Which Statistics Test

1.74K views
0

I’m struggling with statistics and I need someone to for me which test to use when.

Binomial test
One sample t-test
Independent Samples t-test
Paired samples t-test
Cross-tabulation/Chi-square test
Pearson’s r / Regression

In the most basic language possible, for some reason this topic eludes me.

In: Mathematics

Dude you can write half of a book on your question. For me it was mostly experience/exercise to find out which test to use when.
Many of them also work as alternatives to each other, there are e.g. cases where you could use the binomial test but also chi-squared. In this case it often relies on your data, if you got everything needed for chi-squared you will use this one because it‘s quicker/easier.
I‘ll try to explain it really (really) basic:
Binomial test/chi-squared if you have two/more categories for your null hypothesis that are either likely
T-tests if you want to find out significant differences in the means (two tailed if your assumption is both are equal, mono if you assume one is greater, indipendent/non depends on the data)
Pearsons r you want to find out if data is correlated at all
Regression you try to find out how data is connected (regression line is this information if it‘s a linear correlation)