I am a student (senior) summer hire working for Dept of the Army Resource Management. I've been assisting the Analyst in developing a model to predict training costs. The technique we used was simple linear regression. Now I've been asked for my opinion in choosing the best model I feel is better. I've chosen #1. Garrison Training but I want a second opinion before I give my choice. I've taken basic statistics last year in high school and will probably be taking more this year. Could you look over the information and give me a reason my choice would not be best? The dep variable-y, is Training costs in dolars per day.

Thanks for your opinion.

If the column you call r is the correlation coefficient then your desision is correct. The correlation coefficient measures the strength of the linear fit and hence the model with the largest r has the best linear fit.