Stats for psychology Flashcards
What is a factorial design?
when there is more than one IV.
Why would there be three hypotheses for a two-factor ANOVA?
Because the interaction is included.
What should be done after an ANOVA is carried out?
Pairwise comparisons to see which means differ, and simple effect analyses to interpret the interaction.
What do simple effects do?
examine the pattern of results that are associated with one IV as the other IV is changed
When can you skip simple effects and go straight to comparisons?
when one experimental factor has just 2 levels.
What are pairwise comparisons for?
To compare means following a significant interaction effect.
What is the problem with conducting pairwise comparisons and simple effects analyses after an ANOVA? How do you correct this?
greater the risk of a type 1 error. To counteract this, make a Bonferroni correction.
When is it ok not to do a Bonferroni correction?
for tests within the same study that are independent of each other, and if the additional stats were planned.
What are post Bem statistics?
In 2011 lots of psychologists reported false stats and just generally weren’t good. Due to this, there became a new wave of statisticians, divided into two groups: confidence intervals & meta analyses, and Bayesian statistics.
What are confidence intervals?
They provide an estimate of the true population parameter, and quantify how precise it is.
What is the population parameter?
unknowable and fixed values of the DV if we tested everyone that exists.
How is a confidence interval interpreted?
The width of the confidence interval reflects the precision of the sample estimate. A larger confidence interval suggests uncertainty.
What is a meta-analysis?
the statistical combination of several studies results on the same issue.
What are Bayesian statistics?
start with a hypothesis, and conduct research to see how probable it is that the hypothesis is a good explanation of the data collected, rather than the other statistical perspectives we have been learning, which examine how unlikely the data is if the reality is that there is no difference.
