Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
44 Cards in this Set
- Front
- Back
Analysis of variance
|
ANOVA, a technique for testing differences in the means of several groups
|
|
One-way ANOVA
|
an analysis of variance wherein groups are defined on only one independent variable
|
|
Omnibus null hypothesis
|
the hypothesis that all population means are equal
|
|
Assumption of normality
|
sampling distribution is normal
|
|
Assumption of heterogeneity of variance
|
each population has same variance
|
|
Assumption of independence of samples
|
all observations are independent of each other
|
|
Logic of ANOVA
|
average value of sample variances is an estimate of population variances, does not depend on the truth or falsity of the null
|
|
MSwithin
|
aka MSerror, variability among subjects in the same treatment group
|
|
MSbetween groups
|
aka MSgroups or MS treatment, variability among group means
|
|
Average of the sample variances
|
MSerror
|
|
Variance of the sample means multiplied by the sample size
|
MSgroup
|
|
MSerror
|
estimate of pop mean independent of the truth of the null, average variance of each sample in treatment
|
|
MS group
|
pop variance estimate only when the null is true, variance of groups corrected by n to produce estimate of population variance
|
|
When MS error and MS group agree
|
support for the truth of the null
|
|
Sum of squares
|
the sum of the squared deviations about the mean
|
|
SStotal
|
the sum of squared deviations of all the scores, regardless of group membership
|
|
SSgroup
|
the sum of squared deviations of the group means from the grand mean, multiplied by the number of observations
|
|
SSerror
|
the sum of squared residuals or the sum of the squared deviations within each group
|
|
SSerror formula
|
SSerror = SStotal – SSgroup
|
|
SStotal formula
|
SStotal = SSerror + SSgroup
|
|
dftotal for ANOVA
|
N-1, N = number of total participants
|
|
dfgroup for ANOVA
|
k-1, k = number of groups
|
|
dferror
|
k(n-1)
|
|
F ratio/statistic
|
the ratio of MS group over MS error
|
|
Logic of F ratio
|
closer to 1 = null is true because MSerror and MSgroup will be approximately the same
|
|
Bonferroni procedure
|
a multiple comparisons procedure in which the familywise error rate is divided by the number of comparisons
|
|
Logic of Bonferroni
|
omit requirement that F be significant, run all tests at α over c, where c = number of comparisons
|
|
Tukey procedure
|
a multiple comparison procedure designed to hold the familywise error rate at α for a set of comparisons, compares every mean with every other mean
|
|
Violations of assumptions
|
if the largest variance is no more than four or five times the smallest = no problem
|
|
Heterogeneity and unequal sample sizes
|
do not mix
|
|
Eta squared
|
ηˆ2, a measure of the magnitude of effect, SSgroup over SStotal
|
|
Bias of eta squared
|
tends to overestimate the value we would obtain if we were able to measure the whole population
|
|
Omega squared
|
ω^2, a less biased measure of magnitude of effect
|
|
Factors
|
independent variables in ANOVA world
|
|
Two-way factorial design
|
a design involving two independent variables in a way in which every level of every variable is paired with every level of the other variable
|
|
Factorial design
|
an experimental design in which every level of each variable is paired with every level of every variable
|
|
Interaction
|
a situation in a factorial design in which the effects of one independent variable depend on the level of another independent variable
|
|
Notation: i and j
|
i = row and j = column
|
|
Main effect
|
the effect of one independent variable averaged across the levels of the other independent variable
|
|
Simple effect
|
the effect of one independent variable at one level of another
|
|
Notation: nc
|
n = number of participants in the group, c = level
|
|
SScells
|
the sum of squares assessing the differences among the cell means
|
|
Differences in cell means
|
other than sampling error, coming from different levels of the same variable, different levels of differing variables or because of an interaction
|
|
Parallel lines
|
means that there is NO interaction effect present
|