Psyc460 exam Flashcards

(53 cards)

1
Q

Control for familywise error in planned comparisons

A

Desired familywise error divided by number of planned comparisons.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is the variance

A

covariance of a variable with itself

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

ss treatment - ss contrast =

A

ss residuals

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Three rules of contrast coding

A

compare negative and positive
sum to 0
If a variable isn’t in a contrast code it as 0.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is the equivalent of r squared in anova

A

partial eta squared

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Type I error

A

1 - alpha

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Type II error

A

alpha

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Familywise error

A

(1 - alpha) ^c

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Per comparison error rate

A

alpha

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What is chi square

A

A measure of discrepancy

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What does chi square change mean

A

A lower chi square change means the model is an improvement in fit.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What is AIC / BIC

A

Model comparison statistics for CFA, EFA and SEM

with non-nested models

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Advamtages of SEM vs. path analysis

A

less restrictive assumptions, inclusion of latent variables, models measurement error

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What methods test for omitted errors

A

None

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Formula for calculating number of values in variance, co variance matrix

A

p(p+1_ /2

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Formula for centreing around the mean

A

x = raw - mean

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Dummy coding for ANOVA

A

a. k-1=j b. k-1=j interaction: j x j

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

what do straight lines mean in path diagrams

A

causal direct effects

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

what do curved lines mean in path diagram

A

covariance. No causal links.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Out of FA or SEM which one has causal effects.

A

SEM

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Calculating degrees of freedom

A

no. of parameters - values in covariance-variance matrix

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Calculating no. of parameters in SEM model

A

errors + loadings + covariance

23
Q

What are examples of non-nested models

A

CFA, EFA, SEM

24
Q

What are examples of nested models

A

t tests. ANOVA, ANCOVA

25
Two tests for mediation
sobels test, bootstrap in Amos with path analysis
26
Mediation is
When a variable gets in the middle
27
Moderation is
When a variable acts jointly with another to change the strength or direction of a variable
28
Moderation is analogous to
Anova interaction
29
How to choose between anova or multiple regression
Anova for categorical variable | Multiple regression for continuous variables
30
What method uses linear contrasts
Anova.
31
What method uses trend analysis | What sort of treatment variables are needed
Anova It is an application of linear contrasts. Need a quantitative treatment v ariable
32
Process for using multiple regression to test moderation
Centre variables to prevent multi collinearity | Multiply a times b to get interaction variable
33
What does it mean To partial a variable
Control for a variable
34
What does semi partial correlation mean
Find the part of the variance that is unique to the total outcome variance
35
Sensitivity is the probability of a
Hit
36
1 - Specificity
Is the probability of a false alarm. An incorrect hit.
37
In correlation what is a suppressor effect
When you partial out a variable and the correlation between the remaining two variables increases or changes direction
38
Specificity is the probability of
Correctly identifying a non case (negative case)
39
Anova is multiple regression with what type of variables
Categorical
40
How do u choose the number of factors to extract in EFA?
Use Kaiser criterion : include all factors eigenvalue >1 Scree plot
41
What is an orthogonal contrast?
Factors are independent. Not correlated
42
What is an EFA oblique rotation?
Factors are correlated
43
What is trend analysis in anova
An application of linear contrasts | Need a quantitative treatment such as dosage, intensity.
44
What Is sphericity?
Each pair of ivs has the same correlation. Test with mauchleys test If violated use greenhouse geiser, huynh feldt.
45
What is type iii sum of squares
Default used by spss | Test each effect by subtracting all main effects from the main model.
46
What is a generalised linear models
It has a transformation applied. Example are: Logistic regression Cox regression Negative log liklihood Lower is better
47
Fix multicollinearity
Combine remove CFA Cooks distance > 1 is outlier / influential Case. Omot and rerun
48
Trend analysis
When groups differ on a quantitative variable
49
Which methods need sphericity removed?
Repeat measures anova | Mixed models
50
Anova assumptions
Homogeneity of varianve Normality Indepemdence
51
Principal components analysis
Data reduction method | All variance accounted for
52
Principal axis factoring
Factor analysis method Only use shared varianxe More theoretical than paf
53
What is an eigenvalue
How much variance each factor accounts for