Repeated Measures
A --- ---- design involves multiple observations from the same experimental unit under different conditions
repeated measures
What is the number of conditions, factors, and levels in a 2 x 2 x 2 design.
3 factors with 2 levels each, for a total of 8 conditions
What effect is present here?[1]
Interaction between pressure and anxiety
Define what a p-value represents.
The p-value is the probability of obtaining your observed results under the null hypothesis.
Which UGIAs teach Wednesday lab? :)
all of them!
Which assumption of the GLM is violated when we take repeated measures from the same subject, and how do we solve this problem?
1. Independence; 2. by treating subject as a factor in the model
What does it means to have an interaction in a factorial design?
Any of the following:
- The degree to which the effect of one factor depends on the level of the other
- the degree to which the mean for treatments differs from the additive effects of factors
- difference between levels of Factor A is different for different levels of Factor B
- difference between simple effects
What effects are present here?[2]
Main effect of Anxiety
Interaction between Anxiety and Pressure
This is when no effect is present but a researcher rejects the null hypothesis.
Type I Error aka alpha error aka "false alarm"
What is the honor's heirloom?
a junior varsity silver medal for track (but still very honorable)
How do we (generally) determine fixed and random effects in a repeated measures study design?
Fixed: levels of interest; actual levels of the experiment
Random: levels represent a sample from all the possible levels; i.e. subject(we aren't theoretically interested in specific subjects)
What is the meaning of the red box.[4]
The main effect (aka the estimate of effect offset) for level j
What kind of design is this?[3]
Mixed design
Why is the null hypothesis the empty model? (Hint: Think of this in terms of what b1 would mean)
The null hypothesis is that there's no difference between groups --> b1 is the parameter that represents the difference between groups --> if there is no difference between groups, then b1 is zero --> when b1 is zero, the predicted score is b0 for everyone
Two truths and a lie (Which is the lie?):
1. Dr. Geller leads a dance team.
2. Dr. Geller has played League of Legends before.
3. Dr. Geller's son's favorite Pokémon is Kabutops.
Dr. Geller has NOT played league before (y'all thought she had?)
What would it mean if all (e)ij aka (pt)ij are equal to zero?
The effect of treatment is the same for every subject.
note: (pt)ij = person*treatment interaction
How do you calculate the mean square of factor A (MSA)?
SSa/dfa
note: calculate any mean square by dividing a sum-of-squares value by the corresponding degrees of freedom
It is demonstrated by this sentence: "When the wine is cheap, high quality wines are rated 16 points higher than low quality wines."
A simple effect of quality on cheap wines
This is the probability of making one or more false alarms when performing multiple pairwise comparisons
familywise error rate (aka experiment wise error rate)
Note: multiple comparisons lead to "alpha escalation"
Whose birthday is coming up next month?
Inez
What is changed/added to the following code to turn something from a between-subjects design to a repeated measures design:
model = lm(OUTCOME ~ PREDICTORS, data = DATA)
followed by anova(model)
model = aov(OUTCOME ~ PREDICTORS + Error(SUBJECT / PREDICTORS), data = DATA)
followed by summary(model) or tidy(model)
What is the difference between a linear mixed model and a mixed design?
LMM: a combination of fixed and random effects
Mixed design: mixing between and within-subjects factors
In a factorial design table, what do you look at to determine main effects and what do you look at to determine simple effects?
Main effects: are marginal means different
Simple effects: difference cell means (aka are the means from one row different? are the means from one column different?)
This is the number of pairwise comparisons possible for 103 groups.
Formula if j = # of groups: j*(j-1)/2
103*102/2 = 5253
What are the UGIAs' last names
Nguyen, Niu, Segall, Sandoval