English Deutsch Français Italiano Español Português 繁體中文 Bahasa Indonesia Tiếng Việt ภาษาไทย
All categories

3 answers

I would avoid z-test. You need a whole population to run a z-test. If you're just going by sample means, you should use ANOVA (analysis of variance); t-test is only for two samples, but ANOVA can test for difference between several means. you may need to make appropriate adjustments (Bonferroni et al), of course.

2006-12-01 08:16:39 · answer #1 · answered by Gumdrop Girl 7 · 0 0

If by this you mean you're comparing the mean of one sample with the mean of another sample, you'd most likely use a two-sample t-test. If both samples are large, you could use a two-sample z-test (which works essentially the same way).

2006-11-30 14:06:31 · answer #2 · answered by dmb 5 · 0 0

This appears like a project for a t-try, the place you have one exhibit variable (presence of inhibitor) and one non-provide up variable (activity point). there are particular situations for t-assessments, e.g. that the information be distributed extra-or-much less many times, regardless of the undeniable fact that it probable would desire to be wonderful on your project. (you additionally can use ANOVA, which as a result's precisely comparable to a t-try.) §

2016-10-13 10:16:21 · answer #3 · answered by ? 4 · 0 0

fedest.com, questions and answers