Using Genomic Ancestry and Demographic Variables to Study Perception in Human Faces
Open Access
- Author:
- Rogers, Megan Alicia
- Graduate Program:
- Genetics
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- None
- Committee Members:
- Mark Shriver, Thesis Advisor/Co-Advisor
- Keywords:
- Genomic Ancestry
perception
faces
AIMs- Ancestry Informative Markers; WA- West African; ORE- Other - Abstract:
- There is substantial evidence supporting the contention that both interindividual and intergroup variability affect our perceptions of human faces. Indeed, there is substantial variability in how persons perceive faces. Like other primates, humans are highly visual and rely on sight much more than they rely on other senses. Humans show a vast capacity for communicating via facial expressions and an incredible facility for distinguishing and remembering faces. It is also clear that our facial perceptive abilities are environmentally modified and need to be explored because we can then understand the roles of facial phenotypes and perception in in-group biases like colorism or racism. Also, by applying this research, we can promote an acceptance and appreciation of phenotypic diversity. I investigated specifically how people rate genomic ancestry of populations admixed with European and West African ancestry, and the relationship between a person’s demographic background (e.g. education level, socioeconomic status, geographical location, racial/ethnic breakdown of family and friends, etc.), and his or her ability to estimate genomic ancestry in human faces. Additionally, I investigated whether sex differences were present among how the stimulus faces were rated, as well. I first collected demographic survey information from each participant. Then, I presented two sets of 42 photos and asked them to estimate the proportion of African ancestry in each face for one set, and categorize each face as “Black” or “White” in the other set. There were 100 participants collected in Georgia and Pennsylvania. My analyses determined that, raters are better than randomly simulated ancestry estimates (p=0.0001). Also, female stimulus faces were rated significantly (p=1.6e-11) better than male stimulus faces. To test whether demographic factors determined one’s ability to rate genomic ancestry, I ran a multiple regression, and individual t-tests or ANOVAs. The multiple regression via factor analysis and principle components resulted in none of the demographic factors being significant. For the t- tests and ANOVAs, I applied the Bonferroni correction and only one factor was significant: the participants’ self described ability to recognize Black faces (p=0.0006). Three factors did not exceed the significance threshold of 0.00263. These trends, (geographical location before the age of ten (p=0.041), participants’ current neighborhood (p=0.031), and the proportion of European Americans in their high school (p=0.011)), and possible explanations for them will be discussed.