Biology 3401
Analysis of Facial Symmetry and Odour Data
For this lab, we have collected four different data sets:
1) Subjective measures of attractiveness of a group of 100 British students
2) Measurements of facial symmetry from this same group of 100 British students
3) Subjective measures of the attractiveness of the odour of T-shirts
4) Measurement of facial symmetry in the class.
For purposes of this lab, I will be letting you cut some corners on the data analysis because doing it strictly by the book will be cumbersome and time-consuming.
You will notice that the data sheets deal with means and standard deviations of the attractiveness and odour ratings. Strictly speaking, this is incorrect since the mean is a number that can’t show up in the sample. For example, someone’s T-shirt may have an odour rating of 2.45 even though this number was not one of the original rating numbers.
All the data are on the accompanying spreadsheets. There are six sheets in the Excel Workbook
1. Odour & Sex ID (the results of what you thought the sex of the shirt wearer was)
2. Odour Rating (the results of the sniff tests)
3. Male scoring of Brits (how men in our class rated the attractiveness of the Brits)
4. Female scoring of Brits (how women in our class rated the attractiveness of the Brits)
5. Brit Measures (the symmetry measures of the Brits)
6. Class Measures (the symmetry measures taken on yourselves)).
Some Explanatory Notes
1) Odour and Sex ID
This spreadsheet is the results of the question asking you to identify the sex of the wearer of each shirt. The sex of the person doing the rating is in Column A. The sex of the shirt’s wearer is across the top. Each row represents what each rater thought the sex of the wearer of each shirt was. The second last column represents your accuracy at identifying your own shirt (Y = yes and N= no). The last column asks if you were at least correct in identifying the sex of the wearer of the shirt you thought was yours. A ‘Y’ in the second last column would automatically generate a ‘Y’ in the last column
2) Odour Rating
The data in this sheet are the subjective results of the sniff tests rating the shirts on an attractiveness scale of 1 to 10. The numbers in the AK Column represent the mean score given by each rater. The numbers in Row 41 represent the mean score for each shirt. These number can be used to correlate attractiveness of odour with symmetry.
Brits Scoring (Sheets 3 and 4)
The descriptive statistics follow the raw data for each subject. These numbers can be used to correlate with the physical measures on spreadsheet 5.
3) Brit Measurement
These are the measures of symmetry that you did from the pictures of the British students. The measurements needed to be normalized to wipe out the differences due to size of the photograph, overall size of the face etc. To normalize the measures of difference between individuals, I’ve used the following formula for each of the A through I criteria:
Measurement of left side – Measurement of right side
Total of measurements of both sides
I then took the absolute value of each of these calculations and averaged and summed them. These are the numbers that appear for each subject in columns AE and AF respectively. In this calculation, the smaller the number, the more symmetrical the face. A perfectly symmetrical face would have a value of 0 and an absolutely asymmetrical face (one in which all measures on one side were 0) would have a value of 1.
4) Class Measures
The measures of the symmetry of the members of our class are presented in this sheet ini the same way as on the measurement sheet for the British students. The letters identifying each of your are the same on this and the odour preference sheets are the same. This means that the individual identified has ‘AA’ on one sheet is the same person as ‘AA’ on the other. The letters were assigned sequentially to a randomized list.
Analysis of the data
1) The ratings of attractiveness for the Brits and their measurements can be correlated. The equation from this correlation (if it is significant) can be used to generate attractiveness scores for each of the students in the class.
2) The scores for the attractiveness of the class can then be used in a correlation with the odour preference data.
Questions to consider:
1) Is there a relationship between symmetry and perceived attractiveness?
2) Is there a relationship between odour and symmetry or odour and the calculated attractiveness values?
3) Is there a difference in the abilities of males and females to correctly identify the odours of either sex?
4) Are males or females better at identifying their own shirt or a least the correct sex for what they thought was their own shirt?
5) What are some the confounding or extraneous factors that could have altered the results?
6) What are the behavioural and evolutionary implications of what you have found and how does it compare to other studies in the literature?