RM2 Exam 2

1 group posttest only

All participants exposed to IV and measured on DV after

1 group pretest-posttest

Repeated measures, measured on DV before and after

2+ groups posttest only

Each group exposed to different level of IV then measured on DV (experimental/control)

Test for 2 groups posttest only

independent samples t

Test for 3+ groups posttest only

ANOVA

2+ groups pretest-posttest

Repeated measures, measured on DV before and after exposed to different levels of IV (experimental/control)

Experiment

completely random assignment to groups

Quasi-experiment

pre-established groups - classes, hometowns, careers, majors

Therapist wants to test effectiveness of a treatment EXPERIMENT

randomly assigning patients to control group and experimental

Therapist wants to test effectiveness of a treatment QUASI-EXPERIMENT

patients who come in on Mondays are control and patients who come in on Wednesdays

Student wants to study effect of organization on memory EXPERIMENT

giving participants shuffled cards with group numbers on them to assign

Student wants to study effect of organization on memory QUASI-EXPERIMENT

having a research methods class be group 1 and a child development class be group 2

Independent Samples T-Test

Between-Subjects

Step 1 hypothesis testing

state hypothesis (null and alt)

Step 2 hypothesis testing

set alpha level

Step 3 hypothesis testing

obtain random sample

Step 4 hypothesis testing

compare sample to hypothesis (do appropriate calculation)

Step 5 hypothesis testing

test variance/confidence interval - how right was this?

Independent samples t test is appropriate when

2 independent samples, dependent variable is ~normally distributed, interval or ratio, homogeneity of variance

Homogeneity of variance

SD is assumed to be the same for the populations represented by the samples

Dependent Samples T-Test

Within-Subjects

Matched-Pairs

groups matched on a control variable and compared under different conditions of IV

Repeated Measures

2 levels of IV given to the same group OR 1 group compared before and after experiencing IV

Dependent samples t test is appropriate when

2 dependent samples with equal Ns, interval or ratio, dependent variable is ~normally distributed

Nonparametric Statistics

do not require normal distribution or homogeneous variance, nominal or ordinal data

Independent groups with nominal data

1 and 2 way chi squares

1 way chi square

goodness of fit test

2 way chi square

test of independence

1 way chi square is appropriate when

participants of one category are measured alongside 2+ other categories (2x1 chart), each participant falls in one independent category

Expected frequency for chi squares must be

5 or above in all categories

2 way chi square is appropriate when

2 categories measured against another variable with 2+ categories, each participant falls in one independent category

McNemar Test

2 dependent groups with 2 categories (dichotomous)

Cochran Test

3+ dependent groups with 2 categories (dichotomous)

Spearman’s Rho

Correlation for ranked data

Mann Whitney U Test

2 independent samples, ordinal data, N<20

Rank Sums Test

2 independent samples, ordinal data, N>/=20

Wilcoxon T-Test

2 dependent samples, ordinal data

Kruskal-Wallis H Test

3+ independent samples, ordinal data, N>/=5

Friedman Chi Square

3+ dependent samples, N>/=5 if 4 conditions, N>/=10 if 3