Table 2. Classification of the reviewed regression papers.

Rubrics 3 and 5–12 represent papers with imperfect handling of regression assumptions: in rubrics 5–7, it is unclear from whether assumptions are correctly dealt with; in rubrics 8–12, the dealing with assumptions was incorrect.

Class	Reason
Papers without a linear regression model:
1	No Model of Interest
2	Rejection of linear regression on basis of correct assumptions
3	Rejection of linear regression on basis of not meeting incorrect assumptions
Papers with a linear regression model:
4	Correct linear regression
5	Mentioned all correct assumptions but not if the ‘normality assumption’ was tested on the residuals or on X or Y
6	Did not test all but some correct assumptions, included neither normality of variables nor errors
7	Use of linear regression but no indication if any or which assumptions were tested
8	Assumed/tested normally distributed X but not the normality of the errors
9	Assumed/tested normally distributed Y but not the normality of the errors
10	Assumed/tested normally distributed X and Y but not the normality of the errors
11	Assumed/tested normally distributed variables but did not indicate if X or Y or both and did not test the normality of the errors
12	Other misconceptions about assumptions