Assignment # 4 PSYC 3000 A

Due Sat. August 10, 2024

Simple and Multiple Regression

This file contains data for 63 participants who responded to four questionnaires measuring Trait
Anxiety, Fear of Negative Evaluation, High Sensitivity, and Social Phobia.

1. Open the Data file.

To open the data file in SPSS, start the program (and click on ‘New Dataset’ in the initial
overlaid window that appears). Click on ‘File’ (located at the very top right-hand side of the
SPSS window), choose ‘Open’ and then ‘Data…’. Browse the computer to find the
SurveyData.sav file and click on it. If the data itself is not visible when this file opens up, you will
need to click on ‘Data View’ (located at the very bottom right-hand side of the SPSS window).

2. Open a syntax file to recode reversed items and make total scores.

To open the recoding syntax file in SPSS (I made this from the scoring schemes available for
each scale), click on ‘File’ (located at the very top right-hand side of the SPSS window), choose
‘Open’ and then ‘Syntax…’. Browse the computer to find the ScalesRecode.sps file and click on
it. To run the syntax, click on ‘Run’ (located on the top menu bar in the Syntax Window), and
click ‘All’. (Note that the SurveyData.sav data file needs to be open in SPSS for this to run on it.)
TIP: Go to the Output Window after running this syntax, then highlight and delete everything in
the output so that it doesn’t get printed out or exported later on!
AS WELL: If you are having problems with this step – contact me (it is needed to make all the
variables used below)!

3. Regress social phobia against trait anxiety.

From the ‘Analyze’ menu, choose the ‘Regression’ option then the ‘Linear...’ option. Click on
“SPTot” and move it into the ‘Dependent’ box by clicking the arrow. Click on “TATot” and move
it into the ‘Independent(s)’ box by clicking the arrow. Next, click on ‘Statistics’, then click in the
boxes for ‘Confidence intervals’, and ‘Casewise Diagnostics’ (but do not unclick the ‘Estimates’
and ‘Model Fit’ boxes). Click on ‘Continue’. Click on ‘OK’.
 4. Make a scatterplot showing the relation between social phobia against trait anxiety.

Next, from the ‘Graphs’ menu, choose the ‘Scatter/Dot...’ option, and the ‘Simple Scatter’ option (which
should already be selected) and click on ‘Define’. Click on “ SPTot” and move it into the ‘Y-axis:’ box by
clicking the arrow. Click on “TATot” and move it into the ‘X-axis:’ box by clicking the arrow. Click on ‘OK’.
When the scatter plot appears, double click on the plot itself. Then, in the ‘Chart Editor’, find and click
on the ‘Add Fit Line at Total’ icon. Then close the Chart Editor. To deselect the plot, click anywhere
outside of it.

5. Regress social phobia against all three predictors together.

From the ‘Analyze’ menu, choose the ‘Regression’ option then the ‘Linear...’ option. Click on
“SPTot” and move it into the ‘Dependent’ box by clicking the arrow. Click on “TATot” and move
it into the ‘Independent(s)’ box by clicking the arrow (although both variables might already be
in these boxes from the previous analyses). In the same fashion, move “NETot” and “HSTot”
into the ‘Independent(s)’ box. Next, click on ‘Statistics’, then click in the boxes for ‘Confidence
intervals’, and ‘Casewise Diagnostics’ (but do not unclick the ‘Estimates’ and ‘Model Fit’ boxes).
Click on ‘Continue’. Finally, click on ‘Plots’, then under ‘Standardized Residual Plots” click in the
box for “Histogram”. Click on ‘OK’.

6. Get a correlation matrix with all of the bivariate correlations between the four scales.

From the ‘Analyze’ menu, choose the ‘Correlate’ option then the ‘Bivariate...’ option. Click on “SPTot”,
“TATot”, “NETot”, and “HSTot” (in that order) and move them into the ‘Variables’ box. Click on
‘OK’.

Use the output to answer the questions below. To save it as a .pdf file, while in the Output
screen itself, go to the ‘File’ menu and select ‘Export’, then ‘Browse’ to indicate where to save it
and give that .pdf file a name).

Questions:
(i) Which of the four measures have items that need to be reverse scored (refer to the
syntax file – copied on the back page of this assignment - to figure this out).

(ii) For the initial bivariate simple regression using “Trait Anxiety” only as the predictor
of “Social Phobia”, report whether the regression results (using α = .05 as the
level of significance) involving the slope coefficient were significant or not (i.e.,
please report these results like this [with all the blanks filled in with the appropriate values
or variable names]: For the regression of __y___ on __x__, the slope coefficient was
significant/not significant, b = __, t(_df_) = ___, p = ___, 95% C.I. (___, ___)).
Interpret the value of the slope coefficient obtained from this analysis (i.e., what
exactly is it telling us in words by referring to the actual variable names?).
(iii) For the multiple regression, report the regression results (using α = .05 as the level of
significance) for each individual slope coefficient (i.e., again, please report these
results like this [with all the blanks filled in with the appropriate values or variable names] :
For the regression of __y___ on all three predictors, the slope coefficient for
_____ was significant/not significant, b = __, t(_df_) = ___, p = ___, 95% C.I. (___,
___), etc.). Discuss the extent to which the slope value and inferential test results
for the Trait Anxiety variable in the multiple regression are consistent (or not)
with the previous corresponding bivariate regression results for this same
variable (i.e., in [ii]).

(iv) Report the multiple R2 for this multiple regression and report the amount that it
changed from the previous corresponding bivariate r 2 for the regression of Social
Phobia on Trait Anxiety alone. This change represents the further relative
reduction in prediction error due to adding which variables to the regression
model?

(v) Does the histogram of standardized residuals from the multiple regression seem to
indicate that normality is present? Does there seem to be any potential
regression outliers (and indicate why or why not)?

(vi) Is the sample size adequate given the number of variables in the multiple regression
analysis (and indicate why or why not)?
Copy of syntax file:

AUTORECODE VARIABLES=TAX1 TAX3 TAX6 TAX7 TAX10 TAX13 TAX14 TAX16 TAX19
/INTO TAX1_RV TAX3_RV TAX6_RV TAX7_RV TAX10_RV TAX13_RV TAX14_RV TAX16_RV TAX19_RV
/DESCENDING
/PRINT.

AUTORECODE VARIABLES=NE2 NE4 NE7 NE10

*Total Trait Anxiety*

COMPUTE TATot = TAX1_RV + TAX2 + TAX3_RV + TAX4 + TAX5 + TAX6_RV + TAX7_RV + TAX8 + TAX9 + TAX10_RV +
TAX11 + TAX12 + TAX13_RV + TAX14_RV + TAX15 + TAX16_RV + TAX17 + TAX18 + TAX19_RV + TAX20.
EXECUTE.

*Total Fear of Negative Evaluation*

COMPUTE NETot = NE1 + NE2_RV + NE3 + NE4_RV + NE5 + NE6 + NE7_RV + NE8 + NE9 + NE10_RV + NE11 +NE12.
EXECUTE.

*Total High Sensitivity*

COMPUTE HSTot = HS1 + HS2 + HS3 + HS4 + HS5 + HS6 + HS7 + HS8 + HS9 + HS10 + HS11 +HS12 + HS13 + HS14 +
HS15 + HS16 + HS17 + HS18 + HS19 + HS20 + HS21 + HS22 + HS23 + HS24 + HS25 + HS26 + HS27.
EXECUTE.

*Total Social Phobia*

COMPUTE SPTot = SP1 + SP2 + SP3 + SP4 + SP5 + SP6 + SP7 + SP8 + SP9 + SP10 + SP11 +SP12 + SP13 + SP14 + SP15 +
SP16 + SP17 + SP18 + SP19 + SP20.
EXECUTE.