Statistics, Data Science
Which Statistical Test to Use? Follow This Cheat Sheet
Get all the statistical tests clear in 3 minutes!
The Question to be Answered
Before we decide on which test to use, we need to be clear of what we want to solve. Is it:
- To validate whether the population mean is correct
- To compare the difference between 2 groups of data to see whether the difference is statistically significant
- To validate whether there’s a relationship between 2 categorical variables
Null and Alternative Hypothesis
Next, you will be deciding on the hypothesis based on your test objective. A null hypothesis (H0) proposes that no significant difference exists in a set of given observations, and an alternative hypothesis (H1) proposes otherwise.
For rejecting a null hypothesis, a test statistic is calculated. This test-statistic is then compared with a critical value. The critical values are the boundaries of the critical region.
- If the test statistic > critical value, the null hypothesis is rejected.
- If the test statistic ≤ critical value, the null hypothesis is accepted.
P-Value
Other than test statistic, P-value is another important result to look at. P-value is the level of marginal significance within a…