Psych Methods & Stat II Quiz 3

independent groups design

definition: each participant takes part in only one condition of the experiment

strengths: no order effects

limitations: participant differences can bias results

correlated groups design

definition: participants matched on key traits, then split into conditions

strengths: reduces individual differences

limitations: matching is difficult and time-intensive

nonexperimental designs

definition: no manipulation, no random assignment

strengths: real-world, ethical research

limitations: cannot establish cause and effect

experimental designs

definition: manipulation, control, and random assignment

strengths: can show cause and effect

limitations: may lack real-world realism

ex post facto studies

definition: study of existing differences after the event

strengths: can study otherwise unethical/impossible topics

limitations: no definitive cause and effect conclusions

single group posttest only design

definition: one group + treatment + posttest only

strengths: simple and fast

limitations: very weak evidence

single group pretest-posttest design

definition: one group measured before and after treatment

strengths: can track change

limitations: no control group; weak causal evidence

matched participants design

definition: participants matched on key traits, then split into conditions

strengths: reduces individual differences

limitations: matching is difficult and imperfect

counterbalancing

definition: varying the order of conditions in repeated measures

strengths: reduces order effects

limitations: can become complex and resource-heavy

repeated measures designs

definition: same participants in all conditions

strengths: controls individual differences

limitations: order effects

solomon’s 4 group design

definition:

strengths:

limitations:

multilevel, randomized between-subjects design

definition:

strengths:

limitations:

randomized, pretest-posttest control group design

definition:

strengths:

limitations:

randomized, posttest-only control group design

definition: randomly assigned groups, treatment vs. control, posttest only

strengths: strong causal inference without testing effects

limitations: no pretest

type 1 error

rejects the null hypothesis even though it is actually true

type 2 error

fail to reject the null hypothesis even though it is false

random assignment vs. random selection

selection: choosing participants randomly

assignment: assigning participants to groups randomly

how does the f-ratio provide a test of whether there is a difference between groups?

large f-ratio: groups likely differ

small f-ratio: differences likely due to chance

what does the numerator represent?

between group variance

what does the denominator represent?

within group variance

what do we expect the f-ratio to be if the null hypothesis is true?

all groups are equal and results are due to chance

what does effect size mean for ANOVA?

how much of the variation in the dependent variable is explained by the independent variable