Reference no: EM132255549
QUALITATIVE. These are intentionally open-ended. In "real life" an interviewer would be mostly looking for a thoughtful (rather than a canned) response.
What was the most interesting thing that you found out about your data set? (The answer doesn't have to be about regression per se).
Were you surprised by any conclusions from your analysis?
What was the most challenging part of the project and what did you do to address the challenge?
If you could try one thing differently to improve your data and/or your analysis, what would it be?
QUANTITATIVE. These questions will be scored on "correctness" based on the analysis you provided in Part 2.
Simple Regression:
If you could only use ONE of your independent variables to model your independent variable, which would you use and why?
For any of your simple regression models:
How do you interpret the regression coefficients, b1?
How do you interpret the y-intercept, b0? Does b0 live inside or outside of the experimental region?
How do you interpret the confidence interval?
Did multicollinearity impact your analysis for any of the simple regression models? Why or why not?
Multiple Regression
What does your correlation table imply about potential multi-collinearity in your model? Explain your reasoning.
Do you see any signs of multi-collinearity in your multiple regression analysis? Be able to explain your reasoning either way.
Be able to write your estimated multiple regression equation on the white board, IGNORING variables that the model found to be INSIGNIFICANT. If you don't remember what an estimated regression equation is, please see the lecture slides and/or Chapter 7 of your book.
Be able to answer these questions about any of the regression coefficients:
How do you interpret the y-intercept, b0? Does b0 live inside or outside of the experimental region?
How do you interpret the coefficients b1, b2, b3, ...?
How do you interpret the confidence interval for a given variable?
Do you see any big changes in your estimated coefficients in multiple regression versus simple regression? What, if anything, does this say about multi-collinearity in your multiple regression model?
Summary Questions
Of all of the models you created (simple and multiple), which one would you "trust" the most? Meaning, if you were forced to pick only one model to base decisions on, which one would it be and why?
Would you say that your data is well described by a linear model? Why or why not?