In: Statistics and Probability
What does it mean when you get an error message running SPSS for binary linear regression that states the dependent variable has more than two non-missing values. For logistic regression, the dependent variable value must assume exactly two values on the cases being processed. What do I need to do to fix this issue?
You need to check the missing values in the dependent variable. Since your dependent variable has two categories you will have to fill the missing values with only those two values but it is not an easy task. I suggest if the number of rows in your data set is large enough that if you remove few values then the logistic model would not get affected then remove the rows that have missing dependent values. Now run logistic regression on this data. If the data set has a low number of rows then the simplest method is creating one more category with NA. Therefore, you will have three categories instead of 2 categories. Now run a multiple logistic regression in SPSS and discard the results for the NA category.
There are some other ways as well but they are more complex because they involve creating predictive models to predict the missing values then they are imputed and used to create the model along with the original data. In this answer, I have explained the simplest method, which is generally used.