13
Experimental Research Methods (Fach) / L2 (Lektion)
:)
Diese Lektion wurde von tklmn erstellt.
Diese Lektion ist leider nicht zum lernen freigegeben.
Anzeige
- power of a test what is the probability of drawing the correct conclusion h1 is true and h0 is rejected
- type I error rejecting H0 when it is true
- type II error not rejecting h0 when h1 is true
- 4 steps for determining power 1. determine Zcv vor the given h0 2. determine sample mean Xcv that belongs with Zcv for the given sample mean 3.convert critical value Xcv to the Zh1 value for the given H1 4. power is equal to the chance P(Z=Zh1 l H1)
- 4 factors that influence power -alpha -sample size -dispersion -the true population mean
- two important measures of effect size cohens d -> how large is the relative difference in groups partial explained variance -> how much of the variance is explained by group membership
- rules of thumb for interpreting effect sizes small d = 0.2 n² = 0.01 medium d = 0.5 n² = 0.06 large d = 0.8 n² = 0.14
- ANOVA cp,üares means across groups
- why ANOVA instead of multiple T-tests? conducting a large number of t-tests leads to inflated type 1 error rate
- logic of ANOVA split total variance ->look at the distance between groups compared to the distance within groups differences between grand mean and group means can be explained by group membership a larger between than within groups variance is indivative for a treatment effect
- Anzeige
- factors the independent categorical variables
- levels the categories of the factors
- condition each combination of levels