This post is a summary and critical review of the paper, Men’s Faces Convey Information About Their Bodies and Their Behaviour: What You See is What You Get, written by Shoup and Gallup (2008). All of the statistical details to be mentioned can be found in the paper, here: (http://psycnet.apa.org/psycinfo/2008-13605-011).
They started off with sound citation of related research. The most significant of which was to do with the variables that they proposed to test: shoulder-to-hip ratio (SHR), hand grip-strength, sexual behaviour and facial attractiveness; all of which in males. In short summary of their introduction, all of these variables have a tendency to predict each other in various ways (Evans, 1972; Franzoi and Herzog, 1987; Hughes, Dispenza and Gallup, 2004; Gallup, White, and Gallup, 2007). This study was designed to test an amalgamation of the general findings discussed in their introduction by looking at their relationship to facial attractiveness. A valiant cause, especially with the right data analysis. This is, potentially, the kind of study that pulls a lot of ideas together and tests the resulting hypothesis in a way that validates an all-round interesting idea. In this case, the fascinating question being tested is: is female judgement of male facial attractiveness related to male SHR, hand grip-strength and/or sexual behaviour?
The participants used to create the stimuli were undergraduate males with a mean age of 20 years whose grip-strength was measured using a Lafayette Instruments Model 78010, SHR by using the Hughes and Gallop (2003) method and information about sexual behaviour, including age of virginity loss and total number of sexual partners, was collected using a questionnaire. All valid methods of data collection, despite the fact that the information collected from the questionnaire was probably subject to the biggest example of truth-stretching since Grov, Wells and Parsons (2013). A questionnaire was more than likely the most practical option for collecting information about participants’ past behaviour.
The stimuli were photographs taken only from the neck up, with the same lighting and back-drop as each other and all hats and accessories removed. However, there was no mention of controlling the stimuli for facial hair, which is a problem that will be returned to when assessing the discussion section. Females of the same age judged the stimuli with no time limit, using a 5-point Likert scale and the stimuli were randomised and split into two sets which were judged separately by female participants in independent groups. Females were asked not to rate males that they recognised, and menstrual cycle and birth control were noted. However, this is the last to be heard of the menstrual cycle and birth control variable/variables, leading the reader to assume that there was no interesting relationship found in this area.
As has been found in research previous to the current study, specifically the studies mentioned in their introduction, there were significant moderate to low-moderate univariate correlations (r = 0.5 to 0.3) between SHR and grip-strength, grip-strength and number of sexual partners, grip-strength and facial attractiveness and SHR and facial attractiveness. However, a multiple regression was carried out on these variables, finding the only significant semi-partial correlation to be between facial attractiveness and SHR. This is interesting to note, as it tells the reader that, when all of the other variables are accounted for, the relationship that stands out is between these two variables. That is to say, the more attractive that the male face was rated by a female, the higher that male’s SHR was likely to be, and these have only simple positive correlations with the other variables.
In the discussion section, the current study highlighted how the female ratings of male attractiveness accounted for most of the variance in SHR, sexual behaviour and grip strength in male university students. However, the simple correlations and multiple regression displayed in the results section were not enough to state this, and no ANOVAs were shown. In fact, the multiple regression that was shown suggested that a common significant extraneous variable in the simple correlations was SHR or facial attractiveness.
The current study then went on to discuss how hormonal and genetic factors could possibly influence both the face and the SHR, accounting for their relationship. This is where the problem of not controlling the stimuli for facial hair came in. Considering the current study’s notion of hormonal influence, it should be remembered that facial hair is affected by testosterone (Lee, Jaffe and Midgley Jr, 1974). Failure to display information about facial hair of stimuli suggested that it was not controlled when creating the stimuli, and so is a possible extraneous variable to consider.
The current study also discussed a non significant negative correlation between age of first sexual encounter and SHR. Why a non significant correlation was discussed as if it is significant was not explained. This was also linked to promiscuity in the current study, and the negative correlation between age of loss of virginity and number of sexual partners was then mentioned, despite this correlation also being non significant.
It was also mentioned that grip strength is not influenced by environmental factors. This is simply not the case. Various external influences can affect grip strength, both negatively and positively (Dash and Telles, 2001; Krafi, Fitts and Hammond, 1992). This does not matter in relation to the current study’s findings a great deal, and is actually contradictory to their own introduction. It was, however, mentioned as a point in their argument within their discussion section and is categorically incorrect. The current study then goes on to talk in relatively detailed levels about intrasexual selection in grip-strength via handshakes, with scant regard to its own findings, methods and research question.
Considering the research question, which was essentially a test of an amalgamation of various relationships between variables, some form of multilevel linear model (MLM) may have been a more appropriate research method. Although, perhaps this method was avoided because of the independent groups measure implemented, as a MLM can be seen as a measure to correct for the potential problems in a repeated measures design. The findings were still interesting, though, distinctly in the results section of the paper. From these, we can tell that men’s facial attractiveness is very much related to the attractiveness of their body morphologies. If the study’s discussion section is to be followed, there is a very real likelihood that this is due to testosterone levels. That is to say, men’s faces are attached to their bodies. In all honesty, though, this is a very interesting paper from a very interesting field.