A recent study published in the Canadian Journal of Behavioural Science has shed light on how people judge unfamiliar faces on dating apps, particularly when these faces are enhanced with photo filters. Subtle photo filters—those that make small, barely noticeable changes to appearance—tended to increase how much people “liked” a face. In contrast, exaggerated filters, which make faces look more artificial, had no impact on the level of “liking.”
Online dating apps like Tinder rely heavily on first impressions, with users making split-second decisions based on profile pictures. In recent years, the use of photo filters—tools that allow people to alter their appearance—has skyrocketed, particularly on social media platforms and dating apps. Filters range from playful (such as turning someone into an animal) to more subtle adjustments that make a face look smoother, younger, or more symmetrical.
Given the importance of visual cues in attraction, researchers wondered how these filters might influence judgments of trustworthiness and likability, particularly in the context of dating apps where such judgments can have real-world consequences. For example, if people use heavily filtered images that alter their appearance drastically, they might create a false sense of trust, which could later lead to feelings of betrayal when meeting in person.
“I like to investigate aspects of social perception that are relevant in everyday life,” said study author Antonio Olivera-La Rosa, a full professor at Luis Amigó Catholic University in Medellín, Colombia. “The use of facial filters on social networks/dating apps is ubiquitous, people often use filters as ‘virtual makeup’ to increase attractiveness and to receive more positive responses (e.g., liking). However, the effects of facial filters on social perception is surprisingly understudied. When we talked about it with Gordon Ingram and Erick Chuquichambi, it seemed like a logical step, since we had previously investigated facial perception of masks during the pandemic.”
For their study, the researchers recruited 218 participants from a Colombian university, representing a diverse range of ages, sexual orientations, and educational backgrounds. The study was designed to simulate a mobile dating app environment, allowing participants to make rapid judgments about unfamiliar faces, as they would in a real-world dating app like Tinder.
The study used 24 different faces, including 12 male and 12 female faces, each presented in three versions: unfiltered, subtly filtered, and exaggeratedly filtered. The subtle filters made small, difficult-to-detect changes to the face, such as smoothing the skin or enhancing features slightly, while the exaggerated filters made the face look more artificial, such as enlarging the eyes or altering facial symmetry in a noticeable way.
The participants were shown each face in its different versions and asked to rate the faces based on two criteria: how much they liked the face and how trustworthy they found it. These two tasks were designed to mimic the “swiping” action in dating apps, where users make quick judgments about whether they want to connect with someone.
In addition to rating the faces, participants completed a questionnaire that measured their sociosexuality, which refers to their openness to casual sexual relationships. This measure allowed the researchers to see whether participants with more unrestricted sociosexuality—those who are more inclined to engage in short-term, uncommitted sexual encounters—were more likely to trust or like certain types of faces, particularly those that were filtered.
The researchers found that individuals with higher sociosexuality scores were more likely to dislike unattractive faces, but their liking of attractive or neutral faces was not significantly different from others. Individuals with a more unrestricted sociosexuality were also more likely to trust faces compared to those with a more restricted sociosexuality, irrespective of the type of filter applied or the attractiveness level of the faces.
“We found that sociosexuality (one’s propensity for uncommitted sexual behaviour) predicted higher trustworthiness judgments when evaluating faces in an app-like interface,” Olivera-La Rosa told PsyPost. “Therefore, our results suggest that sociosexuality is a psychological predisposition towards trust that may constitute a risk factor in online contexts. As we mention in the paper, ‘if someone you have matched with is more sociosexual than you, their ‘like’ may not mean that they find you as attractive as you find them; while if they are less sociosexual than you, they may not trust you as much as you trust them.'”
As expected, attractive faces were rated as more trustworthy and were liked more than unattractive faces. Neutral faces fell in between. The findings support the “beautiful-is-good” stereotype, where attractive people are often perceived more positively across various traits, including trustworthiness. Interestingly, the bias against unattractive faces was stronger than the positive bias for attractive faces. This means unattractive faces were not only liked less but were also judged more harshly in terms of trust.
Another significant finding was that subtle photo filters generally increased how much participants liked a face, while exaggerated filters had no impact likability. Surprisingly, the level of filtering also did not significantly affect trustworthiness judgments.
“We expected that exaggerated photo filters (e.g., with extremely enlarged eyes), would decrease trustworthiness judgements and liking responses,” Olivera-La Rosa said. “However, this was not the case. In general, our findings suggest that the effects of the ‘strength’ of photo filters on social perception may be limited to attractiveness judgments, with no effect on perceived trustworthiness.”
The researchers also found that unattractive male faces were judged more harshly than unattractive female faces. This finding aligns with other studies suggesting that men who are perceived as unattractive may face greater social penalties, particularly when they are judged as less trustworthy or more threatening.
As with any scientific study, this research had its limitations. The sample was relatively narrow, consisting mainly of young Colombian university students. While this group represents a relevant population for dating apps, it may not reflect how people in other age groups, countries, or cultural backgrounds perceive filtered faces.
In addition, the study focused specifically on a dating app-like context, which may not reflect how people make social judgments in other online environments, such as professional networking or social media platforms that emphasize different kinds of interactions. The researchers suggest that future studies could explore how filters and sociosexuality impact trust in non-dating contexts, like Facebook or LinkedIn.
“With AI, people are going to be using filters even more, and even whole artificial video avatars, so it’s going to be very important to see what factors influence how much people trust these new online visual representations with which they are interacting,” Olivera-La Rosa said.
The study, “An Experimental Study on the Role of Sociosexuality in Judgements of Perceived Trustworthiness in a Mobile Dating-Like Interface,” was authored by Antonio Olivera-La Rosa, Erick G. Chuquichambi, and Gordon P. D. Ingram.