<p>Re post 17. Yes, the analysis is subject to limitations…all analysis is. I give them credit for reporting the confidence interval and highlighting results that are statistically significant and those that are not. It is not reasonable to dismiss the results just because you don’t happen to like the answer.</p>
<p>Neither is it reasonable to accept them just because you like the answer.</p>
<p>And how did I do that, p2n? I raised the question originally about gaming the results. But I do like the rigor added with confidence interval…that is something that could be used to improve other surveys, e.g, PA scores from USNews could be enhanced with just a little extra statistical analysis.</p>
<p>Yes, you did alert readers to concerns about gaming the system to your credit. I’ve seen all sorts of these things gamed by likely high schools kids. A free analysis is worth what you pay for it. It was wise to post a disclaimer. Have fun.</p>
<p>i like the results, i just think that pretending like this study has merit when its sample size is less than 200 people is bad practice–especially since they were not randomly selected.</p>
<p>the math and purpose are good, but the methodology isn’t.</p>
<p>a4autum…I agree with your point regarding a random sample, but you do realize that confidence interval takes into account sample size.</p>
<p>aforautumn and p2n, that’s what the confidence interval is for. Gives the reader an idea of how likely the numbers are valid.</p>
<p>
</p>
<p>Although … it does make me wonder how having to register would skew the results.</p>