Exam #2 Laboratory #2 Correlations and Regression on SPSS



Psyc451 Exam #1 Lab #2 Name:_________________________

Bivariate & Multivariate Regression Dataset ( pack1mod.sav

Walk-Through -- The criterion variable will be graduate GPA & the predictors will be prog, degree, GREA, GREQ, and GREV

Obtain the bivariate model using each predictor, in turn.

a. Fill in the following

|Predictor |b | β |p |a |Does the predictor “work”?|Interpretation of b weight (be sure to use the proper wording for |

| | | | | | |quantitative vs. binary predictors) |

|Prog | | | | | | |

|GREA | | | | | | |

|GREQ | | | | | | |

|GREV | | | | | | |

|Degree | | | | | | |

Get the correlation of each of the predictors with the criterion and the multiple regression using all the predictors

b. Fill in the following

R² = _________ F = ___________ df = ____, ___________ p = __________

| |correlations |multiple regression |

|Predictor | | |

| |r |p |Is this a viable bivariate |b |β |p |Does the predictor contribute to the |

| | | |predictor? | | | |multivariate model? |

|GREA | | | | | | | |

|GREQ | | | | | | | |

|GREV | | | | | | | |

|Degree | | | | | | | |

|Constant (a) | | | | | | | |

c. Does the multiple regression model work? How well?

What variables contribute to the model? What contributors are “most important”?

d. Interpret each multiple regression b weight (be sure to use the proper wording for quantitative vs. binary predictors). Also, for each predictor, compare its bivariate relationship with the criterion (from the simple regression) and its contribution to the multivariate model (from the multiple regression weight). Identify each predictor as one of the following:

1. Predictor is neither correlated nor makes a multivariate contribution

2. Correlation and multivariate contribution have the same (non-zero) sign

3. Predictor is correlated, but does not make a multivariate contribution (probably is collinear with other predictors)

4. Suppressor that is not correlated, but makes a multivariate contribution

5. Suppressor that is correlated and makes a multivariate contribution, but with opposite (non-zero) signs

|Variable |Interpretation |“Type of Predictor” |

|Prog | | |

| | | |

|GREA | | |

| | | |

|GREQ | | |

| | | |

|GREV | | |

| | | |

|Degree | | |

| | | |

|Constant (a) | | |

| | | |

e. Write it up, following the example in the handout (use a single table – don’t worry about the means and std).

Your Turn #1 Dataset ( pack2mod.sav

The criterion variable will be depression & the predictors will be stress, SES, salary satisfaction, financial independence, and marital status.

Obtain the bivariate model using each predictor, in turn.

a. Fill in the following

|Predictor |b | β |p |a |Does the predictor “work”?|Interpretation of b weight |

|stress | | | | | | |

|SES | | | | | | |

|salary satisfaction | | | | | | |

|financial | | | | | | |

|independence | | | | | | |

|marital status | | | | | | |

Get the correlation of each of the predictors with the criterion and the multiple regression using all the predictors

b. Fill in the following

R² = _________ F = ___________ df = ____, ___________ p = __________

| |correlations |multiple regression |

|Predictor | | |

| |r |p |Is this a viable bivariate |b |β |p |Does the predictor contribute to the |

| | | |predictor? | | | |multivariate model? |

|SES | | | | | | | |

|salary | | | | | | | |

|satisfaction | | | | | | | |

|financial | | | | | | | |

|independence | | | | | | | |

|marital | | | | | | | |

|status | | | | | | | |

|Constant (a) | | | | | | | |

c. Does the multiple regression model work? How well?

What variables contribute to the model? What contributors are “most important”?

d. Interpret each multiple regression b weight (be sure to use the proper wording for quantitative vs. binary predictors). Also, for each predictor, compare its bivariate relationship with the criterion (from the simple regression) and its contribution to the multivariate model (from the multiple regression weight). Identify each predictor as one of the following:

1. Predictor is neither correlated nor makes a multivariate contribution

2. Correlation and multivariate contribution have the same (non-zero) sign

3. Predictor is correlated, but does not make a multivariate contribution (probably is collinear with other predictors)

4. Suppressor that is not correlated, but makes a multivariate contribution

5. Suppressor that is correlated and makes a multivariate contribution, but with opposite (non-zero) signs

|Variable |Interpretation |“Type of Predictor” |

|stress | | |

| | | |

|SES | | |

| | | |

|salary | | |

|satisfaction | | |

|financial | | |

|independence | | |

|marital status | | |

| | | |

|Constant (a) | | |

| | | |

e. Write it up, following the example in the handout (use a single table – don’t worry about the means and std).

About the “Your Turn,” your Laboratory Project and the Conference

This is the second week to work on your project. If you do a decent job this week of finding a criterion and some predictors, the rest of the project will go very smoothly (on the other hand, if you dog this assignment you’ll have tons to do to catch up).

You may use the same criterion/predictors you used last week, change some, or change all of them -- your choice. Remember, a “good story” isn’t just “a set of all significant” predictors. Rather, it is a combination of things that do and don’t correlate/have multivariate contributions, so we can tell an interesting story about how these variables relate to the criterion.

So, use this assignment to find a criterion and set of predictors

▪ Have a mix of “kinds of predictors” (demographics, etc.)

▪ Have a mix of significant and nonsignificant multivariate contributors (6-8 predictors is plenty!!!)

You may try multiple criterion variables in any data set, but you must present below data from at least 2 different data sets!

When you are done with the four Your Turns, pick 2 of them to write up like the Walk Through above and the example on the handouts.

Your Turn #2 Note: criterion must be quantitative – predictors must be quantitative or binary

Data set ________________________ criterion variable ________________________________

Obtain the bivariate model using each predictor, in turn.

a. Fill in the following

|Predictor |b | β |p |a |Does the predictor “work”?|Interpretation of b weight |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

Get the correlation of each of the predictors with the criterion and the multiple regression using all the predictors

b. Fill in the following

R² = _________ F = ___________ df = ____, ___________ p = __________

| |correlations |multiple regression |

|Predictor | | |

| |r |p |Is this a viable bivariate |b |β |p |Does the predictor contribute to the |

| | | |predictor? | | | |multivariate model? |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

|Constant (a) | | | | | | | |

c. Does the multiple regression model work? How well?

What variables contribute to the model? What contributors are “most important”?

d. Interpret each multiple regression b weight (be sure to use the proper wording for quantitative vs. binary predictors). Also, for each predictor, compare its bivariate relationship with the criterion (from the simple regression) and its contribution to the multivariate model (from the multiple regression weight). Identify each predictor as one of the following:

1. Predictor is neither correlated nor makes a multivariate contribution

2. Correlation and multivariate contribution have the same (non-zero) sign

3. Predictor is correlated, but does not make a multivariate contribution (probably is collinear with other predictors)

4. Suppressor that is not correlated, but makes a multivariate contribution

5. Suppressor that is correlated and makes a multivariate contribution, but with opposite (non-zero) signs

|Variable |Interpretation |“Type of Predictor” |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

|Constant (a) | | |

Your Turn #3 Note: criterion must be quantitative – predictors must be quantitative or binary

Data set ________________________ criterion variable ________________________________

Obtain the bivariate model using each predictor, in turn.

a. Fill in the following

|Predictor |b | β |p |a |Does the predictor “work”?|Interpretation of b weight |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

Get the correlation of each of the predictors with the criterion and the multiple regression using all the predictors

b. Fill in the following

R² = _________ F = ___________ df = ____, ___________ p = __________

| |correlations |multiple regression |

|Predictor | | |

| |r |p |Is this a viable bivariate |b |β |p |Does the predictor contribute to the |

| | | |predictor? | | | |multivariate model? |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

|Constant (a) | | | | | | | |

c. Does the multiple regression model work? How well?

What variables contribute to the model? What contributors are “most important”?

d. Interpret each multiple regression b weight (be sure to use the proper wording for quantitative vs. binary predictors). Also, for each predictor, compare its bivariate relationship with the criterion (from the simple regression) and its contribution to the multivariate model (from the multiple regression weight). Identify each predictor as one of the following:

1. Predictor is neither correlated nor makes a multivariate contribution

6. Correlation and multivariate contribution have the same (non-zero) sign

7. Predictor is correlated, but does not make a multivariate contribution (probably is collinear with other predictors)

8. Suppressor that is not correlated, but makes a multivariate contribution

9. Suppressor that is correlated and makes a multivariate contribution, but with opposite (non-zero) signs

|Variable |Interpretation |“Type of Predictor” |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

|Constant (a) | | |

Your Turn #4 Note: criterion must be quantitative – predictors must be quantitative or binary

Data set ________________________ criterion variable ________________________________

Obtain the bivariate model using each predictor, in turn.

a. Fill in the following

|Predictor |b | β |p |a |Does the predictor “work”?|Interpretation of b weight |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

Get the correlation of each of the predictors with the criterion and the multiple regression using all the predictors

b. Fill in the following

R² = _________ F = ___________ df = ____, ___________ p = __________

| |correlations |multiple regression |

|Predictor | | |

| |r |p |Is this a viable bivariate |b |β |p |Does the predictor contribute to the |

| | | |predictor? | | | |multivariate model? |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

|Constant (a) | | | | | | | |

c. Does the multiple regression model work? How well?

What variables contribute to the model? What contributors are “most important”?

d. Interpret each multiple regression b weight (be sure to use the proper wording for quantitative vs. binary predictors). Also, for each predictor, compare its bivariate relationship with the criterion (from the simple regression) and its contribution to the multivariate model (from the multiple regression weight). Identify each predictor as one of the following:

1. Predictor is neither correlated nor makes a multivariate contribution

10. Correlation and multivariate contribution have the same (non-zero) sign

11. Predictor is correlated, but does not make a multivariate contribution (probably is collinear with other predictors)

12. Suppressor that is not correlated, but makes a multivariate contribution

13. Suppressor that is correlated and makes a multivariate contribution, but with opposite (non-zero) signs

|Variable |Interpretation |“Type of Predictor” |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

|Constant (a) | | |

Your Turn #5 Note: criterion must be quantitative – predictors must be quantitative or binary

Data set ________________________ criterion variable ________________________________

Obtain the bivariate model using each predictor, in turn.

a. Fill in the following

|Predictor |b | β |p |a |Does the predictor “work”?|Interpretation of b weight |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

| | | | | | | |

Get the correlation of each of the predictors with the criterion and the multiple regression using all the predictors

b. Fill in the following

R² = _________ F = ___________ df = ____, ___________ p = __________

| |correlations |multiple regression |

|Predictor | | |

| |r |p |Is this a viable bivariate |b |β |p |Does the predictor contribute to the |

| | | |predictor? | | | |multivariate model? |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

| | | | | | | | |

|Constant (a) | | | | | | | |

c. Does the multiple regression model work? How well?

What variables contribute to the model? What contributors are “most important”?

d. Interpret each multiple regression b weight (be sure to use the proper wording for quantitative vs. binary predictors). Also, for each predictor, compare its bivariate relationship with the criterion (from the simple regression) and its contribution to the multivariate model (from the multiple regression weight). Identify each predictor as one of the following:

1. Predictor is neither correlated nor makes a multivariate contribution

14. Correlation and multivariate contribution have the same (non-zero) sign

15. Predictor is correlated, but does not make a multivariate contribution (probably is collinear with other predictors)

16. Suppressor that is not correlated, but makes a multivariate contribution

17. Suppressor that is correlated and makes a multivariate contribution, but with opposite (non-zero) signs

|Variable |Interpretation |“Type of Predictor” |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

| | | |

|Constant (a) | | |

When you are done with the four Your Turns, pick 2 of them to write up like the Walk Through above and the example on the handouts.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download