Answers are grounded in this paper and cite their sources.
This research assesses the extent to which facial appearance is related to occupational status attainment. Through the use of Internet technology, a diverse random sample of research participants viewed Navy boot camp photographs and rated sailors across nine dimensions. Sailors who were rated highly attractive were also thought to be intelligent and to possess leadership qualities. And, consistent with social expectancy theory and status generalization theory, sailors who were rated high across these three traits advanced to higher ranks, and did so more rapidly, than those rated low on these traits. The findings of this study underscore the sociological significance of facial appearance as a means by which people are stratified within social structural contexts. Acknowledgments The authors wish to thank David Kenny, whose insights and comments on previous versions of this article were invaluable. The research presented here was funded in part by a grant from the National Science Foundation (0094964) under the Time-Sharing Experiments for the Social Science (T.E.S.S.) program, Diana C. Mutz and Arthur Lupia, Principal Investigators. We also thank three anonymous NSF grant reviewers for comments that helped shape the final version of our study. Notes 1For more information about the effect of initial impressions on others' judgments, see the "thin slices" literature. "Thin slices" of behavior refers to any excerpt of expressive behavior less than five minutes long, not including photographs (Ambady et al. 2000). 2Our sample includes those who served during the Gulf War in 1991, but we are unaware of any haphazard wartime advancements during this period for any of our targets or anyone in the service. 3This methodology was made possible through utilization of a private research company (Knowledge Networks Inc.) hired by T.E.S.S. (Time-Sharing Experiments for the Social Sciences), a National Science Foundation–funded program offering researchers opportunities to test their research ideas on large, diverse, randomly-selected subject populations. 4An alternative method of analysis would be to use Hierarchical Linear Modeling (HLM), with photograph characteristics at level 1 and participant traits at level 2. We experimented with this method and discovered that the participant characteristics (level 2) were not significant predictors of any of our dependent variables (and neither were the cross-level interactions). Thus, we opted for the more conservative (because of the fewer degrees of freedom) analytic strategy. In addition, the results of the HLM analyses were not substantively different from those obtained using the more parsimonious OLS and logistic regression techniques. Note. Standard deviations in parentheses. 5We would like to thank David Kenny, an eminent scholar in the field of person perception, for assistance with the development of our subscales. N = 40; *p < .05. Note. Coefficients are unstandardized. For both models, the Wald χ2 test for estimate significance was used and placed in parentheses. For the logit models, we used McFadden's (adjusted) pseudo-R 2. For the OLS models, we used the adjusted R 2. *p < .05. 6We tested to see if "smiling with teeth" had an independent effect on occupational advancement, and found no such effect.
No takes yet. Share an insight, caveat, or question.
Hochschild et al. (2011) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: