Last week, Penn President Liz Magill reposted a study deeming Penn among the top 10 happiest universities nationwide. Penn placed first in the Keystone State, thereby defeating Penn State University also known as the happy valley dating back to the 1930s. But wait – there’s a catch.
The study, conducted and published by Resume.io – a company that helps job seekers build resumes predicated on rules and structures – aimed to explore which institutions of higher education contained the most cheerful students. First, the software perused Instagram posts from universities' geotagged locations. Then, it assessed them on a happiness metric with Amazon Rekognition API, an AI facial emotion recognition tool. The sampling pool was limited to only public accounts. Essentially, the study evaluated happiness based on external appearance.
At face value, the article seems intriguing in its click-baited headline, but after digging deeper, the unreliability surfaces immediately.
Not only does the study contain flawed methodology by discerning University happiness through social media, but it also speaks to the larger problem of perpetuating Penn Face and inauthenticity. It is harmful to circulate this study without investigating its validity. By reposting it, administrators are only feeding into Penn Face and participating in it themselves.
As president of the University, Liz Magill must be in tune with the mental crisis here. Yet, she hastily reposted this study as a performative act to present Penn mental health in a better light. If she actually read it, she would see its overt flaws.
Instagram, and all forms of social media for that matter, are precisely curated highlight reels. People tend to frame and glamorize their feeds with positivity while omitting the realities behind the scenes, parallel to how the concept of Penn Face functions.
“It makes sense that we’re deemed a happy college if the metric Resume.io is using is running AI on our Instagram posts to see if we express happiness — social media is incredibly curated and Penn Face is always prevalent,” College sophomore Yash Mahajan said. “Not many people would post unhappy pictures on a public social media page.”
Oftentimes how people represent themselves on social media bears no resemblance to their true lives. In 2018, a Twitter account asked for people to share pictures in which their photos projected themselves in a way that did not encapsulate the difficult times hidden behind their smiles. With thousands of comments highlighting this phenomenon, this post was just one example illustrating how smiles in pictures are a futile metric for perceiving happiness.
An additional limitation to the study is considering non-student photos within the sample. While navigating their future academic paths, visitors are attracted to universities to obtain a first-hand glance of a college’s atmosphere and offerings. After doing so, they may take pictures and share them on social media while tagging the location of the college. This is a critical factor distorting the accuracy of this experiment – pictures of non-students could skew outcomes for schools, especially Ivies which receive eager visitors.
Further, confounding variables such as a school’s social culture may also influence the results. Known as the Social Ivy, “Penn, which has a reasonably strong party scene, would probably have more really happy faces posted by students during social events, but schools that are a little bit more low-key would have fewer of those even if the student body is happier most of the time,” Wharton sophomore Aneesh Karuppur said.
Although Penn contains well-rounded students, “all this really means in practice is that there are even more angles from which pressure on students gets compounded. That stress carries through to the social interactions students have around campus,” College sophomore Sajan Srivastava said. Thus, it makes sense that this study was reposted in that “top schools like Penn would love to [appear] that their students have fun despite the rigor of their coursework.”
Srivastava mentioned, “I highly doubt that any AI model is intelligent enough to look past the smiles and see what happiness truly looks like in non-candid photos of students; there’s a reason that just about everyone at Penn has either heard about the so-called “Penn Face” or borne it.”
In this scenario, failing to acknowledge sampling and experimental errors serves as a misrepresentation to prospective applicants and minimizes the lived experience of current students. Despite its illegitimacy, it is unsurprising that this study was reposted as a marketing strategy.
As Mahajan noted, “Posting that this is one of the happiest schools in the country sets us back in destigmatising Penn Face.” Administrators, and particularly the president of this University, cannot continue to give strength to this term.
Overall, this study merely shows which universities boast the highest number of faces detected as ‘happy’ between the restricted timeline between December 2022 and January 2023. Equating social media with human emotions is simply fallacious.
Instead, the study’s misleading title should be revised to qualify its findings and say, “Universities Tagged with the Highest Proportion of Smiles on Instagram,” because that’s all it truly revealed: a picture perfect perception.
RIANE LUMER is a College junior studying political science and journalistic writing from Huntingdon Valley, P.A. Her email address is firstname.lastname@example.org.