Hi, I ran best subsets using one dependent variable (continuous), an independent (continuous) variable, and seven independent (dummy-coded) variables. Only the independent (continuous) variable was shown in the results with significant predictor importance. No other variables (all the nominal-dummy coded ones) were included in the tables or graphs. Does this mean that it did not run correctly? If so, were there any other ways I should have entered the variables in the automatic linear modeling dialog box?

Greetings!

Dear @Rachael Ming

When troubleshooting issues with Best Subsets Regression, ensure that you've correctly specified the response and predictor variables. Verify that the dataset is clean, without missing values or outliers affecting the analysis. Review the algorithm's assumptions, such as linear relationships and homoscedasticity. Consider regularization techniques like LASSO or Ridge regression if multicolumn rarity is present. Experiment with different subsets of predictors and assess model performance using cross-validation. Lastly, consult relevant statistical resources or seek guidance from experts to address any specific challenges you encounter.