Ref: "How do we know whether we use t-test, ANOVA, chi-square, correlation, or regression analysis?" on Quora
-
T-test: Compares the means of a variable from two groups. An independent t-test compares the means of two different groups (e.g., reaction times on a task for women vs. men), whereas a paired t-test compares the means of the same group at two different times (e.g., reaction times for the same people on a task before or after a training period).
-
ANOVA: Compares the means of 3 or more groups. Repeated-measures ANOVA is used when it is the same group of people in each group, as in a paired t-test.
-
Chi-square: Comparing nominal data (e.g, # of correct responses).
-
Correlation: Measuring the strength of a relationship between two continuous variables (e.g., height and age).
-
Regression: Similar to correlation, but used to define the relationship between two continuous variables such that you can then make predictions. For example, after finding the equation that best describes the relationship between height and age in your sample, can then predict the height of someone of an age not in your sample.