SPAUD 505: Inferential Stats, Results Section Part II

inferential statistics

inferences based on sample of data collected regarding the broader population
-used for data analysis

tests of differences between two groups/conditions, single DV

-BSD
-WSD
-parametric statistics
-nonparametric statistics

independent-samples t test

BSD

dependent-samples- t test

WSD
data are paired b/c the same subjects are measured twice

parametric stats

statistics for ratio and interval data, has assumptions that must be fulfilled, depends on degrees of freedom

degrees of freedom depends on

sample size

z and t tests are ______ statistics

parametric statistics

z tests

-parametric stat
-used with large sample size (n > 30)
-standardized scores
-fixed critical values

critical values in z tests depend on

level of significance, hypotheses
(directional or nondirectional)

nonparametric stats

stats used for nominal & ordinal data

Wilcoxon matched-pairs signed ranks test

Used to determine whether two correlated means are significantly different. Can be utilized in place of the related samples t test when parametric assumptions cannot be met, measures over time
-WSD

Mann-Whitney U test

determines whether 2 uncorrelated means differe significantly when data are nonparmetric
(memory: the 'u' reminds you of 'uncorrelated')
-BSD

tests of differences among multiple groups/conditions

Parametric stats
Nonparametric stats

parametric stats for tests of differences between multiple groups/conditions

-one-way ANOVA
-two-way ANOVA
-repeated-measures ANOVA
-mixed ANOVA

analysis of variance (ANOVA)

a statistical technique that determines whether three or more means are statistically different from one another, determines how much variances are due to sampling error vs tx effect

one-way ANOVA

-1 IV: >2 levels
-1 DV

two-way ANOVA

-2 IV (w/ greater than or equal to 2 levels)
-1 DV
-calculates 3 F ratios

two-way ANOVA assesses

main effect per IV & interaction between IV

repeated-measures ANOVA

F test used for WSD
-1 DV

Mixed ANOVA

F test for BSD + WSD
-1 DV

nonparametric stats used to test differences among multiple groups/conditions

-Kruskal-Wallis one-way ANOVA (BSD)
-Friedman's two-way ANOVA test (WSD)

Kruskal-wallis one-way ANOVA

Independent-samples test, BSD, nonparametric

Friedman's two-way ANOVA test

Dependent-samples test, WSD, nonparametric

t tests in inferential statistics

identify which group of means are different from each other

effect size indicates 1. _____/______ significance of data, and 2. extent to which _____/______ of DV is explained by the IV

practical/clinical significance of data

value/change in DV is explained by IV

effect size is an ____ _____ of the plausibility of the H0, it is not interpreted as "statistically sig"

independent index

an effect size of 0 means

there is no relation between the DV and IV

effect size is represented by _____ and _____

Eta squared & Cohen's d

interpret the value of 0.90 for Eta-squared

90% chance the variability in DV is accounted for by IV