SPSS TENTAMEN uitwerkingen Technieken voor causale analyse (MTO-C) 2024

Rubi Luneta
2081276
Tilburg University



SPSS Test Solution Videos
Question 1 — q1 — One-Way Between-Subjects ANOVA - 64315.7.1

Run a One-Way Between-Subjects ANOVA to investigate whether countries di er with respect to
the average score on health. In your analysis, also request the following statistics: descriptive
statistics, test for homogeneity of variances, and the Brown-Forsythe, and the Welch test.

Which country has the highest average score on health?

A Finland
B France
C Germany
D Greece

Analyze > Compare Means and Proportions > One-Way ANOVA > Dependent list (Y), Factor (X) >
Options > Statistics ✓ Descriptive, Homogeneity of variances test, Brown-Forsythe test, Welch
test

ONEWAY health BY cntry
/ES=OVERALL
/STATISTICS DESCRIPTIVES HOMOGENEITY BROWNFORSYTHE WELCH
/MISSING ANALYSIS
/CRITERIA=CILEVEL(0.95).




Question 2 — q2 — One-Way Between-Subjects ANOVA — Levene and Welch test statistic
— 64321.3.6

From the ANOVA output, ll-in the following test statistics:

- the test statistic for the Levene’s test (based on mean)
- the test statistic for the Welch test
test statistic for the Levene’s test (based on the mean): 21.793

test statistic for the Welch test: 94.700




1


fi ff

, Question 3 — q3 — Pearson’s r — 64325.6.1

Estimate Pearson’s r between Health and FeelingEnergetic.

Fill in the value of Pearson’s r as reported by SPSS.

0.416

Analyze > Correlate > Bivariate > Variables: health and FeelingEnergetic

CORRELATIONS
/VARIABLES=health FeelingEnergetic
/PRINT=TWOTAIL NOSIG FULL
/MISSING=PAIRWISE.




Question 4 — q4 — partial correlation coe cient — 64328.6.0

Now estimate the partial correlation between Health and FeelingEnergetic, holding constant for
gender di erences (Female).

Fill in the value of the partial correlation coe cient as reported by SPSS.

0.413

Analyze > Correlate > Partial > Variables: health and FeelingEnergetic > Controlling for: female

PARTIAL CORR
/VARIABLES=health FeelingEnergetic BY female
/SIGNIFICANCE=TWOTAIL
/MISSING=LISTWISE.




Question 5 — q5 — regression with centered variable — 64329.2.9

Compute a new variable named “FeelingEnergetic_cent” which shows the mean-centered scores
of the variable FeelingEnergetic. In your computation of the centered variable, use the mean of
FeelingEnergetic rounded to four digits.
2


ff ffiffi

, Then, regress the variable Health (= dependent variable) on the variable FeelingEnergetic_cent (=
independent variable) and report the estimate of the intercept of the regression equation as
reported in the SPSS output.

Value of constant: 8.455

Analyze > Descriptive Statistics > Descriptives > Variables: FeelingEnergetic

DESCRIPTIVES VARIABLES=FeelingEnergetic
/STATISTICS=MEAN STDDEV MIN MAX.




Transform > Compute Variable > Target Variable: feelingEnergetic_cent > Numeric Expression:
FeelingEnergetic-3.4139

COMPUTE feelingEnergetic_cent=FeelingEnergetic-3.4139.
EXECUTE.

Analyze > Regression > Linear > Dependent: health > Independent(s):
feelingEnergetic_cent

REGRESSION
/MISSING LISTWISE
/STATISTICS COEFF OUTS R ANOVA
/CRITERIA=PIN(.05) POUT(.10) TOLERANCE(.0001)
/NOORIGIN
/DEPENDENT health
/METHOD=ENTER feelingEnergetic_cent.




Question 6 — q6 — regression with dummy variables — 64330.4.1

Make four dummy variables for the categories of the variable SafWorkCond.

Then, run a regression analysis with FeelingEnergetic (USE THE UNCENTERED VARIABLE!) as the
dependent variable and dummy variables for the rst three categories of SafWorkCond as the
independent variables (so use the dummy for the category ‘Not at all satis ed’ as the reference
category).

Fill in the values of the unstandardized regression coe cients that are estimated for the dummy
variables.

Unstandardized b for the dummy for category 1 VERY STATISFIED: 1.649
Unstandardized b for the dummy for category 2 STATISFIED: 1.150
Unstandardized b for the dummy for category 3 NOT VERY STATISFIED: 0.505

3



fi ffi fi