Very basic stats question, I have a measure of the % females breeding each year within one mammal population and I want to test between the years. Firstly, what is the best stats for this kind of analysis? Secondly, I'm hoping to then incorporate aditional categories into the test, i.e. Total young born etc. How will this change the test?
My data looks like this:
Yr 1 Yr 2 Yr 3 Yr 4 Yr 5
% breeding 80% 60% 70% 40% 90%
Since parametric ANOVA and discriminant analysis are relatively robust to small deviations from the normal distribution (Dillon, Goldstein, 1984), these deviations are not an obstacle to the application of these tests.
To compare mean values, Student's t-test and two-way ANOVA with Tukey's post-haw test can be used to compare distress within and between species.
To estimate the relative contribution of each parameter to discrimination to estimate the value of Wilks' Lambda. Comparison of the mean values of correct assignment was carried out using a 2x2 χ2 test, comparison of the percentages of correct assignment of different types to specific individuals according to Wilcoxon's criterion for conjugate rows.
Regards, Sergey Viktorovich Pushkin
Dear Anita Marks
First, test the normality of your data distribution.
Then, if your data did not fellow a normality. You can applied non parametric test (for exemple, Kruskal-Wallis test : K independant sampling).
Second, for testing relative contribution of each parameter (Wilcoxon's criterion is good to apply). I agree with Sergey Viktorovich Pushkin
Good luck,
Hi, depending on your number of data points per year, you should use 1 way ANOVA. I agree that you should take a look at the normality of your distribution (I usually go with Shapiro-Wilk unless my n is very big). When you add components to it you can do 2 way anovas followed by Tukey's posy hoc test. I would advise against doing t tests being 2 given years, this is what post-hoc multiple comparisons are for.
Best,
- Badges
- Science topic
- Similar topics
- Biological Science
- Anatomy
- Tissue