TERM 1
Chi-Square Independence Test
DEFINITION 1
- 2 categorical values a re independent in some population (c ar,
gender, 2 die)Analyze --> Descriptive statistics --> Crosstabs
(under "statistics" click "chi-square"- Nominal and NominalEx. p-
value is 0.000, means there is a 0% c hance to find the observed
(or larger) degree of association betw een the variables if they are
perfectly independent in the pop.
TERM 2
One Sample Chi-Square test (Goodness of Fit)
DEFINITION 2
- 1 categorical value & 1 group of cases (1 die, gender,
brand of smartphone)Analyze --> nonparametric tests -->
legacy dialogues --> chi-square- NominalEx. p-value is 0.073,
which is more than 0.05, therefore we accept the Null
hypothesis and conclude that the brands are equally
attractive to the population (no sig. diff)
TERM 3
Independent Samples T-Test
DEFINITION 3
- 1 metric outcome variable & 2 groups o f cases - Test is
means in 2 pops. on one metric varia ble are equalAnalyze -->
compare means --> independent sa mples t-test- Test variable
(top): Euros (scale/interval/ratio)- Grouping variable (bottom):
gender (nominal/ordinal) *always- De fine groups 1 &2Ex. p-value is
0.116, which is larger than 0.05. Ther efore we accept the Null
hypothesis and there is no sig. diff.
TERM 4
Paired Samples T-Test (before &
after)
DEFINITION 4
- 2 metric outcomes & 1 group of cases Analyze --> compare
means --> paired samples t-test- Sca le: avg. time before & avg.
time after- Look at mean to compar eEx. p-value is 0.0083, which is
smaller than 0.05, meaning the differ ence is significant and we
reject the Null hypoth. "Even a single beer slows people down on
given tasks" or "difference before an d after campaign. Was it
effective?"
TERM 5
One-Way ANOVA
DEFINITION 5
- Testing if 3 or more means are all equal- 1 metric
outcome variable & 3+ groups of casesAnalyze -->
compare means --> one-way ANOVA... Options -->
Descriptive*- Dependent list (top): weight in grams
(scale/interval/ratio)- Factor (bottom): fertilizer used
(nominal/ordinal)
