links0311
03-11-2009, 09:56 AM
I am doing a report on the effects on re-introduced wolves into Yellowstone. The focus of the study is looking at how climate conditions effect species populations. I ran a t-test and an ANOVA test using the statistical programs provided by the College of Saint Benedict and Saint John's University (http://www.physics.csbsju.edu) but I'm not sure how to determine significance. My t values after the significance test are 2.62, 5.24, 8.51, 8.59, 17.5, and 2.01. After my ANOVA test I got F values of 6.89, 4.056, 28.57, 72, 79, and 304.9. I know the larger the F value the more different the two means are, but can I present this data alone as significant or do I have to perform further tests? What do my t-test values mean? I'm lost.
-Links
antoniac
03-11-2009, 01:13 PM
What you need to do is consult critical values charts. Before you can do that, you need to compute what is called degrees of freedom for each statistic. The degrees of freedom for a t-test is one less than your sample size - so if you have an n of 20, your degree of freedom on a t-test are 19. Check on a t-test critical values chart for your t score but check it at the level of your degrees of freedom. Find a critical values chart for ANOVA (I'm sure you can find these on the internet somewhere) and then the charts will tell you whether or not your stats are significant at the .0001 (not usually), the .01 or the .05. If your significance value is above these numbers, than your t or f values are not significant. Here's a website that briefly discusses degrees of freedom for ANOVA: http://www.physics.csbsju.edu/stats/anova.html
Write back if you have more questions -
Antoniac
links0311
03-11-2009, 10:50 PM
My degrees of freedom for all the trials was 19. Looking at a t value chart under the values of .05 and .01 with a degree of freedom of 19 I get a value of 1.72 and 2.53 respectively. Does that mean my individual t values have to be less than 1.72 and 2.53 to be significant? Thanks for the response.
-Links
links0311
03-12-2009, 08:03 AM
My numbers don't seem to be adding up. When I run a t-test I am seeing significance, but when I run a F test I am getting high probabilities (.98) that my values are not significant. How can I get two confounding results using the same numbers?