Journalism blows it big time in reporting a study on the objectification of women.

Oct 29 2013 Published by under Psychology

I picked this up from Jezebel. They used the headline of "Depressing Study: Men Look More At Your Body Than Your Face" and said:

"I'm not a boob man or a butt man or a leg man, I'm a face man," say LIARS. According to new research, no matter what a woman's build, men spend more time looking at women's bodies than they do their faces, ...But it's not only men who are focusing on women from the neck down; women do it to each other, too

they cited USA Today, which is our first tip that they didn't bother to read the study in the first place. This lack of reading may have gotten them further into trouble since Jezebel even had the nerve to criticize the design.

here's a grain of salt to take with this: the study involved 29 women and 36 men, a group so small that it would be almost impossible for it to be at all representative of the population. If subjects were drawn from a pool consisting of a public university community, the results would only reflect the attitudes and behaviors of a tiny slice of American culture and not a boob staring epidemic. And a lot of college kids are kind of awful.

While college sophomore psychology students (see below) are notoriously used in Psych studies and equally notoriously poorly representative of many populations of interest...they are not dismissible as entirely meaningless. Often times results from such studies do indeed hold up when replicated in other populations of interest. Nevertheless, this Jezebel comment is pretty hilarious considering how completely backwards they got the story on the actual findings. Which is in large part due to simply passing on the bit from USA Today instead of reading the paper.

From the USA Today article titled "Yes, men really do ogle women's bodies":

The eyes don't lie: Men really do look at women's bodies more than their faces, according to a new study that used eye-tracking technology to prove what many women have long observed.

But it's not just men who do it -- the study found that women look at other women's bodies, too.
...
Both sexes fixed their gaze more on women's chests and waists and less on faces. Those bodies with larger breasts, narrower waists and bigger hips often prompted longer looks.

Truthy! Just what we always suspected and now here it is in peer-reviewed scientific format!

Let's go to the article, shall we? The study by Gervais and colleagues (2013) is pretty simple. The authors recruited some subjects from the traditional "Psychology Department Participant Pool" (aka, students enrolled in Psych classes) and fitted them with eyetracking devices. They showed them some pictures of women and asked them to rate them for either appearance or personality (separate groups for each of these conditions).

Gervais13-appendixThe stimuli were photographs of 10 real women which were modified slightly. Here is the figure listed as the Appendix in the article so you can see how the photoshopping of the visual stimuli worked out. The major dependent measure was "dwell time", i.e., how long the subject spent with their dominant eye focused on one of three zones of the picture (face, chest, waist). They also measured "first-fixation" but this was somewhat contaminated by the fact the fixation cross used to start a trial was in the center of the screen where the chest would appear. So a missed opportunity there. The dwell time is the major outcome measure for discussion.

One of the main goals of the study was to determine if "High, Average and Low" concordance with what they describe as the "cultural ideal" body shape affected the distribution of gaze time. Also to determine if men and women subjects differed and if the type of rating being requested of the subjects altered dwell time.

The results could not be any clearer. Both men and women spent more time gazing on the Face region then they did on the Chest or Waist region. By a lot. Whether asked to assess Appearance or Personality. In the Appearance condition, women spent 1158 ms on the Face, 463 ms on the Chest and 331 ms on the Waist. Values for men were 1296, 448 and 301 ms respectively. When divided out by the three categories of "cultural ideal", Men's dwell times were 1520, 456, 280 ms for the High ideal and 1628, 366, 246 ms for the Low ideal. The same relationships held for the women viewers.

The authors note in the Results:

A main effect of body part, F(1.09, 66.25)=215.68,p < .0001, ηp 2 = .78, revealed that women’s faces
(M=1486.61, SE =64.17) were gazed at for longer durations than their chests (M=381.68, SE =23.33) and their waists [DM-pretty sure this is a typo and meant to be 'faces'] (M=266.62, SE =16.04) and women’s chests were gazed at for longer durations than their waists, ps<.0001.

So. The lede of both USA Today and Jezebel is completely false.

Now, there IS a portion of blame for the authors because they are at pains to emphasize their findings; again from the Results:

consistent with Hypothesis 1a, participants gazed at women’s faces for shorter durations in the
appearance-focus condition than the personality-focus condition. Participants also gazed at women’s chests and waists for longer durations in the appearance-focus condition than the personality-focus condition.

...and perhaps more tellingly from the first part of the Discussion:

Despite the importance of the objectifying gaze to objectification theory (Fredrickson and Roberts 1997) and the adverse consequences of the gaze on women recipients, no published studies to date have empirically documented the nature of the objectifying gaze—less focus on faces and more focus on sexual body parts—in perceivers. Regarding dwell time, participants gazed at women’s faces for shorter durations and chests and waists for longer durations when they were asked to objectify the women by evaluating their appearance (vs. personality, consistent with Hypothesis 1a) and this effect was exacerbated for women with bodies that fit cultural ideals of beauty (i.e., hourglass shaped women, consistent with Hypothesis 1b).

Very careful phrasing there indeed and I can see how "participants gazed at women’s faces for shorter durations and chests and waists for longer durations" would be very easily misinterpreted in the reader's mind as suggesting that faces were receiving less gaze time than were the other regions of the pictures.

But really. A cursory look at the Tables makes the fact that both men and women spent more time gazing at faces than at chests or waists pretty dang obvious. It almost pops out, thanks to the convenient fact that dwell time differed across the 1,000 ms mark. I just don't see how you could miss this if you read the article.

Of course, the journalists didn't read the article.

And this is considered perfectly acceptable within the profession of Journalism.

Updated:
Scienceblog.com barely escapes by sticking close to the authors words without extraneous interpretive phrasing. "The researchers found that participants focused more on women’s chests and waists and less on faces when they were asked to objectify the women by evaluating their appearance rather than their personality."

ScienceCodex screwed it up, "When asked to focus on a woman's appearance, study participants largely looked at women in "that way" – they quickly moved their eyes to and then dwelled on a woman's breasts and other sexualized body parts.".

The South Jersey Courier-Post fell entirely into the trap, "Men really do look at women’s bodies more than their faces, according to a new study", just like USA Today and Jezebel.

as did CBS Philly, "scientists concluded that participants focused more on the female’s chests and figure when asked to evaluate their appearance than they did on the women’s facial features.".

UPDATE 2: Aha! found the press release from the authors' University. OOOOO, Bad authors! BAD! "When asked to focus on a woman's appearance, study participants largely looked at women in "that way" -- they quickly moved their eyes to and then dwelled on a woman's breasts and other sexualized body parts.". So Science Codex just stenographed this. But the blame lies with the authors who should have reviewed the PR. And given how very precisely this is written I suspect them of willing complicity. The impression is given via "and then dwelled on" that this is gazing more than on the non-"sexualized" body parts, i.e., the face. Or hey, maybe I'm wrong and they include face as sexualized and waist as nonsexualized? In which case this is accurate-ish.
__
Sarah J. Gervais, Arianne M. Holland and Michael D. Dodd. My Eyes Are Up Here: The Nature of the Objectifying Gaze Toward Women, Sex Roles, in press DOI: 10.1007/s11199-013-0316-x
[link]

30 responses so far

  • Hermitage says:

    Jezebel has always had embarrassingly bad science news coverage. It would be more remarkable to find a time where they reported a scientific article correctly.

  • drugmonkey says:

    Keep reading Hermie. It is not just Jezebel that is screwing this up. Also, this is not just a minor lack of correctness on details.

    Jezebel and some of the others are reporting this in a manner that is exactly contradictory to what the authors found. A reversal of sign.

    The study actually found that men and women spend at least twice as much time looking at the face region of the images as they do looking at the chest or waist region.

  • Hermitage says:

    I did ^^, it is surprising that other publications also got it wrong, but leading with Jezebel getting it wrong, to me, is like saying Republicans misreport how SNAP benefits work. It kind of a given it'll be terrabad.

  • jipkin says:

    Why is there no heatmap of gaze position vs time figure in their paper? If you're going to go for sensationalism, you need an accompanying visual. Even the raw data of each point of recorded fixation or whatever would be fun. Even more fun: get rid of the fixation cross confound and report the temporal order/patterns with which men and women view the target picture. Even with the confound, there's data like fixation episode length/number/pattern that they should have. And why are the legs cut off? why not throw those in there. Would you get a gender difference in fixation time of the shoes?

    So much missed opportunity for legitimate sensationalism.

  • drugmonkey says:

    so Hermitage, you are saying this is my own damn fault for following Jezebel? some truth to that.

  • drugmonkey says:

    And why are the legs cut off? why not throw those in there.

    Especially given the Jezebel comment about "leg man", eh? I had that thought as well and I came up with two obvious problems from an experimental perspective. ok, maybe three. First, can you get everything in one frame and still have distinguishable gaze zones? Second, what aspect do you show the models in for maximum competition between face and legs, say? And you have to throw the butt comparo out entirely? probably? Thirdly, I would tentatively hypothesize even greater variance in what an individual observer thinks of as attractive legs or butt versus the variance in breast appearance. (Even within the homogeneous psych-class population. ) oh, and Fourth, how do you photoshop manipulate the leg and/or butt appearance?

    Then we get to the interaction of ethnicity of both observer and models...and all kinds of other interesting factors.

    As with the Implicit Association Tests, I think these folks have an interesting model that could extend to a number of additional comparisons. Some may be difficult to validate but hey, worth the try.

  • drugmonkey says:

    ....oh and of *course* you are all with me in thinking a field study with Google Glass(es) is in order, correct? [and no, not Arikia's little experiment]

  • jipkin says:

    Well DM, if we're going to go that far then why not go all the way and use 3D models where the user can rotate them as they like (the rotation and eye fixation can be recombined digitally). This is possible to do photorealistically too (think google streetview). Then you can slap a wire mesh on the person and move their butt and boobs and legs all around and adjust the textures as needed.

    Soon we will have the video game industry funding scientists to find out what the optimal Lara Croft body type is...

  • Grumble says:

    And we sit around wondering why the public is OK with science budgets getting cut.

  • DJMH says:

    I'm not going to defend Jezebel as a proud purveyor of lay science, but c'mon, if the authors' own PR system cast it like this, I sure wouldn't jump too hard on a website that bills itself as "Celebrity, Sex, Fashion, For women."

    Also Jezebel, unlike many many other sites out there that supposedly cater to women, has an active and interesting feminist take on news of all kinds, and so I cut it slack on something stupid like this.

  • DrugMonkey says:

    DJMH- c'mon but it is *every* news outlet doing very little beyond reprinting the PR. Not *one* bothered to read the paper? And you don't think we scientists need to address this?

  • dr24hours says:

    How many scientists are willing to have their work presented incorrectly in order to get attention? My guess: lots.

  • Ola says:

    Since when is the face a "non sexual" body part? Have they not heard the alternative definition of the abbreviation "DSL"?

  • vuncksa says:

    Towards the comment about the author being complicit in the misrepresentation of the findings.

    This was in the USA Today article as a quote from the first author.

    "We live in a culture in which we constantly see women objectified in interactions on television and in the media. When you turn your own lens on everyday, ordinary women, we focus on those parts, too," says lead author and social psychologist Sarah Gervais of the University of Nebraska-Lincoln.

    "Until now, we didn't have evidence people were actually doing that to women's bodies," she says. "We have women's self-reports, but this is some of the first work to document that people actually engage in this."

    This makes it sound like what she wanted to find and what the study actually found weren't the same so she's trying to spin it.
    However, those sentences could also have been taken out of context.

  • DrugMonkey says:

    Exaggeration is one thing, dr24hrs, but this is a complete reversal of the actual finding. I can't see many being totally happy with that.

  • dr24hours says:

    @DrugMonkey

    Sadly, I can. Easy to rationalize - "They got this totally wrong, but it raised my profile and as a result, people will look at my papers and my future work will likely receive publicity. In fact, the people who matter (like that drugmonkey-guy) will blog about it and put the truth out there and get me even more publicity! And since my work is important and life-changing, the part that really matters is that it gets exposure, even if some of it is wrong in the media."

  • Dr Becca says:

    How many scientists are willing to have their work presented incorrectly in order to get attention? My guess: lots.

    As someone who has had her work misrepresented in the media, I can tell you that this is flat out wrong.

  • DrugMonkey says:

    Sure, dr24hours, the media surge can be exciting. But don't you think most scientists would rather it was for reasons of possible exaggeration over being flat out incorrect? At some point you realize other people in your field might question your acumen.

  • […] case study. Consider how this affects how we think about issues like […]

  • dr24hours says:

    @Doc_Becca, Jebus I hope you're right!

    @DrugMonkey, "Rather"? Of course! I'm sure they'd rather it be for accurate, informative, exciting descriptions. But given a choice strictly between "wrong coverage" and "no coverage", I bet a lot of scis would take "wrong".

  • TallDave says:

    OTOH ScienceBlog is generally terrible at actually examining methodology critically.

  • TallDave says:

    "How many scientists are willing to have their work presented incorrectly in order to get attention?"

    Some mistakes are more convenient for authors than others. Reminds me of when Hansen was interviewed (in the 1990s iirc) and the reporter wrote that he said the West Side Highway would be underwater in 20 years. Well, it didn't happen of course, and when asked Hansen claimed he meant a much less scary 40 years, and the reporter says his notes agreed, but one suspects Hansen was not unhappy to be misquoted with a scarier number at the time. (Of course, neither is remotely likely to actually happen, but modern science is apparently no longer about testable predictions. )

  • sciencedude says:

    " Have they not heard the alternative definition of the abbreviation "DSL"?"

    All that comes up in the first two pages of Google is digital subscriber line, even when I include "alternative definition" in the search.

    Besides the obvious misrepresentation of the actual results, I guess I am wondering more where the actual newsworthiness of this headline is. Of course men look at women's bodies, and the hotter they are, the more I am probably going to look. That does not mean I don't respect women; it is simple biology. Besides, women themselves decide how much they want men to look by how they dress. If they don't want people to look, they simply dress in loose fitting clothing. In the middle east they call it a burka. Unfortunately there, it is not usually a choice.

  • drugmonkey says:

    I guess I am wondering more where the actual newsworthiness of this headline is

    I think the apparently counter-intuitive* finding that men spend more time looking at the face when asked to judge appearance *or* personality of women should have been the amazing and newsworthy lede here....

    *see commentary on media reports amounting to "yeah, so what? like we need a study to prove men look at the boobies?" and similar.

  • Dr Becca says:

    If the instructions to subjects were literally "judge the appearance of this person" then I think it's not all that surprising that they spent more time looking at the face, because I'd still imagine that facial attractiveness is the #1 important thing to people's overall evaluation of attractiveness, regardless of whether they're "boob guys" or whatever. Personality, obviously more face because facial expressions are how we understand people's general 'tudes. They should have asked more specific evaluative questions like, "Do you think this person is intelligent/ is nice / would be a good parent / is fun to hang out with / is good at sports" etc.

  • anonymous postdoc says:

    I once thought about starting a blog whose sole purpose would be to correct the science reporting on Jezebel, since a large number of (presumably) otherwise intelligent people totally fall for it every time, as evidenced by the comment section. My kind of research is routinely pilloried by Jezebel.

    Then I realized the magnitude of work such an operation would require, and the severe and detrimental effect it would have on my ability to do actual science. I therefore decided to simply stop ever reading Jezebel.

  • […] of gobbledygook, here is a review of the article about men ogling women. It sounds like the authors misquoted their own […]

  • Hayden says:

    A bit late, but one of the authors (Sarah Gervais) wrote this in Psychology Today (on the web Nov 1):

    "Now before you chalk this study up to “common sense,” let me highlight a few surprising discoveries from the study that paint a more optimistic picture of objectification.
    Overall, people focused on women’s faces more than their bodies and focusing people’s attention on the personalities of women further reduced the objectifying gaze."

    http://www.psychologytoday.com/blog/power-and-prejudice/201311/my-eyes-are-here

    Her conclusion (message?) is very different from any of the press release reprints:

    "...objectifying gazes communicate to women that they are being seen as sex objects rather than people (Fredrickson & Roberts, 1997). Not surprisingly, the objectifying gaze has several negative consequences for women such as making them anxious about their appearance (Calogero, 2004), decreasing their work performance (Gervais, Vescio, & Allen, 2011), and making them feel like they should speak up less (Saguy, Quinn, Dovidio, & Pratto, 2010).
    So, all you guys and gals out there, if you notice your eyes meandering to places they shouldn’t be, remind yourself that you are interacting with another human being with a personality and hopefully your roving eyes will follow suit. "

    But that could be the sound of someone paddling backwards very quickly....

  • bud says:

    The design of the study has another flaw: the pictures used exactly the same face for each example. The only differences were in the chest and waist region. If I'm trying to decide between different things, I'm going to focus on the difference, and therefore I would expect lesser gaze times on the face.

    Which leads to the question of whether the authors constructed a biased study, expecting to "prove" the objectifying gaze, and when that didn't work out, summarized in a confusing manner to mislead reporters, causing the desired results to be disseminated.

  • drugmonkey says:

    There were ten people used in creating the stimulus set.

Leave a Reply