Question

In SPSS, I am reviewing the test of normality. I am trying to see what is the relationship between age and exercise time. Ages range from 20 - 24 with 100 participants. I know I need to test for the dependent variability normality first, with the age being independent and the exercise time being dependent. I know that greater than 0.05 is normally distributed and less than 0.05 is not normally distributed. But lets say four of my age groups of my Shapiro-Wilk Sig is above 0.05. They are normally distributed, but what if one of them is less than 0.04, what does that mean? Is it still normally distributed since the majority is above 0.05? How would that impact how to proceed?

Answer 1

For four of age groups, the Shapiro-Wilk Sig is above 0.05 that means those groups are normally distributed.

If one of them is less than 0.04 that means that the group is not normally distributed. If we combine this group with other groups then the whole group including this will be normal.

You should not make groups of ages to proceed, use age as a single variable and exercise time as another. Grouping is not required to solve this problem if you have data in the form of groups then break down the groups in a single variable form.

Note: you don’t need to make groups for this data to test for normality or relationship testing. Please share the data set, it will be more helpful to us to explain how you should proceed.