SPSS cheat sheet
Block 2
One-sample t test: ANALYZE COMPARE MEANS ONE-SAMPLE T TEST (tick “effect sizes”)
Independent-samples t test: ANALYZE COMPARE MEANS INDEPENDENT-SAMPLES T TEST add
test variable & grouping variable (define groups)
Paired-samples t test:
Method 1: ANALYZE COMPARE MEANS PAIRED-SAMPLES T TEST add variables 1 and 2
Method 2: TRANSFORM COMPUTE NEW VARIABLE µ1 – µ2 (new variable “differences”)
ONE-SAMPLE T TEST for new “differences” variable
! always check if they ask for µ1 – µ2 or µ2 – µ1 (change var 1 and var 2 when doing the t test)
*when reporting test statistics & p values => always look at the equal variances not assumed row in the output
Effect size: reported with a statistical test (tick “effect sizes” when doing a test)
! always reported positive (even if it appears negative in SPSS)
Confidence interval: reported with a statistical test (to change percentage go to OPTIONS
CONFIDENCE INTERVAL PERCENTAGE change from 95% to 90%)
Report means/medians etc.: ANALYZE COMPARE MEANS MEANS add variables in the
“Dependent List” OPTIONS add “means” and “medians” into the box
Wilcoxon rank-sum test:
1. Change variable from Nominal to Ordinal (in VARIABLE VIEW) (or not?)
2. ANALYZE NONPARAMETRIC TESTS LEGACY DIALOGS 2 INDEPENDENT SAMPLES add test
variable & define groups tick “Mann Whitney”
(look at the p value in the “Asymp. Sig (2-tailed)” row)
Wilcoxon signed-rank test: ANALYZE NONPOARAMETRIC TESTS LEGACY DIALOGS 2 RELATED
SAMPLES add variables tick “Wilcoxon”
! when reporting the p-value, check whether the test is one-sided & split sig. in half
Chi-square test for independence: ANALYZE DESCRIPTIVE STATISTICS CROSSTABS add one
variable in Rows, one in Columns STATISTICS… tick “Chi-square” and “Phi & Cramer’s V”
if you want to see expected frequencies: go to CELLS… tick “Expected”
Chi-square test for goodness-of-fit: ANALYZE NONPARAMETRIC TESTS LEGACY DIALOGS CHI-
SQUARE add variables ok
Block 2
One-sample t test: ANALYZE COMPARE MEANS ONE-SAMPLE T TEST (tick “effect sizes”)
Independent-samples t test: ANALYZE COMPARE MEANS INDEPENDENT-SAMPLES T TEST add
test variable & grouping variable (define groups)
Paired-samples t test:
Method 1: ANALYZE COMPARE MEANS PAIRED-SAMPLES T TEST add variables 1 and 2
Method 2: TRANSFORM COMPUTE NEW VARIABLE µ1 – µ2 (new variable “differences”)
ONE-SAMPLE T TEST for new “differences” variable
! always check if they ask for µ1 – µ2 or µ2 – µ1 (change var 1 and var 2 when doing the t test)
*when reporting test statistics & p values => always look at the equal variances not assumed row in the output
Effect size: reported with a statistical test (tick “effect sizes” when doing a test)
! always reported positive (even if it appears negative in SPSS)
Confidence interval: reported with a statistical test (to change percentage go to OPTIONS
CONFIDENCE INTERVAL PERCENTAGE change from 95% to 90%)
Report means/medians etc.: ANALYZE COMPARE MEANS MEANS add variables in the
“Dependent List” OPTIONS add “means” and “medians” into the box
Wilcoxon rank-sum test:
1. Change variable from Nominal to Ordinal (in VARIABLE VIEW) (or not?)
2. ANALYZE NONPARAMETRIC TESTS LEGACY DIALOGS 2 INDEPENDENT SAMPLES add test
variable & define groups tick “Mann Whitney”
(look at the p value in the “Asymp. Sig (2-tailed)” row)
Wilcoxon signed-rank test: ANALYZE NONPOARAMETRIC TESTS LEGACY DIALOGS 2 RELATED
SAMPLES add variables tick “Wilcoxon”
! when reporting the p-value, check whether the test is one-sided & split sig. in half
Chi-square test for independence: ANALYZE DESCRIPTIVE STATISTICS CROSSTABS add one
variable in Rows, one in Columns STATISTICS… tick “Chi-square” and “Phi & Cramer’s V”
if you want to see expected frequencies: go to CELLS… tick “Expected”
Chi-square test for goodness-of-fit: ANALYZE NONPARAMETRIC TESTS LEGACY DIALOGS CHI-
SQUARE add variables ok
