Research often misses the mark when it comes to studying human attractions. This is not entirely surprising, since they are derived from a complex combination of biological, evolutionary, social, and cultural factors.

Take, for example, the work of Dr. David M. Buss, famous evolutionary psychologist of the University of Texas, who approaches every attraction study with the question: How is this evolutionarily adaptive?,meaning, how does this trait or quality increase the likelihood that a person will reproduce, passing down his or her genes? His work often leaves out gay people, as gay folks weren’t traditionally reproducing.

Then there’s the modern work of Dr. Savin-Williams, of Cornell University, who recently made headlines by advocating for a new emerging sexuality: the mostly straight male. His work focuses specifically on how male attractions manifest themselves in the 21st century. His work is very time and culturally sensitive, and would be rendered completely useless to any man living outside of the United States or living 30 years ago.

While these two prominent researchers both study the nature of human attraction, each explore attractions through the narrow lens of his respective field. Moreover, both men probably don’t agree upon a singular definition of the word “attraction.” What constitutes attraction and what is its purpose? Is attraction physical or emotional? Is it for the sole purpose of reproduction or something else entirely?

Here’s the problem with academic studies. They tend to approach attraction from one specific lens, instead of looking at the whole picture. Or they’ll use only one tool to measure attraction, like hormone levels, pupil size, or self reporting. Approaching attraction from one angle overlooks many of its nuances. For example, one high-profile study from Northwestern in 2005 claimed bisexuality in men didn’t exist because the researchers couldn’t find a physiological bisexual arousal pattern. In the study, the researchers measured the circumference of gay, bisexual, and straight-identified men’s pensises while watching gay male and lesbian porn. Three-quarters of the time, bisexual-identifying men in the study had arousal patterns identical to those of gay men; the rest were indistinguishable from heterosexual men.

The results of this study has since been challenged by another penile circumference study, conducted by researchers at Northwestern in 2011. Akin to the previous study, the gay, bisexual, and straight-identified participants watched lesbian and gay porn. However this time, the bisexual men in the study showed a distinct physiological arousal pattern, meaning they became erect while looking at both types of porn, while straight and gay men did not. The reason for this might have to do with the researchers use of more rigorous scientific methods and a stricter criteria for bisexually-identified men. Nevertheless, measuring the circumference of one’s penis in response to watching porn solely measures physical attraction. There are also romantic, emotional, and aesthetic attractions, which cannot be measured this way. It’s possible, for example, to image getting married or spending a lifetime together with the person one has a crush on. Imagining that doesn’t necessarily include physical aroused. But one may become physically aroused when imagining having sex with that crush.

Additionally, many women and men show physiological arousal signs when experiencing sexual assault or rape, but that doesn’t mean they’re attracted to the aggressor. Other factors, unrelated to arousal, are at play. For women, it’s known that the vagina can become lubricated during rape. This isn’t due to arousal; lubrication acts acts a defense mechanism against tearing and pain.

Dr. Sari M. van Anders and her colleagues at University of Michigan’s Psychology & Women's Studies Department published a paper in The Journal of Sexual Research last November titled, "Heterogeneity in Gender/Sex Sexualities: An Exploration of Gendered Physical and Psychological Traits in Attractions to Women and Men." Their research attempted to use a broader approach to studying attractions, one that focuses more on gender expression as opposed to sex characteristics.

“Previous research has assumed that genitals are what determine people’s attractions to women and/or men.” Dr. van Anders tells NewNowNext. “But the lived experiences of many people, including many sexual minorities, contradict that assumption. We wanted to focus in particular on how gender—that is, the social presentation or femininity, masculinity, and gender diversity—might play a role in attractions as well as sex.”

In addition to genitals, past research also tends to focus on other physical sex characteristics. Thus, a study that explores attraction to a “masculine face” would use an image of a man with a strong jawline and beard. While a strong jawline and beard are examples of male sexual dimorphisms (i.e., physical traits of being born male), sexual dimorphisms don’t take into account gender expression, which must be relevant to understanding attraction.

For instance, one particular man with a strong jawline and beard might typically wear makeup and jewelry. Would the people who found a man attractive for his physical male traits still find him attractive if he were presenting himself in this way?

That’s why the work of Dr. van Anders and her colleagues is so important. They elevated sexuality and attraction research through exploring how humans are attracted to gender expression and not just sex characteristics. For their study, the researchers recruited participants by posting advertisements locally in newspapers, magazines, and websites like Craigslist.

The sample used for their paper consisted of 280 individuals: 174 identified as women (173 cisgender, one transgender), 103 as men (102 cisgender, one transgender), two as genderqueer, and one person didn’t indicate any gender. The age ranged from 18–76, with the average age being 32. Additionally, 192 of the subjects, (69%) identified as straight, 53 (19%) as bisexual/pansexual/queer, and 32 (8%) as homosexual. A negligible percentage of folks identified as asexual, unsure, or “mostly straight.”

The researchers asked the participants two open-ended questions. “What do you find attractive in a woman (or man)?”

Their responses were coded by two researchers, looking for the presence or absence of feminine psychological features, (e.g., sweet, nurturing) feminine physical features (e.g., soft skin, big breasts), gender non-specific psychological features, (e.g., similar interests), gender non-specific physical features, (e.g., smile) and then lastly, masculine psychological features, (e.g., rationality, independence) and masculine physical features (e.g., strong arms, big shoulders).

Interestingly, for both men and women, respondents named gender-nonspecific traits significantly more than either masculine or feminine traits. In fact, in men, only 38% of participants mentioned masculine physical and psychological traits, whereas 55% of participants named physical gender-nonspecific traits, and 47% named psychological gender-nonspecific traits.

For women, 49% of participants named physical feminine traits and 14% named psychological feminine traits, compared to the 59% of participants who named gender-nonspecific physical traits and 45% of gender-nonspecific psychological traits.

It’s particularly interesting to note that even in women, participants preferred masculine psychological traits (31%) to feminine psychological traits (14%). It seems as if people in the study were attracted to a woman who “looks like a woman but acts like a man.”

“This relates to femmephobia as well as [the] objectification of women and femme-identified individuals,” Dr. van Anders says.

There were a number of limitations to the study. For one, the question was very open-ended, so it’s unclear exactly what participants were responding to. Was it sexual attraction? Romantic attraction? Something else? Two, and more notably, the researchers coded many of the vague responses as gender-nonspecific. For example, a number of respondents wrote in “body” for what they found attractive. The participants could have meant a specifically masculine or feminine body, however because they didn’t clarify, their responses were coded as gender-nonspecific.

So to be honest, the results aren’t as compelling as they may initial appear to be. It doesn’t necessarily mean that a growing number of people are more attracted to gender-neutral characteristics than classic masculine or feminine traits.

Nevertheless, this research at least attempts to present a broader approach to attraction—one that’s inclusive of gender and gender expression. As our understanding of gender, sexuality, and attractions become more complex, so too, must academic research follow suit.