Help me for stata and interpetation G) AND H) I put some values :) g) Use the co

Question

Help me for stata and interpetation G) AND H) I put some values :)

g) Use the corr command to find the correlation between tuit17 and tuit18. Explain why using the average tuition over two years might be preferred to adding both tuition variables to the model as separate explanatory variables. Create a variable for average tuition and re-estimate the model using this variable (but exclude tuit 17 and tuit18). What happens to the statistical significance of tuition when you use the average over two years h) Do the results for the average tuition variable make sense when interpreted causally? What might be going on?

Explanation / Answer

g) There are 3 results

1. In the regression equation, the tuit17 and tuit18 p-values are much high (0.52 and0.78), which means they are not significant. By f-test it is showing they are not equal

2. The correlation says they are correlated very closely. So it is advisable to drop one variable and have only one of them.

3. But you also have an interesting rule that the p-value of F-test for testing tuit17 = tuit18 has yielded 0.38.

It is greater than 0.05 which means you cannot reject H0 which states the means are equal (tuit17 = tuit18)

So you are confused which one to drop (tuit17 or tuit18), as both are not significant and both are correlated. In that case it is better to average out the two and rerun it.

h) You have not given the data to average out the tuit17 and tuit18 to rerun the model. What you can do is create an averageTuit column and rerun the regression model.

In the regression results, check for P>|t| for the variable averageTuit. That is the p-value and signifies if this averageTuit variable is making any effect on your dependant variable

Related Questions

Navigate

• Browse (All)
• Browse H
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.