Question

Research indicates that the color red increases men’s attraction to women (Elliot & Niesta, 2008). In the original study, men were shown women’s photographs presented on either a white or red background. Photographs presented on red were rated significantly more attractive than the same photographs mounted on white. In a similar study, a researcher prepares a set of 30 women’s photographs, with 15 mounted on a white background and 15 mounted on red. One picture is identified as the test photograph and appears twice in the set, once on white and once on red. Each male participant looks through the entire set of photographs and rates the attractiveness of each woman on a 10-point scale. The following table summarizes the ratings of the test photograph for a sample of nine men. Are the ratings for the test photograph significantly different when it is presented of a red background compared to a white background?

Participant	White Background	Red Background
A	4	7
B	6	7
C	5	8
D	5	9
E	6	9
F	4	7
G	3	9
H	8	9
I	6	9

1. State the null and alternative hypotheses
2. Compute the test statistic using SPSS
3. What is the effect size (you will need to compute both Cohen’s d from data given in the SPSS output table)

Cohen’s d= M1 – M2/ ((SD1+SD2) /2)

4. Write the Results section for your findings. Include the descriptive statistics, type of statistical test and results of the test and the confidence interval.

Answer 1

d) Result of findings:

The mean and standard deviation of red background rating is 8.22 and 0.972 respectively. The mean and standard deviation of white background rating is 5.22 and 1.481 respectively.

The type of test is paired t test. The test statistic of paired t test is 6.000. The p-value of paired t test is 0.000. This p-value is less than 0.05. Hence we reject null hypothesis at 5% level of significance and conclude that, the ratings for the test photograph significantly different when it is presented of a red background compared to a white background.

The value of effect size is 2.446. Hence we can conclude that the ratings for the test photograph significantly large different when it is presented of a red background compared to a white background.

The 95% confidence interval of the difference is (1.847, 4.153). This confidence interval can not include the zero. Hence we reject null hypothesis at 5% level of significance and conclude that, the ratings for the test photograph significantly different when it is presented of a red background compared to a white background.