Question

Pretend you are conducting an experiment on happiness. For all participants, you measure 4 things

1. how many hours they sleep a day,
2. cortisol levels in their saliva,
3. the number of hours of TV they watch, and finally
4. how many minutes of exercise they get every week.

As all your variables are continuous, you decide to analyze this data using multiple linear regression. As a result, you get the following results.

Overall model: R2=.45, F(3, 145)=14.46, p<.05

Beta values:

hours of sleep: β=.26, p<.05

Cortisol levels: β=.04, p>.05

# hours of TV: β=.13, p>.05

Min. of exercise: β=.33, p<.05

Based on these results, were you able to significantly predict happiness using these 4 variables?

A. yes

B. unsure

C. no

Which one is the answer? why?

Answer 1

First of all, the p-value for F-test is less than 0.05 So it clearly indicates that our model is significant and it is able to model the behavior.

After that those variables which have a p-value of less than 0.05 will be included in the model and those having a p-value greater than 0.05 will not be included in the model.

So

• hours of sleep
• Min. of exercise

will be included in the model.

So based on these variables included in the model, we can significantly predict happiness.

Answer (A). yes

​Please give thumbs up if you like the answer.