The format in which responses are collected affects the statistical analyses that can be performed (Story & Tait, 2019). Researchers should ensure that statistical assumptions are met for whatever tests they plan to do and should plan investigations a-priori. The type of response data (nominal, ordinal, categorical, or continuous) will affect which tests are appropriate. Creation of dummy variables may be used to analyze ordinal responses, but caution should be applied when using tests that assume a continuous response variable due to the possibility of unequal Likert response spacing (Story & Tait, 2019). 

Common tests include (Story & Tait, 2019): 

• T-test

• ANOVA

• Chi-square (Likert responses might be lumped together to enhance power)

• Ordinal logistic regression approaches (for Likert data; advanced approach)