# Regression analysis in SPSS

Question 1

An experiment involves three treatment groups; group A, group B and a control group

(i.e. no treatment). Three hundred subjects are randomly assigned to the groups creating

three groups of one hundred.

a. If the experimenter wants the regression slopes to compare each treatment mean

to the control group mean, how should the codes be set up? Are they orthogonal?

b. If the experimenter wants to perform the following comparisons, how should the

codes be set up? Are these orthogonal?

1. Treatment vs No Treatment

2. Treatment A vs Treatment B

Question 2

The school board of a particular town wants to compare the three high schools in the

town. A multiple regression analysis is conducted using the contrasts shown below. Use

the means table also shown to find the prediction equation: ) 2 c ( b ) 1 c ( b b Yˆ i = 0 + 1 + 2 , i.e.

find the values of b0, b1, and b2. Describe (in a sentence each) the meaning of the

coefficient of c1 and c2.

c1 c2 Mean SAT N

School 1 1 1 School 1 992 100

School 2 -2 0 School 2 988 100

School 3 1 -1 School 3 986 100

Question 3

A researcher is interested to know if playing a musical instrument has a significant effect

on grade point average (gpa). She records whether or not a student plays an instrument

and their gender.

Using the data, run the appropriate analysis and interpret the results. Include the group

means and a plot of these (i.e. a graph of the interaction or non-interaction) in your

interpretation as well as a separate effects ANOVA table.

Question 4

A company is interested in ways of improving employee production. The management

decided to attempt to increase company pride and moral by providing one day a month

that would be different from a normal work day. They could not decide if this day should

include some sort of work related activities (such as training or other workshops, group

A) or if the day should be purely social (trips to the movies or a museum, group B). The

research department randomly assigned 15 people to each group. They also recorded

each worker's salary. Using the data, perform the appropriate analysis in SPSS.

Summarize the results.

https://brainmass.com/statistics/regression-analysis/regression-analysis-spss-192041

#### Solution Preview

Please see the attachments.

Please use SPSS 16 to open the SPSS files.

Question 1

An experiment involves three treatment groups; group A, group B and a control group

(i.e. no treatment). Three hundred subjects are randomly assigned to the groups creating

three groups of one hundred.

Three dummy variables are created as fellows

Control =1 if group = control group

=0 else

GroupA =1 if group = A

=0 else

GroupB =1 if group = B

=0 else

a. If the experimenter wants the regression slopes to compare each treatment mean to the control group mean, how should the codes be set up? Are they orthogonal?

In this case the dummy variables for groups A and Group B are included in the regression model. The regression coefficients of the dummy variable will give the effect of group A and B compared to the control variable.

The dummy variables included in the model are orthogonal as the sum of cross product of these variables are equal to ...

#### Solution Summary

Step by step method for regression analysis in SPSS is discussed here. Regression coefficients, coefficient of determination, scatter diagram and significance of regression model are explained in the solution.