<p>The survey really needs to be taken with a grain of salt. </p>
<p>For one, unemployed graduates are likely to have more time to fill out a survey and are more likely to have remained at their old address. Therefore, the survey is likely skewed to be more representative of the unemployed graduates.</p>
<p>Number two, although the survey may remain open for up to six months, that does not say anything about when the survey was filled out by any of the students. For example, one student may fill out the survey two months after graduation and then find a job in the third month. It is unlikely he/she would take the time to go update their survey response, although I suppose that is possible. </p>
<p>Finally, the survey’s are voluntary response surveys which make them statistically invalid for drawing factual conclusions.</p>
<p>I think the best we can get from the surveys is a general idea of how graduates of one class are faring compared to another class, and how one major is stacking up against a different major. But even then, the survey results are unlikely to accurately reflect reality because of the flaws I mentioned.</p>