Results | research

Results

After all the participants completed the social desirability scale and the surveys, I used a one-way between subjects analysis of variance (ANOVA) to correlate social desirability and details changed. The ANOVA is used when the independent variable has two or more conditions. In my case the conditions were the high and low social desirability groups. The ANOVA allowed me to compare these two groups in respects to the dependent variable, the number of responses changed from survey one to survey two. In order to have significant data, the F-ratio needs to be larger than one and significance (p) of less than .05. I found an F-ratio of .065 and p>.05 meaning there was no significance.

I found that:

There was a change in reported details from survey one to survey two as expected
There was no significant correlation between the number of responses changed and the score on the Paulhus Social Desirability Measure
Although it was not included in my hypotheses, there was a significant negative correlation between confidence levels on the first survey and the number of responses changed. This means that people who were less confident on the first survey changed more of their reported details to match their partner's

This graph supports my first hypotheses. My first hypotheses simply stated that there would be a change in reported details from survey one to survey two. The bars on the graph represent the scores on the surveys. The scores were visibly higher on the second survey indicating that the majority of the participants changed at least one detail. Of all 32 participants three people remained consistent from the first survey to the second.

This graph speaks to my second hypothesis. My second hypothesis stated that participants who scored higher on the Paulhus Social Desirability Scale would change more of their reported details on survey two. As indicated in the graph, this was not supported. There was no significant correlation between the social desirability and details changed, people who scored lower on the scale were just as likely to change answers as those who score higher.

Lower Confidence Levels and Responses Changed

This graph does not support either of my hypotheses but it shows the significance I found with confidence. It explains that people who had a lower confidence levels on the first survey changed significantly more details when they took then second survey. It also shows that people with lower confidence levels on the first survey were significantly less confident on the second survey as well.