How can we define attractiveness in experiment

One year ago, I had a chance to do a research about attractiveness in Hong Kong. My group needs to find out whether there is a relationship between attractiveness and helping behavior. We set the experiment by chosen a boy and a girl from our group members who are the most attractiveness one, they held many books in their hands and pretended dropped it all, let saw would people helped them picked up the books. In some experiments which related to attractiveness or beautiful about human, maybe some of you have done it before by SONA here. I sometimes confused about how can researchers define attractiveness?

By my experience, my group use scale method to choose who is the most attractiveness. My group mates and I chose a normal face picture and gave like hundred people to mark it. High score means the most attractiveness one and low score for the ugly one. However, I did not know was a good method or not. People have different opinion about attractiveness, sometimes may depends on their facial, body or personality. Some people may still think Angelina Jolie is not beautiful woman although most of us think she is.

Langlois and Roggman had proposed an article called “Attractive face are only average” in 1990. They wanted to systematically examine whether averageness is linked with facial attractiveness. They selected photographs of 192 male and female faces, each of which was computer scanned. Then made computer-processed composites of each image. As 2, 4, 8, 16, 32-face composites, these component faces were rated by other people by scale. Finally they found the result that 32-face composite score higher than other. Moreover, Dr. Stephen introduces a system called Marquardt beauty analysis (MBA). It dedicated to proactively researching human visual aesthetics, including its biological and mathematical bases, and to utilizing the results of that research to develop and provide information and technology with which to analyze and positively modify human visual attractiveness. MBA introduces a mask which they think is “fitting beauty”. If people fit with the mask, then they should be defining as attractiveness.

To conclude, sometimes it difficult to consider what is attractiveness in the experiment. Maybe next time we became a researcher and do a topic like this, we may confuse which ways should we choose. People still have different opinions about attractiveness or beauty. What is your opimion?

Reference: Langlois, J. H. & Roggman, L. A. (1990). “Attractive faces are only average.” Psychological Science, 1, 115-121.

Marquardt beauty mask: http://www.beautyanalysis.com/

Chi-square test goodness-of-fit test

This time I want to talk about chi-square test goodness-of-fit test. It is one of the nonparametric tests. As I talked about it last few weeks, nonparametric tests used for nominal or ordinal data and they are less powerful. There are two well-known tests. One is Goodness-of-fit test and the other one is test of independence.

Chi-square test goodness-of-fit test is a nonparametric inferential procedure that determines how well an observed frequency distribution fits an expected distribution. Expected frequency means frequency expected in category if the sample data represent the population. Observed frequency means which participants fall into a category. Let me take an example, suppose a researcher is interested in determining whether the teenage pregnancy rate at a particular high school is different from the rate statewide, say 17%. If N = 80, expected frequency for pregnancy is 14 (= 80 X 17%) and the opposite should be 66 (= 80 X 83%). If you find only 7 pregnant teenagers and 73 who are not, after calculate by Chi-square formula, we can know that the obtained value is larger, so we reject the null hypothesis and conclude the observed frequency of pregnancy is significantly lower than expected by chance.

What I think about goodness-of-fit test is that it is quite general. Whatever discrete or continuous distribution, it seems to be fit to apply it. Goodness-of-fit indices are often used to supplement chi-square tests of lack of fit in evaluating the acceptability of structural equation and other models. A high goodness-of-fit index may be an encouraging sign that a model is useful even when it fails to fit exactly on statistical grounds (Mulaik, 1988). However, there are some disadvantages that we need to think carefully. This test does not assess all aspects of a model’s appropriateness for data. Hypotheses regarding structural coefficients that are predicted to be nonzero in the population but are estimated as free parameters in the model are not directly assessed by goodness-of-fit indices. One can obtain a high goodness-of-fit index value for a model in which certain structural coefficients hypothesized to be nonzero but treated as free parameters turn out to have estimated values of zero(Giere, 1985).

To conclude, goodness-of-fit tests should only be used conditionally on a significant chi-square for the appropriate null model which if one rejects the hypothesis that all structural coefficients are simultaneously equal to zero and on the significance of tests of individual parameters of special salience to a model.

Reference: 1) Giere, R. N. (1985). Constructive realism. In P. M. Churchland & C. A. Hooker (Eds.), Images of science Chicago: University of Chicago Press

2) Mulaik, S. A., James, L. R., Van Alstine, J., Bennett, N., Lind, S., & Stilwell, C. D. (1989). Evaluation of goodness-of-fit index for structural equation models. Psychological Bulletin, 105(3), 430-445. doi:10.1037/0033-2909.105.3.430