double Gs festooned belt buckles; and Marc Jacobs even painted his naked body with the neon pink Louis Vuitton logos inspired by LV’s collab with Stephen Sprouse in 2008 for an ad campaign. If early-2000s fashion had a square inch to spare, it was filled with an alphabet city of letters.

Tongue-in-cheek nods to this super-trend were made; in the brilliant March 2000 editorial “Branded” for Vogue, Helmut Newton took model Angela Lindvall and decorated her string-bikini-clad body with a smattering of logos belonging to all the mega fashion houses—a mishmash of monograms and icons where Chanel’s double Cs intermingled with Fendi’s double Fs to create a nonsense of logos. And Dapper Dan, all the way back in the late 1980s, was cutting up Gucci garment bags and transforming them into jackets in a brilliant reclamation of luxury.

All this emphasis on the bag you carried spawned a growing market of counterfeit bags—and with it a new industry of authenticators. In May 2001 Vogue reported on the issue: “Imitation is the sincerest form of flattery, not the chicest. When you’ve dished out big bucks for the latest leather logo bag but suspect it’s a fake, who ya gonna call? David Colman meets fashion’s new crime fighters: the Authenticators.”

We can thank (or point the finger at) Alexander McQueen for the low-rise trend . Though it reached its zenith in the aughts, the designer debuted tailbone-revealing bumster trousers in his 1993 Taxi Driver collection—his first collection after graduating from Central Saint Martins—and notably again in his 1995 spring-summer collection.

Five or so years later, the trend had gone full-on mainstream. (In 2001 even Levi’s introduced a Superlow jean silhouette.) It was the default rise of all jeans and pants. Britney Spears was a notable proponent of the belly-baring trend.

The low-rise trend also coincided with the rise of jeans for any occasion. Though it was the 1970s that gave us designer denim, brands continued to offer denim: bedazzled with diamatés, distressed to smithereens, painted with floral motifs.

“Designer jeans: Can we count the ways we adore them now?” wrote Sarah Mower in Vogue ’s September 2001 issue . “They’re rocking. Posing and pirouetting. Diving off into a zillion cuts, rinses, and textures, while simultaneously gate-crashing into unheard-of echelons of society. Jeans can be street, ladylike, dressed up, dressed down. They’re cool for everyday. They do black-tie. They’re fantasy and reality.”

Twin forces helped to craft a new trend by the mid decade. Stateside, it was Mary-Kate and Ashley Olsen . Retired from their teenybopper flicks (their last, New York Minute , came out in 2004), the pair was busy crafting a new, unique look that spawned dozens of copycats. Over in the UK, Sienna Miller was making waves for her take on boho : more festival chic, fewer layers.

The sum of the look was greater than its parts—fashionably speaking. It was a movement more to do with styling than individual pieces: peasant dresses over puddle jeans layered with a bolero; a functionless belt (preferably perforated); and a skinny scarf spun with metallic threads for flair.

In addition to the Olsens and Miller, there were Kate Moss , Mischa Barton, Nicole Richie , and other famous adopters. It’s said the force of this look, operating behind the curtain like Oz, was not a fashion designer (though Phoebe Philo did send out superfluous spaghetti-thin scarves for Chloé’s spring-summer 2005 collection) but the LA-based stylist Rachel Zoe .

Though the decade was dominated by denim and It bags, there was room for a few trends born on the runway—especially for women needing something more sophisticated than denim and boho. Enter the bodycon dress.

At his spring 2006 collection, Roland Mouret debuted his Galaxy dress, and without the help of TikTok or Instagram—this was pre–social media, after all—the look went viral.

“You hadn’t been able to open a newspaper or magazine without seeing another young Hollywood A-lister wearing that hot-damn curvy dress: Scarlett Johansson, Cameron Diaz, Rachel Weisz, Keira Knightley, Nicole Kidman—so many, it was verging on a joke,” wrote Mower in Vogue ’s October 2007 issue. “You put it on, and you looked like an icon,” added the designer.

With as many colors as a book of Pantone swatches, the dress came in a double wool crepe that gripped the body like a second skin. The silhouette was below or at the knee and waist cinching, with a square neckline and sculptural cap sleeves. Like the jeans in the Sisterhood of the Traveling Pants , it was universally flattering.

Mouret wasn’t the only designer delivering bodycon dresses. At this time, Hervé Léger (which, in April 2007, had been relaunched by Max Azria) was binding women up like metropolitan-chic mummies in his bandage dresses. No red carpet was complete without at least one bandage dress.

In the aughts, magazines (and tabloids) were still the predominant form of fashion news, along with budding blogs and news websites. All of these were filled with snaps of celebs circulating in the starry constellations of Hollywood. Paris Hilton and her Simple Life costar Nicole Richie were frequent tabloid fodder; so too were Britney Spears, Cameron Diaz, Lindsey Lohan, and newcomer on the block Kim Kardashian.

These off-duty looks produced head-scratching trends adopted by the masses: Juicy Couture velour tracksuits with script on the derrieres, ragged-edge miniskirts paired with Ugg boots , and baby tees emblazoned with words like angel . Page-boy caps, It bags, and a Starbucks Frappuccino finished off the look.

Top Designers of the 2000s

Chloé, Hervé Léger, Roland Mouret, Balenciaga, Burberry, Yves Saint Laurent, Juicy Couture, Marc Jacobs, Versace, Prada, Miu Miu, Dolce & Gabbana, Alexander McQueen, Calvin Klein, Fendi, Tom Ford, Gucci, Martin Margiela, Dries Van Noten, Vivienne Westwood, Comme des Garçons’ Rei Kawakubo, Yohji Yamamoto, John Galliano, Chanel, Oscar de la Renta, Kenzo, Giorgio Armani, Valentino, Michael Kors, Moschino, Alaïa, Lanvin, Roberto Cavalli, Stella McCartney, Viktor & Rolf, Pucci, Gareth Pugh, Christopher Kane, Rodarte, Phillip Lim, Jason Wu, Jimmy Choo, Christian Louboutin, Nina Ricci, Raf Simmons, Isabel Toledo

Men’s Trends of the 2000s

The top of the decade was dominated by the stage looks of boy bands: frosted trips and leather jackets with a metallic sheen. For teenagers, Abercrombie & Fitch’s elitist Americana prep ruled; the more polos the better. Hip-hop and the styles worn in music videos on MTV were chock-full of fashion like Air Jordans, iced-out chains, and baggy jerseys.

By the mid decade, a new look was taking hold thanks to the runways of Dior Homme, which appointed Tunisian-Italian designer Hedi Slimane as creative director. For Dior, Slimane churned out ultra-skinny pants, ties, and jackets inspired by rockers with waifish frames like Pete Doherty. If Kate Moss’s arrival in the 1990s signaled the heroine-chic era for women, Slimane passed the baton onto menswear in the 2000s. It was a look that carried over into the early 2010s and has since been dubbed indie sleaze . A crop of garage rock and post-punk bands (The Strokes, The Killers, The Libertines) epitomized the look.

For men who didn’t dabble in the mall-based or music-video-inspired trends of the aughts, the tailoring of Giorgio Armani remained the height of fashion.

In the Culture

The decade started off with the 9/11 terrorist attack, which left the world reeling. It ended on an optimistic note with the arrival of Barack Obama as the 44th president of the United States. And when his wife Michelle Obama stepped out at the inauguration gala in a white Jason Wu dress , she launched the young designer into the spotlight. In 2001 Apple changed the game with the introduction of the iPod, and then six years later, on June 29, 2007, Apple released its first-generation iPhone, making a leap in technology with the first smartphone.

The decade also saw the rise of reality TV: Survivor aired in 2000, America’s Next Top Model in 2003, and Keeping Up With the Kardashians arrived in 2007.

The first Harry Potter film arrived in 2001, Twitter launched in 2006, and that same year the world got its first Taylor Swift album. In 2007 the first episode of Gossip Girl aired, and Beyoncé and Jay-Z tied the knot in 2008.

Vogue World: Paris will pair select sports—cycling, gymnastics, tennis, tae kwon do, fencing, and breakdancing, among others—with French fashion from every decade since 1924. The show will highlight French designers, current and past, as well as houses that historically present their collections in Paris.

For front-row tickets, email [email protected]

• Celebrity Fashion

The Best and Worst Early 2000s Fashion Trends Essay

My question to the early 2000s — why did we dress like this.

The early 2000s was a strange time for pop culture . A time where women kissing was seen as shocking (Britney, Christina, and Madonna at the 2003 VMAs), but using people and pets as accessories wasn't ( Gwen Stefani 's L.A.M.B, and Paris Hilton 's Tinkerbell). The early aughts was bizarre, for sure, especially between 2003 and 2006, which created some of the worst personal style of the past century. Yeah, it was that bad, which is why we absolutely love it.

Early 2000s fashion sense might be trash, but it was totally iconic. We had a strange obsession with really long torsos. Not sure where that came from, but all signs seem to point to the debut of The Simple Life in 2003. Paris Hilton made Juicy Couture track suits, tiny dogs, and flip phones the height of cool. But a year prior, Christina Aguilera released her career defining hit "Dirrty", which saw a boost in celebrities sporting midriff-baring crop tops and streaky highlights on the red carpet. Then, Laguna Beach premiered in 2005 and set the groundwork for the UGG-wearing, pumpkin-spice-latte-drinking women on today. Arguably, Lauren Conrad 's greatest contribution to fashion history was the "pouf" hairstyle and the revival of the "going-out top" with jeans. Gosh, were the hairstyles bad, but nothing was worse than our obsession with wearing low-rise jeans with everything — even floor length evening gowns. Accessories-wise, it was all about the chain belt, the baker boy cap, and covering everything in Swarovski crystals.

A decade that started with Britney Spears and Justin Timberlake 's matching denim-on-denim couples outfits at the 2001 American Music Awards and ended with the premiere of Keeping Up With the Kardashians and their smoky eye beauty look in 2007 was bound to be one for the books. Ahead, 12 of the best and worst fashion trends from the early aughts — which one were you guilty of wearing?

1 Early 2000s Fashion Trend: Dropped Waist Minidress

Girls Aloud at the 2004 Disney Channel Kids Awards.

Early 2000s Fashion Trend: Dropped Waist Minidress

Beyoncé at the 2003 Essence Awards.

2 Early 2000s Fashion Trend: Dresses Over Jeans

Alicia Keys at the 2004 Grammys.

Early 2000s Fashion Trend: Dresses Over Jeans

Jessica Alba at the 2003 ESPY Awards.

3 Early 2000s Fashion Trend: Long Asymmetrical Tops

Hilary Duff at the 2003 American Music Awards.

Early 2000s Fashion Trend: Long Asymmetrical Tops

Joss Stone at the 2004 VH1 Big in '04 event.

4 Early 2000s Fashion Trend: Cropped Tops

Rihanna at JAY-Z's Teen People Listening Lounge event in 2005.

Early 2000s Fashion Trend: Cropped Tops

Keira Knightly at the 2003 premiere of Pirates of the Caribbean: The Curse of the Black Pearl .

5 Early 2000s Fashion Trend: Criss-Cross Halter Tops

Christina Aguilera at the 2002 Video Music Awards.

Early 2000s Fashion Trend: Crisscross Halter Tops

Gwen Stefani at the 2003 L.A.M.B for LeSportsac launch.

6 Early 2000s Fashion Trend: Tiny Dogs as Accessories

Paris Hilton and Nicole Richie at the 2003 Fox summer press tour.

Early 2000s Fashion Trend: Tiny Dogs as Accessories

Britney Spears in 2003.

7 Early 2000s Fashion Trend: Juicy Couture Tracksuits

Britney Spears shopping in 2003.

Early 2000s Fashion Trend: Juicy Couture Tracksuits

Paris Hilton in 2003.

8 Early 2000s Fashion Trend: Kitsch Couples Style

David Beckham and Victoria Beckham at the 2003 MTV Movie Awards.

Early 2000s Fashion Trend: Kitsch Couples Style

JAY-Z and Beyoncé at the 2004 MTV Video Music Awards.

9 Early 2000s Fashion Trend: Baker Boy Cap

Christina Aguilera at the 2003 American Music Awards.

10 Early 2000s Fashion Trend: Tight Edges

Beyoncé at the 2003 NBA All-Star game.

11 Early 2000s Fashion Trend: The Pouf

Lauren Conrad at MTV's Total Request Live in 2005.

12 Early 2000s Fashion Trend: Streaky Highlights

Jennifer Lopez at the 2004 Man on Fire premiere.

You May Also Like

Now You Know

Entertainment

The love island couples still together prove reality tv can work as cupid.

by Joely Chilcott 2 days ago

The Latest "Priscilla" Trailer Highlights the Ups and Downs of Her Relationship With Elvis

by Eden Arielle Gordon 04/10/23

Why Are So Many of Us Personally Invested in Celebrity Breakups?

by Ali Faccenda 03/10/23

Miley Cyrus Responds to Video of Adele Praising "Used to Be Young" at Las Vegas Show

by Eden Arielle Gordon 30/8/23

Follow. Like. Love.

• Entertainment Celebrity Movies TV Music Books
• Wellness Fitness Nutrition & Food Health Sex & Relationships Parenting & Fertility Astrology Identity Wellness Shopping
• Beauty Beauty News Celebrity Beauty Skin Care Makeup Hair Nails Beauty Reviews Beauty Shopping
• Fashion Fashion News Celebrity Style Street Style Clothing & Accessories Fashion Week Fashion Shopping
• Must Have Celebrity Style Fashion Shopping Beauty Shopping Home Shopping Tech Shopping Parenting Shopping Wellness Shopping Fitness Shopping Gift Guides
• Login / Register

It's a Y2K World, We're Just Living in It: 12 Trends That Defined the Iconic Era

If you grew up in the 2000s , you'll probably be surprised to hear that the decade's fashion trends are once again popular. I was born in 1994 , and if you ask me, the decade still feels like it happened just yesterday. I grew up watching  Britney Spears music videos and vying to get my hands on basically everything Paris Hilton wore. For me, 2000s fashion trends are more nostalgic than those of the  '80s or '90s .

Yes, I'm talking about the teeny-tiny  handbags and even tinier  strappy sandals that defined the decade's aesthetic. I won't lie, though. I'm thankful that not everything we wore in the 2000s is making it into the  2010s because, well, styles like  ultra-low-rise jeans are best left in the past (in my opinion). But if you're as ready as we are to embrace the best of noughties fashion once again, see all the 2000s fashion trends that are still going strong ever since their initial resurgence. 

WHO: Naomi Campbell and Iman, 2004

Whether logo-filled or embellished with buckles, baguette-shaped  bags were considered  the 2000s It trend to own. It was during this era when the Dior Saddle, Louis Vuitton Pochette, and Fendi Zucca bags all featured prominently on the arms of the A-list set.

WHO: Rihanna, 2005

Now, would it even be an early aughts story without a mention of low rises? They were as iconic a silhouette during the decade then as they are now, and it was all about pairing them with a crop top for an exposed midriff moment. Daring? Yes. Nostalgic? Absolutely.

WHO: Nicole Richie, 2006

Shield sunglasses are the accessory trend that's still popping up everywhere from Miu Miu to Le Specs. The good news is that you don't have to spend a fortune to nail this noughties trend.

WHO:  Paris Hilton, 2003

Everything about 2000s fashion was defined by a more is more mentality: more embellishments, more layering, and showing more skin. Skirts were hemmed high to show off as much leg as possible, while waistlines were slung ultra-low to maximize an exposed midriff which brings us to the epitome of early 2000s silhouettes: the micro mini skirt. 

WHO:  Jessica Alba, 2003

Ready or not, this  divisive outfit trend  is bubbling back up again—except that 2023's versions are much more refined and forward.

WHO: Sarah Jessica-Parker, 2001

Soft ruffles are still having a major moment today, and we're thanking icons like Carrie Bradshaw who made the asymmetric silhouette look so good.

WHO: Destiny's Child, 2000

One thing we'll always remember this decade for was how committed everyone was to achieving a standout look. No embellishments were spared when getting dressed, and that means we saw all manner of glitz and glam throughout the decade's most memorable ensembles, but especially the popularity of crystal dressing. Whether it was a crystal mesh dress à la Paris Hilton at her 21st birthday or the coordinating crystal bra tops Destiny's Child donned at the 2000 Hip-Hop Music Awards, more is more when it came to this shimmery trend.

WHO: Britney Spears, 2002

Love them or hate them today, there's no denying that bootcut jeans were the ultimate denim trend of the 2000s with their slim fits and flared-ankle shape.

WHO:  Mariah Carey, 2000

If you grew up in the early 2000s, you know all about the "going out top" (looking at you, Millennials). They were an entire category of our closets back then, defined by statement colors, asymmetric necklines, cropped hems, and pretty embellishments. Mariah Carey's sequin butterfly top is the blueprint here. Essentially, the going-out top was the only thing you needed to own to make an outfit look good—pair one with jeans and you're good to go.

WHO:  Christina Aguilera, 2000

Say what you will, but the 2000s were nothing if not one big experiment in taking fashion risks. And boy did celebs embrace a risqué look. When Christina Aguilera showed up to the MTV VMAs in 2000 in a G-string skirt, it sparked an entire movement of purposefully showing off your thong straps via hip cutouts or built-in straps.

WHO:  Meagan Good, 2003

Trucker hats are continuing to be relevant, but the Von Dutch styles will always be reminiscent of the era. Everyone and their mother sported the iconic brand.

WHO: Kate Moss, 2005

Skinny scarves were a micro-trend that fashion people adored adding to their outfits throughout the Y2K era and now, people are yet again draping themselves in this trend. Buy into it before it sells out.

Anna is an editor on the fashion team at Who What Wear and has been at the company for over five years, having begun her career in the Los Angeles office before relocating to New York, where she's currently based. Having always been passionate about pursuing a career in fashion, she built up her experience interning at the likes of Michael Kors, A.L.C., and College Fashionista before joining the team as a post-graduate assistant editor. Anna has penned a number of interviews with Who What Wear's cover stars over the years, including A-listers Megan Fox, Issa Rae, and Emma Chamberlain. She's earned a reputation for scouting new and emerging brands from across the globe and championing them to our audience of millions. While fashion is her main wheelhouse, Anna led the launch of WWW Travels last year, a new lifestyle vertical that highlights all things travel through a fashion-person lens. She is passionate about shopping vintage, whether it be at a favorite local outpost or an on-the-road discovery, and has amassed a wardrobe full of unique finds. When she's not writing, you can find her shooting street imagery on her film camera, attempting to learn a fourth or fifth language, or planning her next trip across the globe.

By Natalie Munro

Time to find out.

By Grace Lindsay

A foolproof pairing.

Y2K, the grown up way.

By Remy Farrell

By Harriet Davey

And you might already own it.

Yep, they're still a thing.

By Anna LaPlaca

It is possible. We promise.

By Emily Dawes

We've gone into the archives.

By Elinor Block

Bet you'd forgotten about them.

By Natalie Hammond

• Contact Future's experts
• Terms and conditions
• Privacy policy
• Cookies policy
• Accessibility Statement
• Advertise with us

Who What Wear UK is part of Future plc, an international media group and leading digital publisher. Visit our corporate site . © Future Publishing Limited Quay House, The Ambury, Bath BA1 1UA. All rights reserved. England and Wales company registration number 2008885.

• Reports of arrested students receive holds on their accounts, some concerned about Graduation
• Statement from the UC San Diego writing community: “You wrote them in the cruelest way”
• Letter from UC San Diego Faculty to the Students of the Gaza Solidarity Encampment
• Submission: Silenced identities - a letter from a Jewish student
• Violent responses to pro-Palestine encampments violate content neutrality
• Senior Sendoff: Chaz Cepielik-Weeks
• Senior Sendoff: Bailey Bujnosek
• Senior Sendoff: Marcella Barneclo
• Senior Sendoff: Leah Schiffer
• Senior Sendoff: Millie Root

The UCSD Guardian

10 Early 2000s Fashion Trends We Must Revive

8.  Cargo Pants: Ah, the baggy, pocketed, and somehow trendy cargo pants. A wide variety of colors are accepted, but I would personally suggest an olive green. Pair those with a black top and you might as well call yourself Kim Possible, ready to take on the world with pocket room to spare!

Your donation will support the student journalists at University of California, San Diego. Your contribution will allow us to purchase equipment, keep printing our papers, and cover our annual website hosting costs.

The Student News Site of University of California - San Diego

Comments (0)

Cancel reply

Your email address will not be published. Required fields are marked *

• Share full article

Advertisement

Supported by

Notes on the Culture

Gaudy and Gleeful, Early 2000s Fashion Is Making a Comeback

In an otherwise restrained era, the return of hot pink, exposed midriffs and more allows for both escape and self-expression.

By Nick Haramis

“BIMBO SUMMIT.” SO read the headline on the front page of The New York Post on Nov. 29, 2006, over a photograph, taken outside the Beverly Hills Hotel in the small hours of the morning, of the heiress Paris Hilton, the actress Lindsay Lohan and the pop star Britney Spears. Their “unholy alliance,” as the corresponding newspaper item called it, was being presented for scorn: Spears had recently given birth to her second child; Lohan, only the night before, had accused Hilton of throwing a drink at her; and yet here were all three women, wedged uncomfortably into a two-seater sports car well after midnight. The article’s final line could easily have been uttered by Lohan in the 2004 film “Mean Girls”: “And regardless of how long these friendships last: Skanks for the memories!”

Antipathy for the decade’s celebrities — particularly the young women, including Hilton, Lohan and Spears, who had only recently been anointed as its idols — was at its peak, with tabloids and gossip blogs making vicious sport of their public drubbings. This “meeting of the minds,” as The Post also described it, marked both the pinnacle and the beginning of the end of this period: the so-called McBling era, an exuberantly trashy coda to the Y2K years that lasted from roughly 2000 to 2008 and was characterized by trucker hats, fake tans and pubis-baring velour tracksuits. If fashion leading up to the new millennium aspired to an iridescent, digitized new day, McBling (its name a nod to the flashy logomania that was then prevalent) came stumbling downstairs the morning after, hung over and hiding behind a pair of pink Baby Phat aviators. The former style — punctuated by cargo pants, strappy tops and white eye shadow — was an optimistic, if superficial, exploration of how the future might look; the latter led, in every direction, toward unabashed bad taste. “Fashion at the time had a sense of humor about itself,” says the British designer Christian Cowan, who was born in 1994 and whose runway shows Hilton has walked in twice. “It was about being perfect, fake and bright. Everyone knew it was kind of shallow.”

But when the stock market crashed in 2008, it took ostentation down with it. Inconspicuous consumption became the goal. The hipster was ascendant, in worn flannel and skinny jeans, and in the mid-2010s, normcore followed with its aim of, as its name suggests, looking as average as possible in everyday sweats and sneakers. Not long after, millennials claimed their own color: a wallflower shade of day-old salmon that was a clear trauma response to the hot pink ensembles worn at the turn of the aughts by everyone from Aaliyah to Reese Witherspoon in “Legally Blonde” (2001).

ENTER GENERATION Z, today’s teens and 20-somethings who have never known life without social media or the internet. To them, as they navigate world-threatening issues such as climate change and the coronavirus, the idea of the millennium bug seems quaint. To them, a teen idol who forgoes underwear, as Spears had done on several occasions, might be yet another TikTok star. To them, pink should pop. And so it does on a new squad of young style icons — including the 22-year-old rapper Lil Nas X, recently seen onstage in a metallic fuchsia crop top with a matching harness and pleated skirt — who have embraced the brazen plasticity of the noughties. “It’s pretty simple,” says Cowan, whose spring 2022 collection featured baby blue sequined bandanna tops and plumed headpieces in, yes, neon pink. “We want to feel hot, even full of ourselves. It’s the kind of energy we need, especially after two years of boredom.”

Across the spring collections, brands channeled that mood by evoking the bare bellies and garish hues of the early 2000s. At GmbH, the Berlin-based designers Serhat Işik and Benjamin A. Huseby added queer and brown inflections to infamous red-carpet outfits of the period, such as the coordinated denim looks that Spears and her then-boyfriend, Justin Timberlake, wore to the 2001 American Music Awards. (Işik and Huseby’s version, modeled by a man, included a faux fur-trimmed floor-length denim coat.) For his final offering as the creative director of Bottega Veneta, the British designer Daniel Lee invited the Y2K-era legends Lil’ Kim and Mary J. Blige to Detroit to attend his show, which centered on glossy, brightly colored anoraks and blousons — a specific strain of retrofuturism that also set the tone at Celine, Etro, Tom Ford and Louis Vuitton. At Miu Miu, frayed khaki micro- miniskirts and cropped cable-knit wool sweaters called to mind the skimpy garments of the suggestively Sapphic Russian musical duo t.A.T.u., who rose to fame in the early 2000s. And on the Kim Shui runway in New York, models walked in bedazzled halter-neck dresses and skin-baring cowgirl ensembles that evoked the costumes of Christina Aguilera’s “Stripped” era circa 2002.

Yet this return to early-aughts aesthetics isn’t quite nostalgia. Nor is it entirely novel. When asked about fashion’s resurgent interest in the period, Gabriel Held, a 36-year-old Brooklyn-based stylist, pointed to the music video for the Filipina American pop star K Rizz’s 2014 reggaeton song “ Salbahe ,” in which she wears cerulean chaps and bikini bottoms and a rattan cowboy hat. “That,” he says, “is my best rebuttal to the idea that the Y2K revival is a new idea.” The look, which Held credits more to prolific stylists of the time such as Patricia Field and Misa Hylton than to any one designer or celebrity, has lingered for years at the outer periphery of good taste, kept alive both by the endless churn of the internet and by those with the desire and, let’s face it, the courage to stand out.

Still, by reconsidering it now, en masse and through the eyes of a younger generation, designers are offering an enticing do-over of sorts. Rewinding history by just two decades, we cannot only forget the intervening hardships of the Great Recession and the pandemic but also rewrite the past, overlaying a version of events in which we were perhaps kinder to and more tolerant of our pop-cultural heroines. Because in that bygone world we are conjuring, the future still looked bright for the three women in that too-small sports car. And now, some 16 years later, as if nothing had gone awry since, all three of them are making headlines again, for good news: Hilton has a business empire that extends from fragrances to plant-based seafood alternatives; Lohan, just engaged, is preparing her comeback; and Spears, recently released from an oppressive almost 14-year conservatorship , is free once more. We all ended up at the same place, but it’s comforting to imagine the ride here had fewer glitches.

Inside the World of Gen Z

The generation of people born between 1997 and 2012 is changing fashion, culture, politics, the workplace and more..

Many of Harvard’s Generation Z say “sellout” is not an insult, instead it appears to mean something strikingly corporate-minded .

A younger generation of crossword constructors is using an old form to reflect their identities, language and world. Here’s how Gen Z made the puzzle their own .

For many Gen-Zers without much disposable income, Facebook isn’t a place to socialize online — it’s where they can get deals on items  they wouldn’t normally be able to afford.

Dating apps are struggling to live up to investors’ expectations . Blame the members of Generation Z, who are often not willing to shell out for paid subscriptions.

Young people tend to lean more liberal on issues pertaining to relationship norms. But when it comes to dating, the idea that men should pay in heterosexual courtships  still prevails among Gen Z-ers .

We asked Gen Z-ers to tell us about their living situations and the challenges of keeping a roof over their heads. Here’s what they said .

Wunderlabel

Published: 2023-04-12

Fashion of the 2000s

The 2000s will go down in fashion history as the decade of mash-ups. Retro styles from past decades collide with trends from music culture. Globalization also plays its part, not only bringing Asian fashion to Europe, but also paving the way for fast fashion through worldwide supply chains.

The early 2000s (2000-2002)

The transition from the 90s to the first decade of the new millennium is gradual. Many fashion trends continue. However, events such as 9/11 quickly leave their mark on the fashion world. At the same time, the Internet, as a ubiquitous new technology, sets new accents in both the design and consumption of fashion.

Y2K fashion - futuristic, metallic, sexy

As the decade begins, fashion is also influenced by new technologies. Monochrome , futuristic garments in shades of gray, shiny black or metallic colors aim for a dark and technical look - with maximum sex appeal, of course. Mesh tops, rectangular sunglasses with metal frames, boxy skirts, leather skirts, rhinestones or halter necklines are common. Y2K, the acronym for the year 2000, which is actually a computer problem, fits perfectly with this tech-inspired trend.

80s revival - first part

By the end of the '90s, the first '80s revival was underway, building on the early '80s. In addition to tracksuits, the trend is towards the opulent use of denim fabrics, trench coats and polo shirts. The velour tracksuits of Paris Hilton and others are unforgettable. This trend will be replaced by the second wave in 2008, when the fashion of the late eighties will be in the focus of the it-girls and -boys.

Casual and leisure wear- casual and sexy

Many trends from the 90s continued into the 2000s due to the continued influence of pop stars like Britney Spears. Starting around 2001, low-rise pants, bell-bottoms, silk or satin cargo pants, or even white jeans paired with crop tops or hoodies became popular. Belly-free outfits add a lot of sex appeal. Women often wear skirts over pants, wrap dresses with floral patterns, rugby shirts and even functional jackets enter everyday life.

Sex and the City - empowerment and fashion trends

The US series "Sex and the City" has had a significant impact on women's consumer behavior. As self-confident shoppers, they no longer buy what is imposed on them, but follow their own individual style. Of course, the series also sets its own fashion trends - high-heeled stilettos and designer handbags are just two examples.

Fast fashion - the rise of disposable fashion

In the early to mid-2000s, mass consumption of cheap fast fashion spread throughout the industrialized world. In rapid succession, manufacturers adapted fashion from the catwalks of haute couture to be produced as cheaply as possible in third countries. After a journey that often takes the product once around the globe, the articles arrive at H&M, Zara and Co. to supply fashion-conscious consumers with cheap mass goods.

Puffy jackets - not just for hip hoppers

Padded puffer jackets in oversized sizes are no longer just popular with hip-hop followers. In bright colors, the warm down jackets combine functionality and fashion consciousness. They are everywhere at the beginning of the decade.

Jeans - return to denim

After 9/11 and the financial crisis, classic jeans are back in fashion. The working-class fabric is suitable for virtually any occasion. Throughout the decade, trends change in both waist height and leg width. At first, low-rise, flared jeans are in vogue, before super-tight skinny jeans are rediscovered in the middle of the decade. Even denim overalls are often seen on the red carpet. In the end, denim jeans may be one of the decade's few consistent fashion trends.

Clean chic - simple and discreet

Simple cuts and muted colors characterize this look, which is a clear contrast to the 80s revival with its bright colors and artificial materials. The matching outfit is a sleek look hairstyle or a short haircut and, of course, subtle makeup.

Fashion Mid 2000s (2003-2006)

By the middle of the decade, the trend toward a retro look was unmistakable. Above all, the chic of the sixties made a comeback on the catwalks of haute couture - and later on the mass market. Sex appeal is not neglected either. Figure-hugging jeans, thongs and waterfall shirts show a lot of skin.

60s revival - the boho chic

The boho look is a departure from the grunge look of the 90s. Britpop and garage rock bands from the UK bring fashion culture into the mainstream, sparking a fashion revival of the sixties. This includes, for example, light bootcut jeans, vintage shirts, military uniforms, velvet sports jackets, parkas, tweed jackets or western shirts, as popularized by the White Stripes. These days, however, vintage items are no longer bought cheaply in oversize at thrift stores, but in exclusive boutiques. No wonder the style in Paris is called bobo, short for bourgeois-boheme. The high-priced vintage look is often complemented by exclusive designer pieces.

Thong Underwear, Waterfall Shirts and Crop Tops

Low-rise jeans are a perfect match for belly-free shirts. A daring thong underwear is often seen peeking out from under the waistband. Jennifer Lopez and other pop stars make a fashion statement with the chain-shaped waterfall shirt with a plunging neckline and narrow fabric trim at the back. Lots of bare skin is certainly the trend here.

Military Clothing and Pirate-Style – Protest and Pop Culture

This trend has been seen in the UK since around 2005 and will continue until the end of the decade. The preference for military clothing is not only driven by the retro trend of the 80s, but also by movies like Pirates of the Caribbean. This includes harem pants, hussar jackets, uniforms, and even riding boots. But also forms of protest like the Palestinian scarf find their way into the fashion world as a rejection of the war in Iraq, while Che Guevara's clothing style is celebrated as an icon of alternative youth and young adults.

Polo Shirts and Basics - Simple and Stylish

Brands like Abercrombie & Fitch are flooding the market with stylish polo shirts. To really be on trend, the collar is important. Alternatively, the V-neck is spreading through American Apparel and other brands that make affordable basics available everywhere. After all, monochrome basics can be combined with any extravagant outfit - and should be a part of every wardrobe.

Retro Movies - Even More Revivals

Movies from the 60s and 80s also set new fashion trends in the 2000s. The stars of Top Gun and Midnight Cowboy are the role models for the renewed popularity of muscle shirts and hooded sweatshirts, cargo pants and pilot jackets.

Blazer and Shirt - Casual in Everyday Life

The combination of a casual blazer and a simple T-shirt is perfect for any occasion. Even bold all-over prints or the ever-popular Ed Hardy shirts seem made for this style. As an alternative to the T-shirt, hoodies can be worn under a blazer for the perfect combination of casual and chic.

Business Suits and Their Alternatives

In addition to classic business suits in muted colors or pinstripes, Nehru jackets and shirts with choker collars, as popularized by the Beatles, James Bond, or films like The Matrix, become increasingly popular as an alternative to suits. In the U.S., however, the casual look dominates in the middle of the decade, with subtly striped shirts and chino pants, wool sweaters, or sporty jackets.

The Late Years of 2000s Fashion (2007-2009)

Most of the trends from the early and mid-2000s are still wearable at the end of the decade. These include boot-cut jeans, tank tops, cargo pants, boho chic and denim miniskirts. In addition, the fashion world rediscovers India as a source of fashion sophistication, and nerdy geek chic takes hold.

80s Revival - Part Two

While the first revival of the eighties took place at the beginning of the decade, the trend now goes to the bright colors of the late eighties. Headbands, knit sweaters, denim skirts with tights or the Wonderbra are back. XXL shirts, spandex leggings, cuffs and crop tops are back, as are Hawaiian shirts, V-neck sweaters and black leather jackets.

Asian Fashion - Harem Pants and Mangas

Ethnic fashion from India and the Middle East is back in style for summer 2007. This includes harem pants, sarongs, saris and even the kurta. This trend is not only inspired by Bollywood movies, but also by pop stars like Shakira, who has achieved international success with her belly dancing. Younger women, in particular, are also inspired by Japanese pop culture with manga and J-pop, wearing sailor dresses, Lolita outfits and petticoats with knee socks.

Slim Fit Suits

Toward the end of the decade, the cut of the classic suit changed almost everywhere in favor of a slimmer, more body-hugging fit, while in the U.S. the broad-shouldered power suit made a brief comeback.

DIY Fashion - Handmade is in

As an evolution of the vintage look into boho chic, and fueled by the financial crisis of 2008, there is a growing trend toward DIY. Handknit sweaters are all the rage, but scarves and homemade clothing are also in.

Geek Chic - Grandparents' nerdy look

This look is a bit nerdy and always looks as if the wearer has raided their grandparents' wardrobe. Big glasses, horn-rimmed glasses are just as trendy as buttoned-up cardigans and loose-fitting corduroys.

Beauty Trends in the 2000s

Of course, the devil is in the details when it comes to creating the perfect outfit. You have to make sure that your hairstyle is up to date and that your shoes and accessories are a hit or a miss when it comes to joining the elite circle of trendy it-girls and -boys.

Hair - versatile with colored streaks

In the early 2000s, women tend to wear their hair long and straight, with lengths between the ears and the shoulders. Streaks provided a sunny look before Kelly Clarkson started a trend in 2002 with much more visible streaks. Clips and barrettes complete the look. Farrah Fawcett's feathered hairstyle is still seen at the beginning of the decade, while stars like Jennifer Lopez set trends with ballet-inspired updos. Rihanna, Alicia Keys and Janet Jackson are confidently sporting cornrows, dreadlocks and shorter versions of the "afro. The bob is absolutely en vogue in the mid-2000s - from around 2007, mainly in deep black and, towards the end of the decade, with a side parting rather than the out-of-date center parting.

In the early 2000s, men often wear short haircuts, some in the style of pop punk and boy bands, some in spiked styles, some in Iroquois cuts, and some with bleached ends, until natural hair colors come back into fashion at the end of the decade. From 2003 to 2006, the hairstyles of the mods and the long hair of the surfers became more common. By the end of the decade, British men were increasingly using pomade to hold their stylish short haircuts in place. Sideburns and goatees are also popular as a counter-movement to the metrosexual look of previous years. And, of course, there is the iconic Justin Bieber hairstyle, which has been an absolute statement hairstyle since 2008.

Make-up - as versatile as the decade

At the beginning of the new millennium, the glittery makeup of the 90s still dominates as part of the futuristic and bizarre Y2K trend. As an alternative, brown and purple shades are more likely to be combined with lip gloss than lipstick. After the glittery Y2K makeup disappears from the scene around 2004, the bold styling with bright red lips and cat eyes from the 1940s makes a comeback. Soon, however, the smoky eye disappears in favor of eyeliner, mascara and eye shadow. Toward the end of the decade, Lady Gaga heralded a new trend toward more artificiality with false eyelashes, which were further enhanced by eyelash extensions and tinting.

In addition, the demand for self-tanning products skyrocketed during this period, a trend that will continue until about 2009. Botox and waxing for both men and women also reflect the desire for the perfect body.

Jewellery and Accessories - Details for the Party Outfit

Popular accessories include mini shoulder bags (called baguette bags), aviator or other oversized sunglasses, and oversized earrings. Livestrong silicone wristbands are popular, as are trucker baseball caps, preferably Von Dutch. Crosses and rosaries, as well as shutter shades - sunglasses with striped plastic instead of lenses, which Kanye West made into an absolute trend - complete the party styling.

Necklaces and bracelets made from shells are perfect for the summer look. Wide decorative belts are reminiscent of the '70s, while eye-catching necklaces are the perfect complement to elegant evening wear. A slender chain belt also works well.

Trends from the '90s, such as the G-Shock, which was a must-have at the beginning of the decade, continue. Wristbands and headbands are no longer limited to the fitness context, but are worn by hip-hop stars such as Nelly and by the emo scene as a fashion accessory.

Tattoos and piercings - a trend that began in the 90s continues into the 2000s. Tribal motifs remain as popular as donkey antlers, but by the end of the decade, Chinese characters, Sanskrit, and, finally, an old-school Ed Hardy look have replaced the large motifs.

Fedoras - felt hats with wide brims that fold down slightly at the front - fit both men and women, judging by Britney Spears, Justin Timberlake and company.

Tn the 2000s, practical messenger bags became popular in parallel with fixie bikes. Carried over the shoulder on one side, these bags replaced the classic backpack, only to be replaced by it.

Originally designed to protect against the cold, scarves have found their way into everyday styling and are often part of a skinny jeans, blazer and shirt outfit. They also serve as headwear to keep hairstyles in place.

Shoes - Mostly Flat and Comfortable

Flip-flops in every imaginable color and shape are undoubtedly one of the enduring fashion trends of the 2000s. Sneakers by Nike, Adidas or Reebok are still in fashion as part of the 80s revival. Ugg boots in suede with lambskin lining and PVC ballerinas are also popular. The high heels that were so in at the beginning of the decade are replaced by flat ballerinas, pumps with small heels, gladiator sandals, boat shoes or chucks, and opulent platform shoes. Angular toe tips often provide that certain something.

Men in the 2000s wear Budapest , sneakers, Converse All Stars or pointed Winklepickers. In any case, the choice of the right brand is very important to be really "in". It is not uncommon for the entire outfit to match the color of the shoes and not vice versa.

Fashion Icons of the 2000s - Popstars Instead of Models

While supermodels like Naomi Campbell and Claudia Schiffer set the fashion trends in the 90s, in the 2000s it is music, film and television stars who are setting the style. The following list is just a small selection of the most fashionable trendsetters.

• Gisele Bündchen
• Paris Hilton
• Britney Spears
• Avril Lavigne
• Chloë Sevigny
• Nicole Richie
• Lindsay Lohan
• Amy Winehouse
• Hillary Duff
• Sarah Jessica Parker
• Kirsten Dunst
• Kim Kardashiann
• Jennifer Lopez
• Christina Aguilera
• Pharrell Williams
• Liya Kebede
• Miranda Kerr

The Most Popular Designers of the 2000s

High fashion provides the role models that soon reach the mass market through fast fashion. At the same time, the phenomenon of collaborations between the world's top designers and renowned fashion brands such as Puma, Diesel and even H&M appears for the first time in the 2000s.

Christian Audigier – Ultimate Trendsetter of the 2000s

The French designer made his breakthrough as head designer of the Von Dutch brand, setting a fashion trend with trucker baseball caps that didn't even stop at Hollywood stars and celebrities. Even more popular in the 2000s, however, was his collaboration with tattoo artist Don Ed Hardy, which resulted in the Ed Hardy label, which seems to have become omnipresent this decade with its large, colorful prints.

Karl Lagerfeld – Entry into Fast Fashion

Karl Lagerfeld has not only been making a name for himself since the 2000s, as he has been head designer at Chanel since the 1980s. But in 2004, the sought-after eccentric became the first top designer to launch a collection at H&M, crossing the line between the expensive, exclusive designer fashion of haute couture and the cheap, mass-produced goods of fast fashion. No wonder the first collection sold out in no time.

Marc Jacobs – Revival of the Eighties Dresses

In 2009, the designer brings the party dresses of the 80s back to the runway. But as creative director of LVMH, he no longer just dresses celebrities, he also develops his own lines, making him a celebrity in his own right. As a result, he was able to recruit personalities such as Stephen Malkmus and M.I.A. for his campaigns.

Alexander McQueen – On Stage with Lady Gaga

The British designer broke through with his own fashion label in the early 2000s. In 2006, he began a sports collaboration with Puma, but it was his stage outfits for Lady Gaga that made the biggest splash. In 2010, the world-famous designer, one of the most important icons of his time alongside John Galliano and Vivienne Westwood, took his own life.

Diane Von Furstenberg – Comeback of the Wrap Dress

The designer's wrap dress made a comeback in 1997. Celebrities like Gwyneth Paltrow and Jade Jagger wore it on red carpets before it returned to stores in 2003 due to its growing popularity.

Fashion Trends of the 2000s - A Continuation of the Nineties

As is often the case, music cultures influence fashion trends beyond their own scenes. In the 2000s, the hip-hop style is still present in the mainstream, and the distinctive ravers are still making a statement. In addition, indie and emo are revitalizing the fashion world.

Rave Culture - Tattoos, Piercings and Plastic

The fashion trends of the 1990s rave culture continue into the 2000s. Bright colors and spiky hairstyles, colorful dreadlocks or even bald heads are still popular. Piercings and tattoos, midriff-baring tops and latex pants that emphasize the toned body and bare skin are still characteristic.

Indie and Emo - A Continuation of Brit-Pop Plus Hipness

Indie pop from England and North America has been conquering the charts - and the fashion world - since the mid-2000s. The stars of the scene are inspired by the fashion culture of the 1960s, as well as by second-hand clothes or clothes from the Salvation Army. Band shirts with blazers, skinny jeans and narrow ties are just as much a part of the indie style as black-painted nails and cardigans. Closely related is the hipster fashion that developed in the USA at the end of the 2000s. Here, vintage clothing is more an expression of an ironic statement than a symbol of counterculture.

Hip Hop - Biker Jackets, Sneakers and Tracksuits

Hip-hop culture made its fashion mark in the 90s by entering the mainstream. In the 2000s, the culture experienced a retro trend with baggy jeans, basketball jackets, sneakers, sweatpants and tracksuits, baseball caps, down jackets and quilted jackets, and trucker baseball caps. White youth, in particular, imitate the gangsta lifestyle with expensive designer clothing, gold jewelry, and dark jeans, but they rarely achieve the street credibility of real gangsta rappers.

What started out as a continuation of the 90s quickly turns into a much more conservative look with 09/11. With the advent of new technologies - first and foremost the Internet with platforms like MySpace and YouTube - the importance of celebrities in spreading new fashion trends increases enormously. And with the constant availability of online stores, shopping habits are also changing - fashion is becoming faster, cheaper and more global. As a result, hardly any trend of the body-conscious decade lasts longer than two years, always with a nod to past fashion decades.

32 '00s Trends We Love

It wasn't all bad!

The 2000s were pretty maligned as a fashion decade. That's not to say that there weren't some tough sartorial moments—in the late '90s, stars were starting to use the red carpet as their personal runway, and the '00s really upped the ante for shock value. But leaving aside some of the barely-there minis with Uggs and other strange combos we favored, there are some fun style pieces we can lift out of the decade. In fact, the "dressing like it's Y2K/the 2000s" aesthetic came back into popularity in the early 2020s, with people picking the best, most iconic pieces from that decade and giving them a modern spin. I realize this might sound ridiculous—but some '00s fashion trends really can become a part of our classic capsule wardrobes.

Below, 32 '00s trends that we love.

There were a few notable "red carpet hat trends" in the '00s, including the newsboy hat, beret, and trucker hat, but none was more popular—at least as far as Britney Spears was concerned—as the humble fedora. And, if we're being totally honest, she rocked it.

Leaving aside the fact that the low-rise pant isn't for everyone, the '00s took our '90s love of flare pants and took it up a notch—usually with a wider, longer length at the bottom. We've loved flares since the '70s, but this version is a particularly nice update on a classic.

Pants as an Under-Layer

This trend is, in a word, divisive: Plenty of the "'00s worst trends" lists include the mandatory "formal dress over a pair of jeans" look. But it doesn't mean this trend is impossible to achieve; As with Heidi Klum here, the trick is to pick two pieces that look intuitive together, like this shirt dress worn open.

Shorts and Wellies

Was there anything more iconic than Kate Moss attending the Glastonbury Festival in short shorts and wellies? In the '00s, this was the fashion girl look. There's a particular context for this outfit (it's not something you'd wear to a nice dinner), but for mud or grass, this is perfect.

The crop top existed long before the '00s got ahold of it, but this was the crop top decade. We have some...interesting deployments of it, but we also have Mariah Carey looking adorable here in a color blocked outfit. Whether that particular item of clothing is "in" or "out," the '00s has plenty of inspo if you like it.

Stay In The Know

Marie Claire email subscribers get intel on fashion and beauty trends, hot-off-the-press celebrity news, and more. Sign up here.

Cropped Jackets

The '00s were often about layering in interesting ways, including a short jacket with a longer top underneath. A more modern iteration of this look would be to layer a crop top under a cropped jacket and then utilize some high-waisted jeans, then keep the accessories and pumps.

Chunky Belts

Frankly, I love this heart belt on Jennifer Lopez and would wear this look tomorrow. (Rhinestones: Another '00s trend we like!) A statement belt is a classic element for your everyday wardrobe, but when it's whimsical or silly, you just want to keep the rest of the look neutral.

Long Scarves

The scarf as accessory, instead of warm weather staple, was a '00s mainstay. We love the edgy skull pattern here, and the only update to make would be to wrap it around your neck a couple times and match it with something else that's orange in your look—your shoelaces, nail polish, or similar.

Chunky Jewelry

Obviously, chunky jewelry has been around forever, but the '00s saw us going with mixed materials—metals with glass, plastic, or crystal, as with Alexa Chung and her skull-adorned necklace here. The big clear drops might look dated, but it's okay to mix silver and gold in one piece.

Graphic Tees

Paris Hilton, rocking this Forrest Gump-esque shirt, is demonstrating how a graphic tee can be a fun upgrade from a classic shirt. Here, Hilton is probably making a snarky comment to paparazzi, but I kind of love rocking this kind of positivity in your everyday wardrobe.

Midi Skirts

The '00s trend everyone loves to discuss is the mini, but the midi skirt got some serious love too. We often saw the '00s skirts end around knee-length, which can be tricky to pull off, but when it was worn below the knee with boots that came up to match, it looks especially fab.

Oversized Sunglasses

Oversized shades have been with us almost as long as we've had fashion sunglasses, but the '00s took the look and made it wonky: Weird shapes, bedazzled frames, mismatched with the overall look (as with Kirsten Dunst) here: It was all happening. Pick a "surprising" pair and wear one today!

Even though you might associate the romper with the '10s, it began to take hold in the '00s. (After all, if J. Lo wears one, that's usually a sign it's about to become popular!) Thanks to the glittery accessories and sleek hair, the look doesn't have to skew too "youthful."

Bucket Hats

Never has there been an item of headgear more discussed than the bucket hat. If you love it, you've been waiting for the right moment to bring it out. If you hate it, you'll never wear it. So let your own preferences be your guide, and just break it out when it's not too cold.

Granted, this is a costume from Sex and the City , which is arguably one of the more fashionable shows in modern history (so this is not your average streetwear look, in other words). But I still love it as a layering option—whether you're matching it to your pants or just throwing it on over a tee.

Handkerchief Skirts

The handkerchief skirt can be a maxi, midi, and miniskirt all in one garment—but, to keep it from looking dated, make like Gabrielle Union here and keep the hem generally one length. Then, keep the rest of the look simple and neutral, letting the skirt be the standout.

Low-Rise Jeans

You either loved or hated the low-rise jeans of the '00s (I happen to be in the latter category because I have thin hips). Thankfully, jeans have evolved in the intervening years, and the new low-rise pants won't just slide off your body—and they're not quite as low as they used to be.

The pleated mini had a real redemption arc in the early 2020s, especially via influencers on TikTok and Instagram. This look on Rihanna could absolutely work today, but if you'd rather not be so ab-forward, just take a long chunky knit top and layer it over your skirt.

Studded/Patterned Denim

Destiny's Child made it cool to match your outfit with your bffs. Actually, there's not much you'd need to do to update these looks; If you like a longer length on your denim, choose a pair of jeans with a longer inseam and no slits. And you can always go for bedazzling in your tops!

Chunky Boots

Chunky boots never went out of fashion, per se, it's just about how you wear them. The '00s gave us surprise boot pairings, as with this more feminine, folded miniskirt on Gwyneth Paltrow . To modernize, just make something in the rest of you look bulkier and black, like a blazer.

Cargo Pants

This look on Britney is just so cute, and she's actually balancing out top and bottom effectively. Using a faux fur jacket and top that's feminine, then leaning into a chunkier, heavier vibe in the pants, then taking it back to delicate in the shiny shoes is actually cohesive.

A shorter jacket and a long top underneath: Peak '00s! There's ways to make this look dated, but there's also ways to make it look classic. Keep the length of the tunic not too long; It should just peek out underneath the jacket. Opt for a pair of pants in a similar color.

Rhinestones

We really wanted to sparkle in the '00s, ya know? Nelly Furtado here has sparkly nails, eyeshadow, earrings, jewelry, AND outfit. You don't need to do that much (unless you want to!) so pick one or two and really lean in. An easy way is to grab a bedazzled tee and just rock it.

Wallet Chains

Much like a chunky belt, the wallet chain is one of those surprising, counterintuitive pieces that, when deployed effectively, can bring a look together. Britney Spears wears hers messily, but the key is to have yours look sleek (and match with your belt, as she does).

Trucker Hats

The trucker hat of the '00s was made graphic and feminized so as not to look too jarring with the rest of an outfit (it could still look jarring, tbh, but that's another story). Since it's an unusual shape, match the color to your shoes to tie the look together. Or, go for a heavy baseball cap instead.

The tube top was the slightly more conservative older sister to the crop top—like, if you didn't want to show that much stomach, the tube top was your friend. It's a fabulous layering piece; Seriously, put it under a big, heavy blazer and see how great it looks.

Platform Shoes

While never "out" of fashion, it felt like we got tired of platforms after wearing them every day in the '00s. But the platform espadrille, as Kate Moss is wearing here, is truly a timeless piece. And if you don't want to lean in on the chunkiness, just opt for a sleeker shape.

Boyfriend Jeans

Long, slouchy, tapered, cuffed: The boyfriend jean was everywhere in the '00s. Now, it's basically just considered a classic piece of denim, no different than the skinny or wide-leg, but the boyfriend jeans of today also come in high-rise and more tailored shapes.

Butterflies

It didn't really matter where: On tops, pants, shoes, and accessories, butterflies were everywhere. (And we got a little tired of them by the end of the decade.) Now, as with everything else on this list, it's all about moderation. A little butterfly clip is the cutest touch to your look.

Shoulder Bags

Whether sleek and structured or soft and colorful, a shoulder bag was on everyone's shoulder in the '00s. If you like shoulder bags already, you'll know that you can wear them anytime. But if you'd like to dip a toe into the style, go with a big similar to Hilton's: simple, neutral color, sleek.

Chunky Necklaces

Why wear one of your necklaces from your jewelry box when you can wear allllll of them? Kate Beckinsale is giving us the blueprint for how to make it work: Have all the pieces kind of go together, and use varying lengths so they look complementary and feel delightfully jumbled.

Column Dresses

To a degree, the Calvin Klein-esque column dress of the '90s was replaced by more body-con styles in the '00s and '10s. But the red carpet (and sometimes runway) still gave us some gorgeous iterations of the classic style. You could wear this anytime—just hem it if you don't like the pooling at the bottom.

Katherine’s a contributing syndications editor at Marie Claire who covers fashion, culture, and lifestyle. In her role, she writes stories that are syndicated by MSN and other outlets. She’s been a full-time freelancer for over a decade and has had roles with  Cosmopolitan  (where she covered lifestyle, culture, and fashion SEO content) and  Bustle  (where she was their movies and culture writer). She has bylines in  New York Times ,  Parents ,  InStyle , Refinery29, and elsewhere. Her work has also been syndicated by  ELLE ,  Harper’s Bazaar ,  Seventeen ,  Good Housekeeping , and  Women’s Health , among others. In addition to her stories reaching millions of readers, content she's written and edited has qualified for a Bell Ringer Award and received a Communicator Award. 

Katherine has a BA in English and art history from the University of Notre Dame and an MA in art business from the Sotheby's Institute of Art (with a focus on marketing/communications). She covers a wide breadth of topics: she's written about  how to find the very best petite jeans ,  how sustainable travel has found its footing on Instagram , and  what it's like to be a professional advice-giver in the modern world . Her personal essays have run the gamut from  learning to dress as a queer woman  to  navigating food allergies as a mom . She also has deep knowledge of SEO/EATT, affiliate revenue, commerce, and social media; she regularly edits the work of other writers. She speaks at writing-related events and podcasts about freelancing and journalism, mentors students and other new writers, and consults on coursework. Currently, Katherine lives in Boston with her husband and two kids, and you can follow her on  Instagram . If you're wondering about her last name, it’s “I go to dinner,” not “Her huge ego,” but she responds to both.

Meanwhile, Meghan’s former ‘Suits’ co-star and onscreen love interest Patrick J. Adams teased a major new development for the brand over the weekend.

By Rachel Burchfield Published 3 June 24

"It just keeps getting better and better."

By Halie LeSavage Published 3 June 24

After a brief hiatus, the singer is carrying her collection once again.

• Contact Future's experts
• Advertise Online
• Terms and conditions
• Privacy policy
• Cookies policy

Marie Claire is part of Future plc, an international media group and leading digital publisher. Visit our corporate site . © Future US, Inc. Full 7th Floor, 130 West 42nd Street, New York, NY 10036.

2000's Fashion

Fashion in the 2000's

The fashion styles of the 2000's are heavily influenced by the entertainment world around them. From the world of sports, music and television the 2000's was a time of great commercial success - especially in the early 2000s. For men the look mimics much of the style trends of bands like NSync and worked to conform to social trends of the time, by the mid-2000's the Yuppie look of the 1980's was back - with a new twist. Rugby shirts and popped collars were heavily in style with an emphasis on name brand apparel to show your wealth. However, by the end of the 2000's the United States was in the midst of a severe economic downturn where the look of men began to become more subdued and refined.

Women's fashion during the 2000's was about emulating the pop stars of the day. Women were no longer looking to the fashion industry for insight, they were looking to the musicians and television stars they watched and listened to every day. Crop tops, short skirts and body jewelry was increasingly popular; however, this look was countered with baggy camo printed pants and jackets that ushers in a more relaxed bohemian look of the 1960's and 1970's. This look also brought with it a new shoe trend - the Ugg boots which became the go to shoe for celebrities and young women alike. Additionally , hair highlights also became increasingly popular.

By the time the 2000's came to an end fashion was in a confused state, it was a blend of the television and entertainment boom and the desire to be like the celebrities we watched everyday.

These trends eventually lead to an explosion of individuality we experience throughout the 2010s.

Task 1: Watch the Introductory Video to this Unit (5 minutes)

Task 2: review the provided images and attached links for fashion during the 2000's (20 minutes), task 3: go to google classroom and respond to the following prompts (25 minutes).

Respond to prompts using complete sentences and appropriate punctuation and grammar

What are TWO (2) similarities and TWO (2) difference you see in the fashion trends of the 2000's compared to the fashion trends from the 1990's

Look at this area of pop culture as a whole

What are the general trends you see from decade to decade

What new trends emerged in this decade that are either completely unique or built off a previous idea/idea/way if thinking

You can list or similarities and differences or write them out using complete sentences.

Why do you think th e similarities you selected carried over from the previous decade

Is it cultural, technological, innovative, something else

Do you think it will last much longer than this decade

What is ONE (1) fashion trend that emerged in the 2000's that you think will continue on to the next decade

You can select anything from this area of pop culture that emerged during the decade

Men's Fashion

Technology and Y2K had a huge impact on fashion in the early 2000s. The color palette was filled with shiny black tones and reflective metallics. While some Y2K trends were worn daily, many of these looks were reserved for going out. For men, Y2K looks usually involved leather jackets, a statement dress shirt, and chunky shoes. If you need inspiration, *NSYNC were the trendsetters of the day . Casual clothing and leisurewear were the other big trends of the early 2000s. Denim became a staple for men and women, going beyond jeans to shirts, jackets, and hats. ( https://www.centralcasting.com/quick-guide-2000s-fashion/ )

Check out more information about Men's Fashion in the 2000's here

Women's Fashion

In many ways, the 2000s were a decade where the celebrity obsession peaked. Instead of models being fashion trendsetters, celebrities as Paris Hilton and Nicole Richie were style icons. The early years of the 2000s saw clothes as: crop tops, vests, blazers, halterneck tops, v-shaped jumpers, low waisted jeans and skirts. Layered tank tops in different colors were popular and should be matched with cargo pants. The flowy life also worked for the bohemian style that was trending and Ugg boots were all the rage. The 2000s w ere the decade of highlights and colored hair. The highlights weren’t to look natural but often contrasting one’s hair color. As in blonde hair and brown highlights or the other way around. ( https://www.thefashionfolks.com/blog/21th-century-fashion-history-2000-2010/ )

Check out more information about Women's Fashion in the 2000's here

• Close Menu Search
• Suggestions and Questions

The Highlighter

Gen Z and the Return of 2000s Fashion

How the younger generations are bringing back what’s considered one of the worst eras for fashion.

Lindsay Grimm , Staff Writer | September 16, 2020

Daryn Okada

This screenshot from the movie Mean Girls, shows one of the best movies in terms of showcasing 2000s fashion.

Along with being the start of a new decade, 2020 has brought about the revival of a handful of fashion trends from the early 2000s. Given the concept of the “20-year rule” of fashion, which states that something popular now will see a resurgence in twenty years, the comeback of the fashion from the 2000s is hardly a surprise. Y2K fashion, the style from the early 2000s, is heavily influenced by technology, globalization, and the emerging market for fast-fashion. However, given how despised trends from the 2000s seem to be on the internet, this revival of 2000s fashion can be mostly attributed to nostalgia. 

Generation Z and Millenials have recently started to make up a larger portion of the economy and are estimated to account for $350 billion of spending power in the U.S. By having these younger generations make up so much of the economy, fashion companies are starting to listen to their demands and create clothing and accessories tailored to their tastes and trends, Y2K fashion being one of them.

Logo prints, small camis, baggy pants, bright colors, cardigans, small purses, and fun hair accessories are a handful of the staple pieces seen in Gen Z’s revival of Y2K fashion. These trends are both created and spread on social media, most notably Instagram and TikTok. Videos under #Y2K on TikTok have collectively garnered 468.3 million views and videos #2000sfashion has 34.3 million views. 

Social media is undoubtedly to thank for the spread of these fashion trends, but what sparked the comeback is most likely a wave of nostalgia. The generations that grew up watching movies like Mean Girls , Legally Blonde , Bring It On , and other staple 2000s movies now have the money and ability to emulate the fashion in the movies they grew up watching. Rather than relying on their parents for money and clothes, Generation Z has reached the stage in which most of them have their own source of income and ability to style themselves. 

Although there is no doubt about the resurgence of 2000s fashion, this revival seems eager to leave some of the more gauche parts of 2000s fashion in the past. There will probably be no celebrities wearing denim dresses and suits on the red carpet, and nobody has tried to revive wearing dresses over jeans. 

Given the cyclical nature of fashion, it was only a matter of time until these 2000s trends made a comeback. I’m sure that by 2030, these trends that are on fire right now will be denounced as trashy pandemic-era fashion, and the skinny jeans and millennial pink of the 2010s will be all the rage.

Lindsay is a senior at Rocky and is passionate about video games and feminism. After high school, she plans to move to Los Angeles to attend college and...

Global Leaders and there work with PHS Food Pantry

The Duck Pond

Maddy Winokur – Her Student Teaching Journey

THE EMPORIUM SPORTS BAR

Senior Assassin Weekly

Invasive Species: A Problem at RMHS

What is rocky listening too.

Get employed! Summer is around the corner

Stepping into spring break in Fort Collins

The Bo’cycle Shop

PEERS Spread the Love

Ko-la What?

Colorado’s Very Own Clothing Company

Jayda Ratzlaff – A Teen Business Owner

The Community Experience of Rocky Mountain High School

Comments (0)

Cancel reply

Your email address will not be published. Required fields are marked *

Viral Chatter

The 2000s Are Back: 12 Fashion Trends Making a Comeback Today

Posted: June 4, 2024 | Last updated: June 4, 2024

Fashion, like history, has a tendency to repeat itself. The 2000s, with its distinctive style and bold choices, are experiencing a renaissance in the fashion world. From the runways of Paris to the streets of New York, these trends are being revived with a modern twist.

1. Low-Rise Jeans

Low-rise jeans, once the hallmark of early 2000s fashion, are making a fierce return. These hip-hugging pants were iconic for their daring cut and are now being embraced by a new generation. Designers are updating this trend with new fits and styles, making them more comfortable and versatile.

Pairing low-rise jeans with crop tops and oversized jackets creates a balanced look that is both nostalgic and contemporary. This trend is seen everywhere, from high-fashion runways to street-style influencers, proving its versatility and enduring appeal.

2. Tiny Sunglasses

Tiny sunglasses, popularized by celebrities like Paris Hilton and Lindsay Lohan in the early 2000s, are once again a must-have accessory. These small, often oval or rectangular frames add a chic, edgy vibe to any outfit.

Modern iterations of tiny sunglasses come in a variety of colors and lens tints, allowing for greater personalization. This trend is a favorite among influencers and celebrities, making frequent appearances in fashion magazines and social media feeds.

3. Cargo Pants

Cargo pants, with their multitude of pockets and utilitarian design, are back in style. This trend is not just about functionality; it’s about making a fashion statement. The 2000s saw cargo pants as a casual staple, and today, they are being reimagined in sleeker, more tailored fits.

Pairing cargo pants with fitted tops or blazers creates a balanced silhouette that combines practicality with style. This trend is ideal for those who want to blend comfort with a touch of rugged sophistication.

4. Velour Tracksuits

Velour tracksuits, famously worn by celebrities like Jennifer Lopez and the Kardashian sisters, are back in the limelight. These comfy yet stylish sets epitomize the 2000s leisurewear trend, and today, they are being embraced for their retro charm.

Modern velour tracksuits come in an array of colors and cuts, making them suitable for both lounging and streetwear. Brands are capitalizing on the nostalgia factor, reintroducing these tracksuits with contemporary designs that appeal to today’s fashion-savvy consumers.

5. Butterfly Clips

Butterfly clips, the quintessential 2000s hair accessory, are fluttering back into fashion. These colorful, whimsical clips add a playful touch to any hairstyle and are perfect for channeling a carefree, youthful vibe.

Today’s butterfly clips are often used to create intricate hairstyles or to add a pop of color to simpler looks. They have been spotted on fashion runways and in everyday street style, proving that this accessory is both versatile and timeless.

6. Platform Sneakers

Platform sneakers, which gave us extra height and a boost of confidence in the 2000s, are making a triumphant return. These shoes are being revamped with modern materials and innovative designs, making them a fashionable choice for today’s trendsetters.

Whether paired with jeans, skirts, or dresses, platform sneakers add a bold statement to any outfit. This trend is a favorite among younger fashion enthusiasts who appreciate its blend of comfort and style.

7. Denim Skirts

Denim skirts, especially the mini versions, were a staple of 2000s fashion and are now back with a vengeance. These skirts are being reimagined with new cuts, embellishments, and washes, making them a versatile addition to any wardrobe.

Pairing denim skirts with tights, boots, or oversized sweaters creates a look that is both modern and nostalgic. This trend is a great way to embrace the 2000s style while keeping it fresh and relevant.

8. Trucker Hats

Trucker hats, once synonymous with casual, laid-back style, are experiencing a resurgence. These hats are being updated with new designs and logos, making them a popular accessory for both men and women.

Trucker hats add a sporty, casual element to any outfit and are perfect for bad hair days or adding a touch of 2000s flair. They are versatile enough to be worn with anything from athleisure to casual streetwear.

9. Slip Dresses

Slip dresses, with their delicate, minimalist design, were a 2000s staple and are now back in full swing. These dresses are being styled in various ways, from casual daywear to elegant evening outfits.

Layering slip dresses over t-shirts or under blazers creates a modern look that is both chic and effortless. This trend is perfect for those who appreciate the simplicity and elegance of 2000s fashion.

10. Chunky Highlights

Chunky highlights, the bold and striking hair trend of the 2000s, are making a return. This hair trend involves thick, contrasting streaks of color, often in blonde or bright hues, creating a dramatic look.

Modern interpretations of chunky highlights are more refined, often blending more seamlessly with the natural hair color. This trend is a fun way to experiment with hair color and add a touch of 2000s flair to your look.

11. Statement Belts

Statement belts, a key accessory in the 2000s, are back in vogue. These belts, often oversized and adorned with large buckles, are a focal point of any outfit.

Wearing a statement belt over dresses, blazers, or high-waisted pants can cinch the waist and add an element of interest to your ensemble. This trend is perfect for those who want to add a bold, stylish touch to their outfits.

12. Logomania

Logomania, the trend of wearing clothing emblazoned with brand logos, is experiencing a comeback. This 2000s trend is all about making a statement and showing off your favorite brands in a bold way.

Today’s logo mania is often seen in streetwear and high fashion, with brands reintroducing their iconic logos in new and creative ways. This trend is ideal for those who love to make a statement with their fashion choices.

The resurgence of these 2000s trends proves that fashion is cyclical and that what goes around comes around. Whether you’re reliving your teenage years or discovering these styles for the first time, there’s no denying the charm and appeal of 2000s fashion.

More for You

If you have this much money in your 401(k), you're doing better than most—here's how to save even more

Reacher Season 3: Release, Plot, Cast & News

I found Dad washing the car with a paintbrush. His dogs knew exactly what was happening

World’s biggest solar farm goes online, big enough to power a country

5 Rules Every Jeep Owner Needs To Know About 'Ducking'

Claims that Donald Trump is no longer able to visit 37 countries now that he’s a convicted felon need context

42-year-old yacht captain's side hustle brings in $124,000 a year: ‘It's the easiest, simplest thing to do’

Do you get a jolt when you're falling asleep? Here's the reason behind that and other strange bodily reactions.

"I Texted My Dad To Immediately Pick Me Up": 23 People Are Recalling The Bizarre "House Rules" They Experienced At Someone Else's Home When They Were A Kid

Enormous Great White Shark Found Washed Up on a Beach After an Even Larger Sea Animal Killed It

BBC’s new Agatha Christie drama is here - and it sounds brilliant

I tried Red Lobster's $30 endless shrimp for the first time. I see why the deal might have contributed to the chain's bankruptcy.

Matt Gaetz Splits With Donald Trump

Tipping frustration reaching the boiling point

Average US household income ranked by state – see how you compare

Adam Levine is returning to 'The Voice' as a coach, joined by newcomer Kelsea Ballerini

Man lives debt-free in cob home that only cost $200 to build: ‘It’s just beautiful’

Here's Why European Trucks Look So Different From What We See In America

Fighter Jets Land on Highway in Emergency Training Exercise

Virginia governor says state will abandon California emissions standards by the end of the year

Capri Pants Are Back: The 15 Best Pairs to Shop Now

I never thought I would be the one to tell you that capri pants are having a resurgence. This summer, a style that many millennials vowed to never wear again suddenly feels cool again. Their popularity might be, at least in part, another symptom of the early aughts nostalgia wave that continues to flow through the fashion world, but there’s something else there. Maybe it’s a backlash to the baggy jeans and wide leg trousers that have dominated our wardrobes lately. A slim cut pair of capri pants feels fresh and feminine, more flattering than a pair of shorts and way more put-together than a legging—and, to give the style a bit more credit, it actually originated in the late 1940s, becoming popular in the ’50s and early ’60s. So there are more classic ways to wear them that feel crisp and fresh. With so many brands jumping on the bandwagon, there are plenty of options to choose from, including a bold printed pair by Connor Ives and a sleek and simple set from Wardrobe NYC. Scroll through our edit of the best capris to wear this summer.

• Asia Pacific
• Middle East
• United Kingdom
• United States
• The Great Reboot
• Reuters Next
• Aerospace & Defense
• Autos & Transportation
• Environment
• Healthcare & Pharmaceuticals
• Media & Telecom
• Retail & Consumer
• Sustainable Business
• Future of Health
• Future of Money
• Reuters Impact
• Legal Industry
• Transactional
• Asian Markets
• Carbon Markets
• Commodities
• Emerging Markets
• European Markets
• Global Market Data
• Rates & Bonds
• U.S. Markets
• Macro Matters
• Reuters Momentum
• Motor Sports
• Oddly Enough

THE WIDER IMAGE

In albania, two women take on a nation with a rooftop wedding.

Edlira Mara (left), Alba Ahmetaj (centre), 44, and their twin daughters get ready for their wedding ceremony in central Tirana, Albania, May 19, 2024. REUTERS/Florion Goga

By Florion Goga

Filed May 31, 2024, 11:09 a.m. GMT

Photography by Florion Goga

Writing by Fatos Bytyci and Florion Goga

Filed June 03, 2024, 08:00 a.m. GMT

In many ways Alba Ahmetaj and Edlira Mara lead an ordinary life. They brush their twin daughters’ hair before school and play fight with them in their flat on weekend mornings. They have matching shoulder tattoos that mark their 14 years together.

But in their fight to be treated like other families, the lesbian couple did something extraordinary.

At dusk on Sunday, May 19, friends cheered as they stood out on the rooftop of the mayor’s office in central Tirana, kissed, exchanged rings and got married.

Their marriage is not acknowledged by the state - Albanian law does not recognise same-sex civil unions. It has prompted outrage from the political right and the powerful religious community.

But for Alba and Edlira, it was a real expression of love, a cry for equality and, as far as they know, the first wedding of its kind in the Muslim-majority Balkan country.

“There are two people in love ... and now they have finalised it with this beautiful ceremony,” Edlira said after the wedding. “Society will never be ready ... What does this mean? That I cannot live?”

While much of western Europe has made strides towards marriage equality, governments in much of the centre and east oppose change.

In Albania, religion was prohibited for half a century under communism. Today, the country is known for its tolerance among Muslims, Catholics and Orthodox Christians. These faiths are united in their opposition to same-sex marriage.

When plans for the couple’s wedding became public, social media was flooded with thousands of threatening comments. Police officers guarded the building during the ceremony.

Two days later, opposition parties held a protest against the mayor over separate corruption allegations. But the speakers turned on Alba and Edlira too, accusing them of destroying family values.

The furore over the wedding was just the latest stage in what the couple say has been a long struggle to get the same rights as heterosexual couples.

When their daughters were born three years ago, the couple said they both wanted to be registered as parents, but that was not allowed under law. The children are registered under Edlira, the biological mother.

“Our society is very patriarchal and homophobic,” Alba said before the wedding. “If you see comments on Facebook or Instagram ... you will see how little tolerance we have as a nation.”

One of her twin daughters hugs Edlira as their dog Rainy looks at her empty plate, during breakfast at home in Tirana, Albania, May 18, 2024. REUTERS/Florion Goga

Alba does her daughter’s hair at home in Tirana, Albania, May 18, 2024.  REUTERS/Florion Goga

Alba, Edlira and their twin daughters have breakfast at home in Tirana, Albania, May 18, 2024. REUTERS/Florion Goga

Alba and one of her twin daughters play at home in Tirana, Albania, May 18, 2024.  REUTERS/Florion Goga

Edlira and one of her twin daughters feed their cat Pufi through the window of their apartment, in Tirana, Albania, May 18, 2024. REUTERS/Florion Goga

Alba and Edlira pick up their twin daughters from nursery in Tirana, Albania, May 17, 2024. REUTERS/Florion Goga

Alba picks up one of her twin daughters, at a park in Tirana, Albania, May 18, 2024.  REUTERS/Florion Goga

Alba inflates balloons for her twin daughters as Edlira laughs, a day before their wedding ceremony, at home in Tirana, Albania, May 18, 2024. REUTERS/Florion Goga

Alba and Edlira make arrangements for their wedding, a day before the ceremony, at home in Tirana, Albania, May 18, 2024.  REUTERS/Florion Goga

Alba and Edlira’s twin daughters play at their apartment in Tirana, Albania, May 18, 2024.  REUTERS/Florion Goga

Amid the struggle, the wedding became a bright spot. To prepare, the family blew up balloons in their apartment. On the evening of the ceremony, friends helped the couple into their big white wedding dresses.

They walked, holding their daughters’ hands, towards the altar through a crowd of friends who threw white rose petals. Around them were the sights of downtown Tirana and mountains beyond, covered in mist.

They were wed by two British pastors.

“We are fighting against 90 percent of the population,” Edlira said. “Both of us are changing a lot of things.”

The Wider Image

Photography: Florion Goga

Writing: Fatos Bytyci and Florion Goga

Photo editing and design: Marta Montana Gomez and Maye-E Wong

Text editing: Edward McAllister and Andrew Heavens

LICENSE THIS STORY

• Follow Reuters Investigates

Other Reuters investigations

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 03 June 2024

Applying large language models for automated essay scoring for non-native Japanese

• Wenchao Li 1 &
• Haitao Liu 2  

Humanities and Social Sciences Communications volume  11 , Article number:  723 ( 2024 ) Cite this article

12 Accesses

1 Altmetric

Metrics details

• Language and linguistics

Recent advancements in artificial intelligence (AI) have led to an increased use of large language models (LLMs) for language assessment tasks such as automated essay scoring (AES), automated listening tests, and automated oral proficiency assessments. The application of LLMs for AES in the context of non-native Japanese, however, remains limited. This study explores the potential of LLM-based AES by comparing the efficiency of different models, i.e. two conventional machine training technology-based methods (Jess and JWriter), two LLMs (GPT and BERT), and one Japanese local LLM (Open-Calm large model). To conduct the evaluation, a dataset consisting of 1400 story-writing scripts authored by learners with 12 different first languages was used. Statistical analysis revealed that GPT-4 outperforms Jess and JWriter, BERT, and the Japanese language-specific trained Open-Calm large model in terms of annotation accuracy and predicting learning levels. Furthermore, by comparing 18 different models that utilize various prompts, the study emphasized the significance of prompts in achieving accurate and reliable evaluations using LLMs.

Similar content being viewed by others

Scoring method of English composition integrating deep learning in higher vocational colleges

ChatGPT-3.5 as writing assistance in students’ essays

Detecting contract cheating through linguistic fingerprint

Conventional machine learning technology in aes.

AES has experienced significant growth with the advancement of machine learning technologies in recent decades. In the earlier stages of AES development, conventional machine learning-based approaches were commonly used. These approaches involved the following procedures: a) feeding the machine with a dataset. In this step, a dataset of essays is provided to the machine learning system. The dataset serves as the basis for training the model and establishing patterns and correlations between linguistic features and human ratings. b) the machine learning model is trained using linguistic features that best represent human ratings and can effectively discriminate learners’ writing proficiency. These features include lexical richness (Lu, 2012 ; Kyle and Crossley, 2015 ; Kyle et al. 2021 ), syntactic complexity (Lu, 2010 ; Liu, 2008 ), text cohesion (Crossley and McNamara, 2016 ), and among others. Conventional machine learning approaches in AES require human intervention, such as manual correction and annotation of essays. This human involvement was necessary to create a labeled dataset for training the model. Several AES systems have been developed using conventional machine learning technologies. These include the Intelligent Essay Assessor (Landauer et al. 2003 ), the e-rater engine by Educational Testing Service (Attali and Burstein, 2006 ; Burstein, 2003 ), MyAccess with the InterlliMetric scoring engine by Vantage Learning (Elliot, 2003 ), and the Bayesian Essay Test Scoring system (Rudner and Liang, 2002 ). These systems have played a significant role in automating the essay scoring process and providing quick and consistent feedback to learners. However, as touched upon earlier, conventional machine learning approaches rely on predetermined linguistic features and often require manual intervention, making them less flexible and potentially limiting their generalizability to different contexts.

In the context of the Japanese language, conventional machine learning-incorporated AES tools include Jess (Ishioka and Kameda, 2006 ) and JWriter (Lee and Hasebe, 2017 ). Jess assesses essays by deducting points from the perfect score, utilizing the Mainichi Daily News newspaper as a database. The evaluation criteria employed by Jess encompass various aspects, such as rhetorical elements (e.g., reading comprehension, vocabulary diversity, percentage of complex words, and percentage of passive sentences), organizational structures (e.g., forward and reverse connection structures), and content analysis (e.g., latent semantic indexing). JWriter employs linear regression analysis to assign weights to various measurement indices, such as average sentence length and total number of characters. These weights are then combined to derive the overall score. A pilot study involving the Jess model was conducted on 1320 essays at different proficiency levels, including primary, intermediate, and advanced. However, the results indicated that the Jess model failed to significantly distinguish between these essay levels. Out of the 16 measures used, four measures, namely median sentence length, median clause length, median number of phrases, and maximum number of phrases, did not show statistically significant differences between the levels. Additionally, two measures exhibited between-level differences but lacked linear progression: the number of attributives declined words and the Kanji/kana ratio. On the other hand, the remaining measures, including maximum sentence length, maximum clause length, number of attributive conjugated words, maximum number of consecutive infinitive forms, maximum number of conjunctive-particle clauses, k characteristic value, percentage of big words, and percentage of passive sentences, demonstrated statistically significant between-level differences and displayed linear progression.

Both Jess and JWriter exhibit notable limitations, including the manual selection of feature parameters and weights, which can introduce biases into the scoring process. The reliance on human annotators to label non-native language essays also introduces potential noise and variability in the scoring. Furthermore, an important concern is the possibility of system manipulation and cheating by learners who are aware of the regression equation utilized by the models (Hirao et al. 2020 ). These limitations emphasize the need for further advancements in AES systems to address these challenges.

Deep learning technology in AES

Deep learning has emerged as one of the approaches for improving the accuracy and effectiveness of AES. Deep learning-based AES methods utilize artificial neural networks that mimic the human brain’s functioning through layered algorithms and computational units. Unlike conventional machine learning, deep learning autonomously learns from the environment and past errors without human intervention. This enables deep learning models to establish nonlinear correlations, resulting in higher accuracy. Recent advancements in deep learning have led to the development of transformers, which are particularly effective in learning text representations. Noteworthy examples include bidirectional encoder representations from transformers (BERT) (Devlin et al. 2019 ) and the generative pretrained transformer (GPT) (OpenAI).

BERT is a linguistic representation model that utilizes a transformer architecture and is trained on two tasks: masked linguistic modeling and next-sentence prediction (Hirao et al. 2020 ; Vaswani et al. 2017 ). In the context of AES, BERT follows specific procedures, as illustrated in Fig. 1 : (a) the tokenized prompts and essays are taken as input; (b) special tokens, such as [CLS] and [SEP], are added to mark the beginning and separation of prompts and essays; (c) the transformer encoder processes the prompt and essay sequences, resulting in hidden layer sequences; (d) the hidden layers corresponding to the [CLS] tokens (T[CLS]) represent distributed representations of the prompts and essays; and (e) a multilayer perceptron uses these distributed representations as input to obtain the final score (Hirao et al. 2020 ).

AES system with BERT (Hirao et al. 2020 ).

The training of BERT using a substantial amount of sentence data through the Masked Language Model (MLM) allows it to capture contextual information within the hidden layers. Consequently, BERT is expected to be capable of identifying artificial essays as invalid and assigning them lower scores (Mizumoto and Eguchi, 2023 ). In the context of AES for nonnative Japanese learners, Hirao et al. ( 2020 ) combined the long short-term memory (LSTM) model proposed by Hochreiter and Schmidhuber ( 1997 ) with BERT to develop a tailored automated Essay Scoring System. The findings of their study revealed that the BERT model outperformed both the conventional machine learning approach utilizing character-type features such as “kanji” and “hiragana”, as well as the standalone LSTM model. Takeuchi et al. ( 2021 ) presented an approach to Japanese AES that eliminates the requirement for pre-scored essays by relying solely on reference texts or a model answer for the essay task. They investigated multiple similarity evaluation methods, including frequency of morphemes, idf values calculated on Wikipedia, LSI, LDA, word-embedding vectors, and document vectors produced by BERT. The experimental findings revealed that the method utilizing the frequency of morphemes with idf values exhibited the strongest correlation with human-annotated scores across different essay tasks. The utilization of BERT in AES encounters several limitations. Firstly, essays often exceed the model’s maximum length limit. Second, only score labels are available for training, which restricts access to additional information.

Mizumoto and Eguchi ( 2023 ) were pioneers in employing the GPT model for AES in non-native English writing. Their study focused on evaluating the accuracy and reliability of AES using the GPT-3 text-davinci-003 model, analyzing a dataset of 12,100 essays from the corpus of nonnative written English (TOEFL11). The findings indicated that AES utilizing the GPT-3 model exhibited a certain degree of accuracy and reliability. They suggest that GPT-3-based AES systems hold the potential to provide support for human ratings. However, applying GPT model to AES presents a unique natural language processing (NLP) task that involves considerations such as nonnative language proficiency, the influence of the learner’s first language on the output in the target language, and identifying linguistic features that best indicate writing quality in a specific language. These linguistic features may differ morphologically or syntactically from those present in the learners’ first language, as observed in (1)–(3).

我-送了-他-一本-书

Wǒ-sòngle-tā-yī běn-shū

1 sg .-give. past- him-one .cl- book

“I gave him a book.”

Agglutinative

彼-に-本-を-あげ-まし-た

Kare-ni-hon-o-age-mashi-ta

3 sg .- dat -hon- acc- give.honorification. past

Inflectional

give, give-s, gave, given, giving

Additionally, the morphological agglutination and subject-object-verb (SOV) order in Japanese, along with its idiomatic expressions, pose additional challenges for applying language models in AES tasks (4).

足-が 棒-に なり-ました

Ashi-ga bo-ni nar-mashita

leg- nom stick- dat become- past

“My leg became like a stick (I am extremely tired).”

The example sentence provided demonstrates the morpho-syntactic structure of Japanese and the presence of an idiomatic expression. In this sentence, the verb “なる” (naru), meaning “to become”, appears at the end of the sentence. The verb stem “なり” (nari) is attached with morphemes indicating honorification (“ます” - mashu) and tense (“た” - ta), showcasing agglutination. While the sentence can be literally translated as “my leg became like a stick”, it carries an idiomatic interpretation that implies “I am extremely tired”.

To overcome this issue, CyberAgent Inc. ( 2023 ) has developed the Open-Calm series of language models specifically designed for Japanese. Open-Calm consists of pre-trained models available in various sizes, such as Small, Medium, Large, and 7b. Figure 2 depicts the fundamental structure of the Open-Calm model. A key feature of this architecture is the incorporation of the Lora Adapter and GPT-NeoX frameworks, which can enhance its language processing capabilities.

GPT-NeoX Model Architecture (Okgetheng and Takeuchi 2024 ).

In a recent study conducted by Okgetheng and Takeuchi ( 2024 ), they assessed the efficacy of Open-Calm language models in grading Japanese essays. The research utilized a dataset of approximately 300 essays, which were annotated by native Japanese educators. The findings of the study demonstrate the considerable potential of Open-Calm language models in automated Japanese essay scoring. Specifically, among the Open-Calm family, the Open-Calm Large model (referred to as OCLL) exhibited the highest performance. However, it is important to note that, as of the current date, the Open-Calm Large model does not offer public access to its server. Consequently, users are required to independently deploy and operate the environment for OCLL. In order to utilize OCLL, users must have a PC equipped with an NVIDIA GeForce RTX 3060 (8 or 12 GB VRAM).

In summary, while the potential of LLMs in automated scoring of nonnative Japanese essays has been demonstrated in two studies—BERT-driven AES (Hirao et al. 2020 ) and OCLL-based AES (Okgetheng and Takeuchi, 2024 )—the number of research efforts in this area remains limited.

Another significant challenge in applying LLMs to AES lies in prompt engineering and ensuring its reliability and effectiveness (Brown et al. 2020 ; Rae et al. 2021 ; Zhang et al. 2021 ). Various prompting strategies have been proposed, such as the zero-shot chain of thought (CoT) approach (Kojima et al. 2022 ), which involves manually crafting diverse and effective examples. However, manual efforts can lead to mistakes. To address this, Zhang et al. ( 2021 ) introduced an automatic CoT prompting method called Auto-CoT, which demonstrates matching or superior performance compared to the CoT paradigm. Another prompt framework is trees of thoughts, enabling a model to self-evaluate its progress at intermediate stages of problem-solving through deliberate reasoning (Yao et al. 2023 ).

Beyond linguistic studies, there has been a noticeable increase in the number of foreign workers in Japan and Japanese learners worldwide (Ministry of Health, Labor, and Welfare of Japan, 2022 ; Japan Foundation, 2021 ). However, existing assessment methods, such as the Japanese Language Proficiency Test (JLPT), J-CAT, and TTBJ Footnote 1 , primarily focus on reading, listening, vocabulary, and grammar skills, neglecting the evaluation of writing proficiency. As the number of workers and language learners continues to grow, there is a rising demand for an efficient AES system that can reduce costs and time for raters and be utilized for employment, examinations, and self-study purposes.

This study aims to explore the potential of LLM-based AES by comparing the effectiveness of five models: two LLMs (GPT Footnote 2 and BERT), one Japanese local LLM (OCLL), and two conventional machine learning-based methods (linguistic feature-based scoring tools - Jess and JWriter).

The research questions addressed in this study are as follows:

To what extent do the LLM-driven AES and linguistic feature-based AES, when used as automated tools to support human rating, accurately reflect test takers’ actual performance?

What influence does the prompt have on the accuracy and performance of LLM-based AES methods?

The subsequent sections of the manuscript cover the methodology, including the assessment measures for nonnative Japanese writing proficiency, criteria for prompts, and the dataset. The evaluation section focuses on the analysis of annotations and rating scores generated by LLM-driven and linguistic feature-based AES methods.

Methodology

The dataset utilized in this study was obtained from the International Corpus of Japanese as a Second Language (I-JAS) Footnote 3 . This corpus consisted of 1000 participants who represented 12 different first languages. For the study, the participants were given a story-writing task on a personal computer. They were required to write two stories based on the 4-panel illustrations titled “Picnic” and “The key” (see Appendix A). Background information for the participants was provided by the corpus, including their Japanese language proficiency levels assessed through two online tests: J-CAT and SPOT. These tests evaluated their reading, listening, vocabulary, and grammar abilities. The learners’ proficiency levels were categorized into six levels aligned with the Common European Framework of Reference for Languages (CEFR) and the Reference Framework for Japanese Language Education (RFJLE): A1, A2, B1, B2, C1, and C2. According to Lee et al. ( 2015 ), there is a high level of agreement (r = 0.86) between the J-CAT and SPOT assessments, indicating that the proficiency certifications provided by J-CAT are consistent with those of SPOT. However, it is important to note that the scores of J-CAT and SPOT do not have a one-to-one correspondence. In this study, the J-CAT scores were used as a benchmark to differentiate learners of different proficiency levels. A total of 1400 essays were utilized, representing the beginner (aligned with A1), A2, B1, B2, C1, and C2 levels based on the J-CAT scores. Table 1 provides information about the learners’ proficiency levels and their corresponding J-CAT and SPOT scores.

A dataset comprising a total of 1400 essays from the story writing tasks was collected. Among these, 714 essays were utilized to evaluate the reliability of the LLM-based AES method, while the remaining 686 essays were designated as development data to assess the LLM-based AES’s capability to distinguish participants with varying proficiency levels. The GPT 4 API was used in this study. A detailed explanation of the prompt-assessment criteria is provided in Section Prompt . All essays were sent to the model for measurement and scoring.

Measures of writing proficiency for nonnative Japanese

Japanese exhibits a morphologically agglutinative structure where morphemes are attached to the word stem to convey grammatical functions such as tense, aspect, voice, and honorifics, e.g. (5).

食べ-させ-られ-まし-た-か

tabe-sase-rare-mashi-ta-ka

[eat (stem)-causative-passive voice-honorification-tense. past-question marker]

Japanese employs nine case particles to indicate grammatical functions: the nominative case particle が (ga), the accusative case particle を (o), the genitive case particle の (no), the dative case particle に (ni), the locative/instrumental case particle で (de), the ablative case particle から (kara), the directional case particle へ (e), and the comitative case particle と (to). The agglutinative nature of the language, combined with the case particle system, provides an efficient means of distinguishing between active and passive voice, either through morphemes or case particles, e.g. 食べる taberu “eat concusive . ” (active voice); 食べられる taberareru “eat concusive . ” (passive voice). In the active voice, “パン を 食べる” (pan o taberu) translates to “to eat bread”. On the other hand, in the passive voice, it becomes “パン が 食べられた” (pan ga taberareta), which means “(the) bread was eaten”. Additionally, it is important to note that different conjugations of the same lemma are considered as one type in order to ensure a comprehensive assessment of the language features. For example, e.g., 食べる taberu “eat concusive . ”; 食べている tabeteiru “eat progress .”; 食べた tabeta “eat past . ” as one type.

To incorporate these features, previous research (Suzuki, 1999 ; Watanabe et al. 1988 ; Ishioka, 2001 ; Ishioka and Kameda, 2006 ; Hirao et al. 2020 ) has identified complexity, fluency, and accuracy as crucial factors for evaluating writing quality. These criteria are assessed through various aspects, including lexical richness (lexical density, diversity, and sophistication), syntactic complexity, and cohesion (Kyle et al. 2021 ; Mizumoto and Eguchi, 2023 ; Ure, 1971 ; Halliday, 1985 ; Barkaoui and Hadidi, 2020 ; Zenker and Kyle, 2021 ; Kim et al. 2018 ; Lu, 2017 ; Ortega, 2015 ). Therefore, this study proposes five scoring categories: lexical richness, syntactic complexity, cohesion, content elaboration, and grammatical accuracy. A total of 16 measures were employed to capture these categories. The calculation process and specific details of these measures can be found in Table 2 .

T-unit, first introduced by Hunt ( 1966 ), is a measure used for evaluating speech and composition. It serves as an indicator of syntactic development and represents the shortest units into which a piece of discourse can be divided without leaving any sentence fragments. In the context of Japanese language assessment, Sakoda and Hosoi ( 2020 ) utilized T-unit as the basic unit to assess the accuracy and complexity of Japanese learners’ speaking and storytelling. The calculation of T-units in Japanese follows the following principles:

A single main clause constitutes 1 T-unit, regardless of the presence or absence of dependent clauses, e.g. (6).

ケンとマリはピクニックに行きました (main clause): 1 T-unit.

If a sentence contains a main clause along with subclauses, each subclause is considered part of the same T-unit, e.g. (7).

天気が良かった の で (subclause)、ケンとマリはピクニックに行きました (main clause): 1 T-unit.

In the case of coordinate clauses, where multiple clauses are connected, each coordinated clause is counted separately. Thus, a sentence with coordinate clauses may have 2 T-units or more, e.g. (8).

ケンは地図で場所を探して (coordinate clause)、マリはサンドイッチを作りました (coordinate clause): 2 T-units.

Lexical diversity refers to the range of words used within a text (Engber, 1995 ; Kyle et al. 2021 ) and is considered a useful measure of the breadth of vocabulary in L n production (Jarvis, 2013a , 2013b ).

The type/token ratio (TTR) is widely recognized as a straightforward measure for calculating lexical diversity and has been employed in numerous studies. These studies have demonstrated a strong correlation between TTR and other methods of measuring lexical diversity (e.g., Bentz et al. 2016 ; Čech and Miroslav, 2018 ; Çöltekin and Taraka, 2018 ). TTR is computed by considering both the number of unique words (types) and the total number of words (tokens) in a given text. Given that the length of learners’ writing texts can vary, this study employs the moving average type-token ratio (MATTR) to mitigate the influence of text length. MATTR is calculated using a 50-word moving window. Initially, a TTR is determined for words 1–50 in an essay, followed by words 2–51, 3–52, and so on until the end of the essay is reached (Díez-Ortega and Kyle, 2023 ). The final MATTR scores were obtained by averaging the TTR scores for all 50-word windows. The following formula was employed to derive MATTR:

\({\rm{MATTR}}({\rm{W}})=\frac{{\sum }_{{\rm{i}}=1}^{{\rm{N}}-{\rm{W}}+1}{{\rm{F}}}_{{\rm{i}}}}{{\rm{W}}({\rm{N}}-{\rm{W}}+1)}\)

Here, N refers to the number of tokens in the corpus. W is the randomly selected token size (W < N). \({F}_{i}\) is the number of types in each window. The \({\rm{MATTR}}({\rm{W}})\) is the mean of a series of type-token ratios (TTRs) based on the word form for all windows. It is expected that individuals with higher language proficiency will produce texts with greater lexical diversity, as indicated by higher MATTR scores.

Lexical density was captured by the ratio of the number of lexical words to the total number of words (Lu, 2012 ). Lexical sophistication refers to the utilization of advanced vocabulary, often evaluated through word frequency indices (Crossley et al. 2013 ; Haberman, 2008 ; Kyle and Crossley, 2015 ; Laufer and Nation, 1995 ; Lu, 2012 ; Read, 2000 ). In line of writing, lexical sophistication can be interpreted as vocabulary breadth, which entails the appropriate usage of vocabulary items across various lexicon-grammatical contexts and registers (Garner et al. 2019 ; Kim et al. 2018 ; Kyle et al. 2018 ). In Japanese specifically, words are considered lexically sophisticated if they are not included in the “Japanese Education Vocabulary List Ver 1.0”. Footnote 4 Consequently, lexical sophistication was calculated by determining the number of sophisticated word types relative to the total number of words per essay. Furthermore, it has been suggested that, in Japanese writing, sentences should ideally have a length of no more than 40 to 50 characters, as this promotes readability. Therefore, the median and maximum sentence length can be considered as useful indices for assessment (Ishioka and Kameda, 2006 ).

Syntactic complexity was assessed based on several measures, including the mean length of clauses, verb phrases per T-unit, clauses per T-unit, dependent clauses per T-unit, complex nominals per clause, adverbial clauses per clause, coordinate phrases per clause, and mean dependency distance (MDD). The MDD reflects the distance between the governor and dependent positions in a sentence. A larger dependency distance indicates a higher cognitive load and greater complexity in syntactic processing (Liu, 2008 ; Liu et al. 2017 ). The MDD has been established as an efficient metric for measuring syntactic complexity (Jiang, Quyang, and Liu, 2019 ; Li and Yan, 2021 ). To calculate the MDD, the position numbers of the governor and dependent are subtracted, assuming that words in a sentence are assigned in a linear order, such as W1 … Wi … Wn. In any dependency relationship between words Wa and Wb, Wa is the governor and Wb is the dependent. The MDD of the entire sentence was obtained by taking the absolute value of governor – dependent:

MDD = \(\frac{1}{n}{\sum }_{i=1}^{n}|{\rm{D}}{{\rm{D}}}_{i}|\)

In this formula, \(n\) represents the number of words in the sentence, and \({DD}i\) is the dependency distance of the \({i}^{{th}}\) dependency relationship of a sentence. Building on this, the annotation of sentence ‘Mary-ga-John-ni-keshigomu-o-watashita was [Mary- top -John- dat -eraser- acc -give- past] ’. The sentence’s MDD would be 2. Table 3 provides the CSV file as a prompt for GPT 4.

Cohesion (semantic similarity) and content elaboration aim to capture the ideas presented in test taker’s essays. Cohesion was assessed using three measures: Synonym overlap/paragraph (topic), Synonym overlap/paragraph (keywords), and word2vec cosine similarity. Content elaboration and development were measured as the number of metadiscourse markers (type)/number of words. To capture content closely, this study proposed a novel-distance based representation, by encoding the cosine distance between the essay (by learner) and essay task’s (topic and keyword) i -vectors. The learner’s essay is decoded into a word sequence, and aligned to the essay task’ topic and keyword for log-likelihood measurement. The cosine distance reveals the content elaboration score in the leaners’ essay. The mathematical equation of cosine similarity between target-reference vectors is shown in (11), assuming there are i essays and ( L i , …. L n ) and ( N i , …. N n ) are the vectors representing the learner and task’s topic and keyword respectively. The content elaboration distance between L i and N i was calculated as follows:

\(\cos \left(\theta \right)=\frac{{\rm{L}}\,\cdot\, {\rm{N}}}{\left|{\rm{L}}\right|{\rm{|N|}}}=\frac{\mathop{\sum }\nolimits_{i=1}^{n}{L}_{i}{N}_{i}}{\sqrt{\mathop{\sum }\nolimits_{i=1}^{n}{L}_{i}^{2}}\sqrt{\mathop{\sum }\nolimits_{i=1}^{n}{N}_{i}^{2}}}\)

A high similarity value indicates a low difference between the two recognition outcomes, which in turn suggests a high level of proficiency in content elaboration.

To evaluate the effectiveness of the proposed measures in distinguishing different proficiency levels among nonnative Japanese speakers’ writing, we conducted a multi-faceted Rasch measurement analysis (Linacre, 1994 ). This approach applies measurement models to thoroughly analyze various factors that can influence test outcomes, including test takers’ proficiency, item difficulty, and rater severity, among others. The underlying principles and functionality of multi-faceted Rasch measurement are illustrated in (12).

\(\log \left(\frac{{P}_{{nijk}}}{{P}_{{nij}(k-1)}}\right)={B}_{n}-{D}_{i}-{C}_{j}-{F}_{k}\)

(12) defines the logarithmic transformation of the probability ratio ( P nijk /P nij(k-1) )) as a function of multiple parameters. Here, n represents the test taker, i denotes a writing proficiency measure, j corresponds to the human rater, and k represents the proficiency score. The parameter B n signifies the proficiency level of test taker n (where n ranges from 1 to N). D j represents the difficulty parameter of test item i (where i ranges from 1 to L), while C j represents the severity of rater j (where j ranges from 1 to J). Additionally, F k represents the step difficulty for a test taker to move from score ‘k-1’ to k . P nijk refers to the probability of rater j assigning score k to test taker n for test item i . P nij(k-1) represents the likelihood of test taker n being assigned score ‘k-1’ by rater j for test item i . Each facet within the test is treated as an independent parameter and estimated within the same reference framework. To evaluate the consistency of scores obtained through both human and computer analysis, we utilized the Infit mean-square statistic. This statistic is a chi-square measure divided by the degrees of freedom and is weighted with information. It demonstrates higher sensitivity to unexpected patterns in responses to items near a person’s proficiency level (Linacre, 2002 ). Fit statistics are assessed based on predefined thresholds for acceptable fit. For the Infit MNSQ, which has a mean of 1.00, different thresholds have been suggested. Some propose stricter thresholds ranging from 0.7 to 1.3 (Bond et al. 2021 ), while others suggest more lenient thresholds ranging from 0.5 to 1.5 (Eckes, 2009 ). In this study, we adopted the criterion of 0.70–1.30 for the Infit MNSQ.

Moving forward, we can now proceed to assess the effectiveness of the 16 proposed measures based on five criteria for accurately distinguishing various levels of writing proficiency among non-native Japanese speakers. To conduct this evaluation, we utilized the development dataset from the I-JAS corpus, as described in Section Dataset . Table 4 provides a measurement report that presents the performance details of the 14 metrics under consideration. The measure separation was found to be 4.02, indicating a clear differentiation among the measures. The reliability index for the measure separation was 0.891, suggesting consistency in the measurement. Similarly, the person separation reliability index was 0.802, indicating the accuracy of the assessment in distinguishing between individuals. All 16 measures demonstrated Infit mean squares within a reasonable range, ranging from 0.76 to 1.28. The Synonym overlap/paragraph (topic) measure exhibited a relatively high outfit mean square of 1.46, although the Infit mean square falls within an acceptable range. The standard error for the measures ranged from 0.13 to 0.28, indicating the precision of the estimates.

Table 5 further illustrated the weights assigned to different linguistic measures for score prediction, with higher weights indicating stronger correlations between those measures and higher scores. Specifically, the following measures exhibited higher weights compared to others: moving average type token ratio per essay has a weight of 0.0391. Mean dependency distance had a weight of 0.0388. Mean length of clause, calculated by dividing the number of words by the number of clauses, had a weight of 0.0374. Complex nominals per T-unit, calculated by dividing the number of complex nominals by the number of T-units, had a weight of 0.0379. Coordinate phrases rate, calculated by dividing the number of coordinate phrases by the number of clauses, had a weight of 0.0325. Grammatical error rate, representing the number of errors per essay, had a weight of 0.0322.

Criteria (output indicator)

The criteria used to evaluate the writing ability in this study were based on CEFR, which follows a six-point scale ranging from A1 to C2. To assess the quality of Japanese writing, the scoring criteria from Table 6 were utilized. These criteria were derived from the IELTS writing standards and served as assessment guidelines and prompts for the written output.

A prompt is a question or detailed instruction that is provided to the model to obtain a proper response. After several pilot experiments, we decided to provide the measures (Section Measures of writing proficiency for nonnative Japanese ) as the input prompt and use the criteria (Section Criteria (output indicator) ) as the output indicator. Regarding the prompt language, considering that the LLM was tasked with rating Japanese essays, would prompt in Japanese works better Footnote 5 ? We conducted experiments comparing the performance of GPT-4 using both English and Japanese prompts. Additionally, we utilized the Japanese local model OCLL with Japanese prompts. Multiple trials were conducted using the same sample. Regardless of the prompt language used, we consistently obtained the same grading results with GPT-4, which assigned a grade of B1 to the writing sample. This suggested that GPT-4 is reliable and capable of producing consistent ratings regardless of the prompt language. On the other hand, when we used Japanese prompts with the Japanese local model “OCLL”, we encountered inconsistent grading results. Out of 10 attempts with OCLL, only 6 yielded consistent grading results (B1), while the remaining 4 showed different outcomes, including A1 and B2 grades. These findings indicated that the language of the prompt was not the determining factor for reliable AES. Instead, the size of the training data and the model parameters played crucial roles in achieving consistent and reliable AES results for the language model.

The following is the utilized prompt, which details all measures and requires the LLM to score the essays using holistic and trait scores.

Please evaluate Japanese essays written by Japanese learners and assign a score to each essay on a six-point scale, ranging from A1, A2, B1, B2, C1 to C2. Additionally, please provide trait scores and display the calculation process for each trait score. The scoring should be based on the following criteria:

Moving average type-token ratio.

Number of lexical words (token) divided by the total number of words per essay.

Number of sophisticated word types divided by the total number of words per essay.

Mean length of clause.

Verb phrases per T-unit.

Clauses per T-unit.

Dependent clauses per T-unit.

Complex nominals per clause.

Adverbial clauses per clause.

Coordinate phrases per clause.

Mean dependency distance.

Synonym overlap paragraph (topic and keywords).

Word2vec cosine similarity.

Connectives per essay.

Conjunctions per essay.

Number of metadiscourse markers (types) divided by the total number of words.

Number of errors per essay.

Japanese essay text

出かける前に二人が地図を見ている間に、サンドイッチを入れたバスケットに犬が入ってしまいました。それに気づかずに二人は楽しそうに出かけて行きました。やがて突然犬がバスケットから飛び出し、二人は驚きました。バスケット の 中を見ると、食べ物はすべて犬に食べられていて、二人は困ってしまいました。(ID_JJJ01_SW1)

The score of the example above was B1. Figure 3 provides an example of holistic and trait scores provided by GPT-4 (with a prompt indicating all measures) via Bing Footnote 6 .

Example of GPT-4 AES and feedback (with a prompt indicating all measures).

Statistical analysis

The aim of this study is to investigate the potential use of LLM for nonnative Japanese AES. It seeks to compare the scoring outcomes obtained from feature-based AES tools, which rely on conventional machine learning technology (i.e. Jess, JWriter), with those generated by AI-driven AES tools utilizing deep learning technology (BERT, GPT, OCLL). To assess the reliability of a computer-assisted annotation tool, the study initially established human-human agreement as the benchmark measure. Subsequently, the performance of the LLM-based method was evaluated by comparing it to human-human agreement.

To assess annotation agreement, the study employed standard measures such as precision, recall, and F-score (Brants 2000 ; Lu 2010 ), along with the quadratically weighted kappa (QWK) to evaluate the consistency and agreement in the annotation process. Assume A and B represent human annotators. When comparing the annotations of the two annotators, the following results are obtained. The evaluation of precision, recall, and F-score metrics was illustrated in equations (13) to (15).

\({\rm{Recall}}(A,B)=\frac{{\rm{Number}}\,{\rm{of}}\,{\rm{identical}}\,{\rm{nodes}}\,{\rm{in}}\,A\,{\rm{and}}\,B}{{\rm{Number}}\,{\rm{of}}\,{\rm{nodes}}\,{\rm{in}}\,A}\)

\({\rm{Precision}}(A,\,B)=\frac{{\rm{Number}}\,{\rm{of}}\,{\rm{identical}}\,{\rm{nodes}}\,{\rm{in}}\,A\,{\rm{and}}\,B}{{\rm{Number}}\,{\rm{of}}\,{\rm{nodes}}\,{\rm{in}}\,B}\)

The F-score is the harmonic mean of recall and precision:

\({\rm{F}}-{\rm{score}}=\frac{2* ({\rm{Precision}}* {\rm{Recall}})}{{\rm{Precision}}+{\rm{Recall}}}\)

The highest possible value of an F-score is 1.0, indicating perfect precision and recall, and the lowest possible value is 0, if either precision or recall are zero.

In accordance with Taghipour and Ng ( 2016 ), the calculation of QWK involves two steps:

Step 1: Construct a weight matrix W as follows:

\({W}_{{ij}}=\frac{{(i-j)}^{2}}{{(N-1)}^{2}}\)

i represents the annotation made by the tool, while j represents the annotation made by a human rater. N denotes the total number of possible annotations. Matrix O is subsequently computed, where O_( i, j ) represents the count of data annotated by the tool ( i ) and the human annotator ( j ). On the other hand, E refers to the expected count matrix, which undergoes normalization to ensure that the sum of elements in E matches the sum of elements in O.

Step 2: With matrices O and E, the QWK is obtained as follows:

K = 1- \(\frac{\sum i,j{W}_{i,j}\,{O}_{i,j}}{\sum i,j{W}_{i,j}\,{E}_{i,j}}\)

The value of the quadratic weighted kappa increases as the level of agreement improves. Further, to assess the accuracy of LLM scoring, the proportional reductive mean square error (PRMSE) was employed. The PRMSE approach takes into account the variability observed in human ratings to estimate the rater error, which is then subtracted from the variance of the human labels. This calculation provides an overall measure of agreement between the automated scores and true scores (Haberman et al. 2015 ; Loukina et al. 2020 ; Taghipour and Ng, 2016 ). The computation of PRMSE involves the following steps:

Step 1: Calculate the mean squared errors (MSEs) for the scoring outcomes of the computer-assisted tool (MSE tool) and the human scoring outcomes (MSE human).

Step 2: Determine the PRMSE by comparing the MSE of the computer-assisted tool (MSE tool) with the MSE from human raters (MSE human), using the following formula:

\({\rm{PRMSE}}=1-\frac{({\rm{MSE}}\,{\rm{tool}})\,}{({\rm{MSE}}\,{\rm{human}})\,}=1-\,\frac{{\sum }_{i}^{n}=1{({{\rm{y}}}_{i}-{\hat{{\rm{y}}}}_{{\rm{i}}})}^{2}}{{\sum }_{i}^{n}=1{({{\rm{y}}}_{i}-\hat{{\rm{y}}})}^{2}}\)

In the numerator, ŷi represents the scoring outcome predicted by a specific LLM-driven AES system for a given sample. The term y i − ŷ i represents the difference between this predicted outcome and the mean value of all LLM-driven AES systems’ scoring outcomes. It quantifies the deviation of the specific LLM-driven AES system’s prediction from the average prediction of all LLM-driven AES systems. In the denominator, y i − ŷ represents the difference between the scoring outcome provided by a specific human rater for a given sample and the mean value of all human raters’ scoring outcomes. It measures the discrepancy between the specific human rater’s score and the average score given by all human raters. The PRMSE is then calculated by subtracting the ratio of the MSE tool to the MSE human from 1. PRMSE falls within the range of 0 to 1, with larger values indicating reduced errors in LLM’s scoring compared to those of human raters. In other words, a higher PRMSE implies that LLM’s scoring demonstrates greater accuracy in predicting the true scores (Loukina et al. 2020 ). The interpretation of kappa values, ranging from 0 to 1, is based on the work of Landis and Koch ( 1977 ). Specifically, the following categories are assigned to different ranges of kappa values: −1 indicates complete inconsistency, 0 indicates random agreement, 0.0 ~ 0.20 indicates extremely low level of agreement (slight), 0.21 ~ 0.40 indicates moderate level of agreement (fair), 0.41 ~ 0.60 indicates medium level of agreement (moderate), 0.61 ~ 0.80 indicates high level of agreement (substantial), 0.81 ~ 1 indicates almost perfect level of agreement. All statistical analyses were executed using Python script.

Results and discussion

Annotation reliability of the llm.

This section focuses on assessing the reliability of the LLM’s annotation and scoring capabilities. To evaluate the reliability, several tests were conducted simultaneously, aiming to achieve the following objectives:

Assess the LLM’s ability to differentiate between test takers with varying levels of oral proficiency.

Determine the level of agreement between the annotations and scoring performed by the LLM and those done by human raters.

The evaluation of the results encompassed several metrics, including: precision, recall, F-Score, quadratically-weighted kappa, proportional reduction of mean squared error, Pearson correlation, and multi-faceted Rasch measurement.

Inter-annotator agreement (human–human annotator agreement)

We started with an agreement test of the two human annotators. Two trained annotators were recruited to determine the writing task data measures. A total of 714 scripts, as the test data, was utilized. Each analysis lasted 300–360 min. Inter-annotator agreement was evaluated using the standard measures of precision, recall, and F-score and QWK. Table 7 presents the inter-annotator agreement for the various indicators. As shown, the inter-annotator agreement was fairly high, with F-scores ranging from 1.0 for sentence and word number to 0.666 for grammatical errors.

The findings from the QWK analysis provided further confirmation of the inter-annotator agreement. The QWK values covered a range from 0.950 ( p  = 0.000) for sentence and word number to 0.695 for synonym overlap number (keyword) and grammatical errors ( p  = 0.001).

Agreement of annotation outcomes between human and LLM

To evaluate the consistency between human annotators and LLM annotators (BERT, GPT, OCLL) across the indices, the same test was conducted. The results of the inter-annotator agreement (F-score) between LLM and human annotation are provided in Appendix B-D. The F-scores ranged from 0.706 for Grammatical error # for OCLL-human to a perfect 1.000 for GPT-human, for sentences, clauses, T-units, and words. These findings were further supported by the QWK analysis, which showed agreement levels ranging from 0.807 ( p  = 0.001) for metadiscourse markers for OCLL-human to 0.962 for words ( p  = 0.000) for GPT-human. The findings demonstrated that the LLM annotation achieved a significant level of accuracy in identifying measurement units and counts.

Reliability of LLM-driven AES’s scoring and discriminating proficiency levels

This section examines the reliability of the LLM-driven AES scoring through a comparison of the scoring outcomes produced by human raters and the LLM ( Reliability of LLM-driven AES scoring ). It also assesses the effectiveness of the LLM-based AES system in differentiating participants with varying proficiency levels ( Reliability of LLM-driven AES discriminating proficiency levels ).

Reliability of LLM-driven AES scoring

Table 8 summarizes the QWK coefficient analysis between the scores computed by the human raters and the GPT-4 for the individual essays from I-JAS Footnote 7 . As shown, the QWK of all measures ranged from k  = 0.819 for lexical density (number of lexical words (tokens)/number of words per essay) to k  = 0.644 for word2vec cosine similarity. Table 9 further presents the Pearson correlations between the 16 writing proficiency measures scored by human raters and GPT 4 for the individual essays. The correlations ranged from 0.672 for syntactic complexity to 0.734 for grammatical accuracy. The correlations between the writing proficiency scores assigned by human raters and the BERT-based AES system were found to range from 0.661 for syntactic complexity to 0.713 for grammatical accuracy. The correlations between the writing proficiency scores given by human raters and the OCLL-based AES system ranged from 0.654 for cohesion to 0.721 for grammatical accuracy. These findings indicated an alignment between the assessments made by human raters and both the BERT-based and OCLL-based AES systems in terms of various aspects of writing proficiency.

Reliability of LLM-driven AES discriminating proficiency levels

After validating the reliability of the LLM’s annotation and scoring, the subsequent objective was to evaluate its ability to distinguish between various proficiency levels. For this analysis, a dataset of 686 individual essays was utilized. Table 10 presents a sample of the results, summarizing the means, standard deviations, and the outcomes of the one-way ANOVAs based on the measures assessed by the GPT-4 model. A post hoc multiple comparison test, specifically the Bonferroni test, was conducted to identify any potential differences between pairs of levels.

As the results reveal, seven measures presented linear upward or downward progress across the three proficiency levels. These were marked in bold in Table 10 and comprise one measure of lexical richness, i.e. MATTR (lexical diversity); four measures of syntactic complexity, i.e. MDD (mean dependency distance), MLC (mean length of clause), CNT (complex nominals per T-unit), CPC (coordinate phrases rate); one cohesion measure, i.e. word2vec cosine similarity and GER (grammatical error rate). Regarding the ability of the sixteen measures to distinguish adjacent proficiency levels, the Bonferroni tests indicated that statistically significant differences exist between the primary level and the intermediate level for MLC and GER. One measure of lexical richness, namely LD, along with three measures of syntactic complexity (VPT, CT, DCT, ACC), two measures of cohesion (SOPT, SOPK), and one measure of content elaboration (IMM), exhibited statistically significant differences between proficiency levels. However, these differences did not demonstrate a linear progression between adjacent proficiency levels. No significant difference was observed in lexical sophistication between proficiency levels.

To summarize, our study aimed to evaluate the reliability and differentiation capabilities of the LLM-driven AES method. For the first objective, we assessed the LLM’s ability to differentiate between test takers with varying levels of oral proficiency using precision, recall, F-Score, and quadratically-weighted kappa. Regarding the second objective, we compared the scoring outcomes generated by human raters and the LLM to determine the level of agreement. We employed quadratically-weighted kappa and Pearson correlations to compare the 16 writing proficiency measures for the individual essays. The results confirmed the feasibility of using the LLM for annotation and scoring in AES for nonnative Japanese. As a result, Research Question 1 has been addressed.

Comparison of BERT-, GPT-, OCLL-based AES, and linguistic-feature-based computation methods

This section aims to compare the effectiveness of five AES methods for nonnative Japanese writing, i.e. LLM-driven approaches utilizing BERT, GPT, and OCLL, linguistic feature-based approaches using Jess and JWriter. The comparison was conducted by comparing the ratings obtained from each approach with human ratings. All ratings were derived from the dataset introduced in Dataset . To facilitate the comparison, the agreement between the automated methods and human ratings was assessed using QWK and PRMSE. The performance of each approach was summarized in Table 11 .

The QWK coefficient values indicate that LLMs (GPT, BERT, OCLL) and human rating outcomes demonstrated higher agreement compared to feature-based AES methods (Jess and JWriter) in assessing writing proficiency criteria, including lexical richness, syntactic complexity, content, and grammatical accuracy. Among the LLMs, the GPT-4 driven AES and human rating outcomes showed the highest agreement in all criteria, except for syntactic complexity. The PRMSE values suggest that the GPT-based method outperformed linguistic feature-based methods and other LLM-based approaches. Moreover, an interesting finding emerged during the study: the agreement coefficient between GPT-4 and human scoring was even higher than the agreement between different human raters themselves. This discovery highlights the advantage of GPT-based AES over human rating. Ratings involve a series of processes, including reading the learners’ writing, evaluating the content and language, and assigning scores. Within this chain of processes, various biases can be introduced, stemming from factors such as rater biases, test design, and rating scales. These biases can impact the consistency and objectivity of human ratings. GPT-based AES may benefit from its ability to apply consistent and objective evaluation criteria. By prompting the GPT model with detailed writing scoring rubrics and linguistic features, potential biases in human ratings can be mitigated. The model follows a predefined set of guidelines and does not possess the same subjective biases that human raters may exhibit. This standardization in the evaluation process contributes to the higher agreement observed between GPT-4 and human scoring. Section Prompt strategy of the study delves further into the role of prompts in the application of LLMs to AES. It explores how the choice and implementation of prompts can impact the performance and reliability of LLM-based AES methods. Furthermore, it is important to acknowledge the strengths of the local model, i.e. the Japanese local model OCLL, which excels in processing certain idiomatic expressions. Nevertheless, our analysis indicated that GPT-4 surpasses local models in AES. This superior performance can be attributed to the larger parameter size of GPT-4, estimated to be between 500 billion and 1 trillion, which exceeds the sizes of both BERT and the local model OCLL.

Prompt strategy

In the context of prompt strategy, Mizumoto and Eguchi ( 2023 ) conducted a study where they applied the GPT-3 model to automatically score English essays in the TOEFL test. They found that the accuracy of the GPT model alone was moderate to fair. However, when they incorporated linguistic measures such as cohesion, syntactic complexity, and lexical features alongside the GPT model, the accuracy significantly improved. This highlights the importance of prompt engineering and providing the model with specific instructions to enhance its performance. In this study, a similar approach was taken to optimize the performance of LLMs. GPT-4, which outperformed BERT and OCLL, was selected as the candidate model. Model 1 was used as the baseline, representing GPT-4 without any additional prompting. Model 2, on the other hand, involved GPT-4 prompted with 16 measures that included scoring criteria, efficient linguistic features for writing assessment, and detailed measurement units and calculation formulas. The remaining models (Models 3 to 18) utilized GPT-4 prompted with individual measures. The performance of these 18 different models was assessed using the output indicators described in Section Criteria (output indicator) . By comparing the performances of these models, the study aimed to understand the impact of prompt engineering on the accuracy and effectiveness of GPT-4 in AES tasks.

Based on the PRMSE scores presented in Fig. 4 , it was observed that Model 1, representing GPT-4 without any additional prompting, achieved a fair level of performance. However, Model 2, which utilized GPT-4 prompted with all measures, outperformed all other models in terms of PRMSE score, achieving a score of 0.681. These results indicate that the inclusion of specific measures and prompts significantly enhanced the performance of GPT-4 in AES. Among the measures, syntactic complexity was found to play a particularly significant role in improving the accuracy of GPT-4 in assessing writing quality. Following that, lexical diversity emerged as another important factor contributing to the model’s effectiveness. The study suggests that a well-prompted GPT-4 can serve as a valuable tool to support human assessors in evaluating writing quality. By utilizing GPT-4 as an automated scoring tool, the evaluation biases associated with human raters can be minimized. This has the potential to empower teachers by allowing them to focus on designing writing tasks and guiding writing strategies, while leveraging the capabilities of GPT-4 for efficient and reliable scoring.

PRMSE scores of the 18 AES models.

This study aimed to investigate two main research questions: the feasibility of utilizing LLMs for AES and the impact of prompt engineering on the application of LLMs in AES.

To address the first objective, the study compared the effectiveness of five different models: GPT, BERT, the Japanese local LLM (OCLL), and two conventional machine learning-based AES tools (Jess and JWriter). The PRMSE values indicated that the GPT-4-based method outperformed other LLMs (BERT, OCLL) and linguistic feature-based computational methods (Jess and JWriter) across various writing proficiency criteria. Furthermore, the agreement coefficient between GPT-4 and human scoring surpassed the agreement among human raters themselves, highlighting the potential of using the GPT-4 tool to enhance AES by reducing biases and subjectivity, saving time, labor, and cost, and providing valuable feedback for self-study. Regarding the second goal, the role of prompt design was investigated by comparing 18 models, including a baseline model, a model prompted with all measures, and 16 models prompted with one measure at a time. GPT-4, which outperformed BERT and OCLL, was selected as the candidate model. The PRMSE scores of the models showed that GPT-4 prompted with all measures achieved the best performance, surpassing the baseline and other models.

In conclusion, this study has demonstrated the potential of LLMs in supporting human rating in assessments. By incorporating automation, we can save time and resources while reducing biases and subjectivity inherent in human rating processes. Automated language assessments offer the advantage of accessibility, providing equal opportunities and economic feasibility for individuals who lack access to traditional assessment centers or necessary resources. LLM-based language assessments provide valuable feedback and support to learners, aiding in the enhancement of their language proficiency and the achievement of their goals. This personalized feedback can cater to individual learner needs, facilitating a more tailored and effective language-learning experience.

There are three important areas that merit further exploration. First, prompt engineering requires attention to ensure optimal performance of LLM-based AES across different language types. This study revealed that GPT-4, when prompted with all measures, outperformed models prompted with fewer measures. Therefore, investigating and refining prompt strategies can enhance the effectiveness of LLMs in automated language assessments. Second, it is crucial to explore the application of LLMs in second-language assessment and learning for oral proficiency, as well as their potential in under-resourced languages. Recent advancements in self-supervised machine learning techniques have significantly improved automatic speech recognition (ASR) systems, opening up new possibilities for creating reliable ASR systems, particularly for under-resourced languages with limited data. However, challenges persist in the field of ASR. First, ASR assumes correct word pronunciation for automatic pronunciation evaluation, which proves challenging for learners in the early stages of language acquisition due to diverse accents influenced by their native languages. Accurately segmenting short words becomes problematic in such cases. Second, developing precise audio-text transcriptions for languages with non-native accented speech poses a formidable task. Last, assessing oral proficiency levels involves capturing various linguistic features, including fluency, pronunciation, accuracy, and complexity, which are not easily captured by current NLP technology.

Data availability

The dataset utilized was obtained from the International Corpus of Japanese as a Second Language (I-JAS). The data URLs: [ https://www2.ninjal.ac.jp/jll/lsaj/ihome2.html ].

J-CAT and TTBJ are two computerized adaptive tests used to assess Japanese language proficiency.

SPOT is a specific component of the TTBJ test.

J-CAT: https://www.j-cat2.org/html/ja/pages/interpret.html

SPOT: https://ttbj.cegloc.tsukuba.ac.jp/p1.html#SPOT .

The study utilized a prompt-based GPT-4 model, developed by OpenAI, which has an impressive architecture with 1.8 trillion parameters across 120 layers. GPT-4 was trained on a vast dataset of 13 trillion tokens, using two stages: initial training on internet text datasets to predict the next token, and subsequent fine-tuning through reinforcement learning from human feedback.

https://www2.ninjal.ac.jp/jll/lsaj/ihome2-en.html .

http://jhlee.sakura.ne.jp/JEV/ by Japanese Learning Dictionary Support Group 2015.

We express our sincere gratitude to the reviewer for bringing this matter to our attention.

On February 7, 2023, Microsoft began rolling out a major overhaul to Bing that included a new chatbot feature based on OpenAI’s GPT-4 (Bing.com).

Appendix E-F present the analysis results of the QWK coefficient between the scores computed by the human raters and the BERT, OCLL models.

Attali Y, Burstein J (2006) Automated essay scoring with e-rater® V.2. J. Technol., Learn. Assess., 4

Barkaoui K, Hadidi A (2020) Assessing Change in English Second Language Writing Performance (1st ed.). Routledge, New York. https://doi.org/10.4324/9781003092346

Bentz C, Tatyana R, Koplenig A, Tanja S (2016) A comparison between morphological complexity. measures: Typological data vs. language corpora. In Proceedings of the workshop on computational linguistics for linguistic complexity (CL4LC), 142–153. Osaka, Japan: The COLING 2016 Organizing Committee

Bond TG, Yan Z, Heene M (2021) Applying the Rasch model: Fundamental measurement in the human sciences (4th ed). Routledge

Brants T (2000) Inter-annotator agreement for a German newspaper corpus. Proceedings of the Second International Conference on Language Resources and Evaluation (LREC’00), Athens, Greece, 31 May-2 June, European Language Resources Association

Brown TB, Mann B, Ryder N, et al. (2020) Language models are few-shot learners. Advances in Neural Information Processing Systems, Online, 6–12 December, Curran Associates, Inc., Red Hook, NY

Burstein J (2003) The E-rater scoring engine: Automated essay scoring with natural language processing. In Shermis MD and Burstein JC (ed) Automated Essay Scoring: A Cross-Disciplinary Perspective. Lawrence Erlbaum Associates, Mahwah, NJ

Čech R, Miroslav K (2018) Morphological richness of text. In Masako F, Václav C (ed) Taming the corpus: From inflection and lexis to interpretation, 63–77. Cham, Switzerland: Springer Nature

Çöltekin Ç, Taraka, R (2018) Exploiting Universal Dependencies treebanks for measuring morphosyntactic complexity. In Aleksandrs B, Christian B (ed), Proceedings of first workshop on measuring language complexity, 1–7. Torun, Poland

Crossley SA, Cobb T, McNamara DS (2013) Comparing count-based and band-based indices of word frequency: Implications for active vocabulary research and pedagogical applications. System 41:965–981. https://doi.org/10.1016/j.system.2013.08.002

Article   Google Scholar  

Crossley SA, McNamara DS (2016) Say more and be more coherent: How text elaboration and cohesion can increase writing quality. J. Writ. Res. 7:351–370

CyberAgent Inc (2023) Open-Calm series of Japanese language models. Retrieved from: https://www.cyberagent.co.jp/news/detail/id=28817

Devlin J, Chang MW, Lee K, Toutanova K (2019) BERT: Pre-training of deep bidirectional transformers for language understanding. Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics, Minneapolis, Minnesota, 2–7 June, pp. 4171–4186. Association for Computational Linguistics

Diez-Ortega M, Kyle K (2023) Measuring the development of lexical richness of L2 Spanish: a longitudinal learner corpus study. Studies in Second Language Acquisition 1-31

Eckes T (2009) On common ground? How raters perceive scoring criteria in oral proficiency testing. In Brown A, Hill K (ed) Language testing and evaluation 13: Tasks and criteria in performance assessment (pp. 43–73). Peter Lang Publishing

Elliot S (2003) IntelliMetric: from here to validity. In: Shermis MD, Burstein JC (ed) Automated Essay Scoring: A Cross-Disciplinary Perspective. Lawrence Erlbaum Associates, Mahwah, NJ

Google Scholar  

Engber CA (1995) The relationship of lexical proficiency to the quality of ESL compositions. J. Second Lang. Writ. 4:139–155

Garner J, Crossley SA, Kyle K (2019) N-gram measures and L2 writing proficiency. System 80:176–187. https://doi.org/10.1016/j.system.2018.12.001

Haberman SJ (2008) When can subscores have value? J. Educat. Behav. Stat., 33:204–229

Haberman SJ, Yao L, Sinharay S (2015) Prediction of true test scores from observed item scores and ancillary data. Brit. J. Math. Stat. Psychol. 68:363–385

Halliday MAK (1985) Spoken and Written Language. Deakin University Press, Melbourne, Australia

Hirao R, Arai M, Shimanaka H et al. (2020) Automated essay scoring system for nonnative Japanese learners. Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), pp. 1250–1257. European Language Resources Association

Hunt KW (1966) Recent Measures in Syntactic Development. Elementary English, 43(7), 732–739. http://www.jstor.org/stable/41386067

Ishioka T (2001) About e-rater, a computer-based automatic scoring system for essays [Konpyūta ni yoru essei no jidō saiten shisutemu e − rater ni tsuite]. University Entrance Examination. Forum [Daigaku nyūshi fōramu] 24:71–76

Hochreiter S, Schmidhuber J (1997) Long short- term memory. Neural Comput. 9(8):1735–1780

Article   CAS   PubMed   Google Scholar  

Ishioka T, Kameda M (2006) Automated Japanese essay scoring system based on articles written by experts. Proceedings of the 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics, Sydney, Australia, 17–18 July 2006, pp. 233-240. Association for Computational Linguistics, USA

Japan Foundation (2021) Retrieved from: https://www.jpf.gp.jp/j/project/japanese/survey/result/dl/survey2021/all.pdf

Jarvis S (2013a) Defining and measuring lexical diversity. In Jarvis S, Daller M (ed) Vocabulary knowledge: Human ratings and automated measures (Vol. 47, pp. 13–44). John Benjamins. https://doi.org/10.1075/sibil.47.03ch1

Jarvis S (2013b) Capturing the diversity in lexical diversity. Lang. Learn. 63:87–106. https://doi.org/10.1111/j.1467-9922.2012.00739.x

Jiang J, Quyang J, Liu H (2019) Interlanguage: A perspective of quantitative linguistic typology. Lang. Sci. 74:85–97

Kim M, Crossley SA, Kyle K (2018) Lexical sophistication as a multidimensional phenomenon: Relations to second language lexical proficiency, development, and writing quality. Mod. Lang. J. 102(1):120–141. https://doi.org/10.1111/modl.12447

Kojima T, Gu S, Reid M et al. (2022) Large language models are zero-shot reasoners. Advances in Neural Information Processing Systems, New Orleans, LA, 29 November-1 December, Curran Associates, Inc., Red Hook, NY

Kyle K, Crossley SA (2015) Automatically assessing lexical sophistication: Indices, tools, findings, and application. TESOL Q 49:757–786

Kyle K, Crossley SA, Berger CM (2018) The tool for the automatic analysis of lexical sophistication (TAALES): Version 2.0. Behav. Res. Methods 50:1030–1046. https://doi.org/10.3758/s13428-017-0924-4

Article   PubMed   Google Scholar  

Kyle K, Crossley SA, Jarvis S (2021) Assessing the validity of lexical diversity using direct judgements. Lang. Assess. Q. 18:154–170. https://doi.org/10.1080/15434303.2020.1844205

Landauer TK, Laham D, Foltz PW (2003) Automated essay scoring and annotation of essays with the Intelligent Essay Assessor. In Shermis MD, Burstein JC (ed), Automated Essay Scoring: A Cross-Disciplinary Perspective. Lawrence Erlbaum Associates, Mahwah, NJ

Landis JR, Koch GG (1977) The measurement of observer agreement for categorical data. Biometrics 159–174

Laufer B, Nation P (1995) Vocabulary size and use: Lexical richness in L2 written production. Appl. Linguist. 16:307–322. https://doi.org/10.1093/applin/16.3.307

Lee J, Hasebe Y (2017) jWriter Learner Text Evaluator, URL: https://jreadability.net/jwriter/

Lee J, Kobayashi N, Sakai T, Sakota K (2015) A Comparison of SPOT and J-CAT Based on Test Analysis [Tesuto bunseki ni motozuku ‘SPOT’ to ‘J-CAT’ no hikaku]. Research on the Acquisition of Second Language Japanese [Dainigengo to shite no nihongo no shūtoku kenkyū] (18) 53–69

Li W, Yan J (2021) Probability distribution of dependency distance based on a Treebank of. Japanese EFL Learners’ Interlanguage. J. Quant. Linguist. 28(2):172–186. https://doi.org/10.1080/09296174.2020.1754611

Article   MathSciNet   Google Scholar  

Linacre JM (2002) Optimizing rating scale category effectiveness. J. Appl. Meas. 3(1):85–106

PubMed   Google Scholar  

Linacre JM (1994) Constructing measurement with a Many-Facet Rasch Model. In Wilson M (ed) Objective measurement: Theory into practice, Volume 2 (pp. 129–144). Norwood, NJ: Ablex

Liu H (2008) Dependency distance as a metric of language comprehension difficulty. J. Cognitive Sci. 9:159–191

Liu H, Xu C, Liang J (2017) Dependency distance: A new perspective on syntactic patterns in natural languages. Phys. Life Rev. 21. https://doi.org/10.1016/j.plrev.2017.03.002

Loukina A, Madnani N, Cahill A, et al. (2020) Using PRMSE to evaluate automated scoring systems in the presence of label noise. Proceedings of the Fifteenth Workshop on Innovative Use of NLP for Building Educational Applications, Seattle, WA, USA → Online, 10 July, pp. 18–29. Association for Computational Linguistics

Lu X (2010) Automatic analysis of syntactic complexity in second language writing. Int. J. Corpus Linguist. 15:474–496

Lu X (2012) The relationship of lexical richness to the quality of ESL learners’ oral narratives. Mod. Lang. J. 96:190–208

Lu X (2017) Automated measurement of syntactic complexity in corpus-based L2 writing research and implications for writing assessment. Lang. Test. 34:493–511

Lu X, Hu R (2022) Sense-aware lexical sophistication indices and their relationship to second language writing quality. Behav. Res. Method. 54:1444–1460. https://doi.org/10.3758/s13428-021-01675-6

Ministry of Health, Labor, and Welfare of Japan (2022) Retrieved from: https://www.mhlw.go.jp/stf/newpage_30367.html

Mizumoto A, Eguchi M (2023) Exploring the potential of using an AI language model for automated essay scoring. Res. Methods Appl. Linguist. 3:100050

Okgetheng B, Takeuchi K (2024) Estimating Japanese Essay Grading Scores with Large Language Models. Proceedings of 30th Annual Conference of the Language Processing Society in Japan, March 2024

Ortega L (2015) Second language learning explained? SLA across 10 contemporary theories. In VanPatten B, Williams J (ed) Theories in Second Language Acquisition: An Introduction

Rae JW, Borgeaud S, Cai T, et al. (2021) Scaling Language Models: Methods, Analysis & Insights from Training Gopher. ArXiv, abs/2112.11446

Read J (2000) Assessing vocabulary. Cambridge University Press. https://doi.org/10.1017/CBO9780511732942

Rudner LM, Liang T (2002) Automated Essay Scoring Using Bayes’ Theorem. J. Technol., Learning and Assessment, 1 (2)

Sakoda K, Hosoi Y (2020) Accuracy and complexity of Japanese Language usage by SLA learners in different learning environments based on the analysis of I-JAS, a learners’ corpus of Japanese as L2. Math. Linguist. 32(7):403–418. https://doi.org/10.24701/mathling.32.7_403

Suzuki N (1999) Summary of survey results regarding comprehensive essay questions. Final report of “Joint Research on Comprehensive Examinations for the Aim of Evaluating Applicability to Each Specialized Field of Universities” for 1996-2000 [shōronbun sōgō mondai ni kansuru chōsa kekka no gaiyō. Heisei 8 - Heisei 12-nendo daigaku no kaku senmon bun’ya e no tekisei no hyōka o mokuteki to suru sōgō shiken no arikata ni kansuru kyōdō kenkyū’ saishū hōkoku-sho]. University Entrance Examination Section Center Research and Development Department [Daigaku nyūshi sentā kenkyū kaihatsubu], 21–32

Taghipour K, Ng HT (2016) A neural approach to automated essay scoring. Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, Austin, Texas, 1–5 November, pp. 1882–1891. Association for Computational Linguistics

Takeuchi K, Ohno M, Motojin K, Taguchi M, Inada Y, Iizuka M, Abo T, Ueda H (2021) Development of essay scoring methods based on reference texts with construction of research-available Japanese essay data. In IPSJ J 62(9):1586–1604

Ure J (1971) Lexical density: A computational technique and some findings. In Coultard M (ed) Talking about Text. English Language Research, University of Birmingham, Birmingham, England

Vaswani A, Shazeer N, Parmar N, et al. (2017) Attention is all you need. In Advances in Neural Information Processing Systems, Long Beach, CA, 4–7 December, pp. 5998–6008, Curran Associates, Inc., Red Hook, NY

Watanabe H, Taira Y, Inoue Y (1988) Analysis of essay evaluation data [Shōronbun hyōka dēta no kaiseki]. Bulletin of the Faculty of Education, University of Tokyo [Tōkyōdaigaku kyōiku gakubu kiyō], Vol. 28, 143–164

Yao S, Yu D, Zhao J, et al. (2023) Tree of thoughts: Deliberate problem solving with large language models. Advances in Neural Information Processing Systems, 36

Zenker F, Kyle K (2021) Investigating minimum text lengths for lexical diversity indices. Assess. Writ. 47:100505. https://doi.org/10.1016/j.asw.2020.100505

Zhang Y, Warstadt A, Li X, et al. (2021) When do you need billions of words of pretraining data? Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, Online, pp. 1112-1125. Association for Computational Linguistics. https://doi.org/10.18653/v1/2021.acl-long.90

Download references

This research was funded by National Foundation of Social Sciences (22BYY186) to Wenchao Li.

Author information

Authors and affiliations.

Department of Japanese Studies, Zhejiang University, Hangzhou, China

Department of Linguistics and Applied Linguistics, Zhejiang University, Hangzhou, China

You can also search for this author in PubMed   Google Scholar

Contributions

Wenchao Li is in charge of conceptualization, validation, formal analysis, investigation, data curation, visualization and writing the draft. Haitao Liu is in charge of supervision.

Corresponding author

Correspondence to Wenchao Li .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Ethical approval

Ethical approval was not required as the study did not involve human participants.

Informed consent

This article does not contain any studies with human participants performed by any of the authors.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplemental material file #1, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Li, W., Liu, H. Applying large language models for automated essay scoring for non-native Japanese. Humanit Soc Sci Commun 11 , 723 (2024). https://doi.org/10.1057/s41599-024-03209-9

Download citation

Received : 02 February 2024

Accepted : 16 May 2024

Published : 03 June 2024

DOI : https://doi.org/10.1057/s41599-024-03209-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies