assumptions in between subject anova Flashcards
define normality
refers to a normal distribution
in between subject ANOVA, it is assumed that the random variable Eij has a normal distribution in all groups
normality justifies the use of the F distribution to evaluate the observed F statistic
which tests are used to assess normality
descriptive and inferential statistics:
- tests for skewness
- kolmogorov-smirnov and shapiro wilk tests
visual methods:
- histograms
- normal quantile (q-q) plot
skewness
skewness represents symmetry and whether the distribution has a long tail in one direction
should be about 0
>0 = positive/right skew
<0 = negative/left skew
kolmogorov smirnov (k-s) test
- very general, but usually less power than S-W
- conceptually, compares sample scores to a set of scores generated from e.g., a normal distribution with a sample mean and standard deviation
shapiro wilk (s-w) test
usually more powerful than k-s test, but only for normal distributions
how to use histograms to assess normality
construct separate histograms for each group and judge it the plots are roughly symmetric
how to use a normal quantile plot to assess normality
- compute percentile rank for each score
(sort from smallest to largest, what percentage of scores are below score x) - calculate theoretical z scores from percentile rank - if scores were normal, what would the z score b
- calculate actual z scores
- plot the observed vs theoretical z scores
if data are close to normal then the points will look like a straight line
what happens when the assumption of normality is not met
type 1 error rates that are lower than the nominal value
usually not as concerning as a violation that results in excessive type one error rates
define homogeneity of variance (i.e., homoscedasticity)
an assumption that states that the variance of the random variable Eij is the same for all groups - a bit of variance is okay, refers to systematic causes that would lead different groups to have different variances
which tests are used to assess homogeneity of variance
- Fmax test
- Levene’s test
- Brown and Forsythe test
Fmax test
calculate the sample variance for each group and find the largest and smallest variances
Fmax= maxs^2g÷mins^2g
Fmax value is compared against a critical value
assumes each group has an equal number of observations (wont use)
levene’s test
measures how much each score deviates from its group mean
Uses absolute deviation scores, Zij, instead of the original scores to run ANOVA
Zij=Yij-Yj
what happens when homogeneity of variance is violated
excessive type 1 error rates (inflates the observed value of the F statistic)
increases in sample size and having equal groups can alleviate some of the inflation
more problematic when smaller groups have larger variance