<p>hey.
what'd you guys write for inventiveness? any advice for thinking of something to write?</p>
<p>Also...</p>
<p>how in the world do i see if my data is statisically significant? i did a simple correlation, and i want to know if the coefficient of correlation is significant.</p>
<p>thanks</p>
<p>(and yes, i'm aware this is due soon)</p>
<p>for inventiveness i wrote about my previous research. Perhaps you could write about previous projects you've done etc. if your R^2 value is either close to 1 or -1, then there's definitely some correlation. Also, you can try hypothesis testing...t tests, z tests, etc to determine if your data is statistically significant.</p>
<p>hahah, i just asked a statistics teacher and she said the same thing as you</p>
<p>but what exactly is r? is it the coefficient of correlation or something else? t test i'll have to look up...</p>
<p>and past research.. great idea ! too bad i don't really remember the results of my last research project (it was pretty lame.. even though it was a continuation of one that won intel)</p>
<p>alrightie.. how do you do a t test?...</p>
<p>you said you could compare your data with your hypothesis? how?</p>
<p>sorry, its been a while since i've done a full blown t-test. i know you can do it on excel. wikipedia should help you understand a bit more.</p>
<p>ehh.. honestly I don't even want to bother with it.</p>
<p>Do you think it's safe to say that a coefficient of correlation of about .2 is not significant?</p>
<p>make sure you've fitted the data with the correct curve though...log fit, polynomial fit, etc</p>
<p>it doesn't fit ANY curve. it's just a bunch of random dots :[</p>
<p>ahhhh well~ thank you so much for the helpfuladvice</p>
<p><a href="http://www.graphpad.com%5B/url%5D">http://www.graphpad.com</a></p>
<p>can get a 30 day free trial, work your numbers see if you can get anything out of it.</p>