research methods exam 2 anova

what is the problem of alpha inflation and how do we resolve it?

alpha inflation (aka type 1 error inflation) happens when you run too many statistical tests at once. each test has a small chance of giving a false positive. you can fix it by adjusting your criteria for significance.

what are some features of the f-distribution?

• always positive

• centered around 1, not 0

• skewed distribution

• varies depending on 2 diff types of degrees of freedom

• always one tailed

what is the logic of the f-test?

the f-test checks whether the differences between group means are big enough to be unlikely due to chance. it compares between-groups variability and within-groups variability. between-groups variability is how different the group means are from the overall mean. within-groups variability is how spread out the data is inside each group.

what does it mean when one is bigger than the other (between groups variability or within groups variability)

when between is bigger than within: this means the group means are noticeably different from each other

when within is bigger than between: the group means are not very different from each other

oneway randomized ANOVA

• one between-subjects IV with 3 or more levels

• one continuous DV

• examples: during what time of day (morning, noon, afternoon, night) do people stereotype the least? what kind of treatment (psychoanalytic, CBT, group therapy) is most effective at treating bipolar disorder?

what are some weaknesses of oneway randomized ANOVA and how can they be resolved

• assumes equal variance

• assumes normal distribution

• can only test one independent variable

• doesn’t tell which groups differ

• sensitive to outliers

• to resolve: run tests, transform data, use different ANOVA

oneway repeated measures ANOVA

• 1 IV with 3 or more levels

• each participant completes dependent measures in all conditions of the IV

• examples: do people differentially remember white, black, and east Asian faces? do people feel differently about their self-efficacy after eating a cookie, a granola bar, or a piece of carrot? does cognitive performance improve after each of 5 counseling sessions?

what are some strengths and weaknesses of repeated measures designs and how can they be resolved

pros: participants serve as their own controls (reduced error, more power), need fewer participants

cons: order effects, practice effects, may guess hypothesis, longer studies, limits possibilities for experimental manipulations

why would we need to conduct post hoc tests for a oneway ANOVA

because a significant ANOVA result tells you that at least one group is different, but it doesn’t tell you which groups are different

what are some of the common post hoc tests

Tukey’s HSD, Bonferroni correction, Scheffe’s test, Games-Howell

what is effect size and how is it different from a hypothesis test

effect size: measures how big or meaningful a result is

hypothesis test: measures if an effect is likely to be real, or could it just be due to random chance

what is the name of the statistic we use to calculate effect size for the F-test

eta squared

factorial ANOVA

• looking at more than one IV at the same time and want to see how they interact (2 or more IVs)

• one continuous DV

• example: how does type of therapy and medication dose affect depression scores?

what are the three different kinds of factorial ANOVAs

• between-subjects factorial ANOVA: each participant experiences only one combination of the IVs

• within-subjects factorial ANOVA: all participants experience all levels of the IVs

• mixed factorial ANOVA: one IV is between-subjects, one is within-subjects

what are the three different kinds of effects we can look for in a factorial ANOVA, how are they detected on a graph

main effects - the impact of a single IV on the DV. on a graph, differences in the average height of bars between levels of the factor would indicate a main effect

interaction effects - occurs when the impact of one factor depends on the level of the other factors. on a graph, crossing patterns across different groups can indicate an interaction

simple effects - looks at the effect of one factor at a specific level of the other factor. look at the differences within specific conditions