Lecture 12: Multilevel Modelling (Alternative) Flashcards

(36 cards)

1
Q

What is the main goal of hypothesis testing in psychology?

A

To explain variation in behavioural data.

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

What are the two sources of behavioural variation?

A

Systematic and random variation.

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

What does NHST stand for?

A

Null Hypothesis Significance Testing.

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

What is the p-value in NHST?

A

The probability of the data assuming H₀ is true.

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

What p-value threshold is commonly used to reject H₀?

A

p < .05.

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

Does NHST test the probability that H₀ is true given the data?

A

No.

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

What does NHST actually tell us?

A

P(Data | H₀).

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

What is a common misinterpretation of a low p-value?

A

That H₀ is false.

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

In NHST, is H₁ ever formally tested?

A

No.

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

What statistical analogy shows the importance of base rates?

A

HIV testing.

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

What does Bayes’ Theorem calculate?

A

Posterior probability of a hypothesis.

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

What two components does Bayes’ Theorem combine?

A

Prior probability and data likelihood.

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

What does Bayesian updating involve?

A

Revising prior beliefs based on new evidence.

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

What do Bayes Factors compare?

A

H₀ and H₁.

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

What is a credible interval?

A

The range of parameter values most plausible given the data.

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

What must researchers justify in Bayesian analysis?

A

Their choice of priors.

17
Q

Why are Bayesian methods more common now?

A

Advances in computing and analytic tools.

18
Q

What does MLM stand for in statistics?

A

Multi-Level Modelling.

19
Q

When is MLM used?

A

When data have a nested or hierarchical structure.

20
Q

What is a common example of nested data in psychology?

A

Students within classes or schools.

21
Q

Why can’t traditional regression be used with nested data?

A

It assumes observations are independent.

22
Q

What statistical paradox illustrates why MLM is needed?

A

Simpson’s paradox.

23
Q

What does MLM do with variance?

A

Splits it across levels of the data structure.

24
Q

What are random intercepts?

A

Different baseline outcomes for each group.

25
What are random slopes?
Predictor effects that vary across groups.
26
What does the null model in MLM check for?
Whether there’s meaningful variation between groups.
27
What is added in Stage 2 of MLM model building?
Level-1 predictors.
28
What is added in Stage 3 of MLM model building?
Level-2 predictors and cross-level interactions.
29
What is a key benefit of MLM over aggregating data?
It preserves within-group variance and statistical power.
30
Can MLM handle repeated measures data?
Yes.
31
What are fixed effects in MLM?
Effects assumed to be constant across all groups.
32
What are random effects in MLM?
Effects allowed to vary across groups.
33
In the speed dating example, what required MLM?
Predicting outcomes using both interaction- and individual-level variables.
34
What is the main risk of using standard regression on nested data?
Misleading or incorrect conclusions.
35
Why is MLM increasingly important in psychology?
Because much psychological data is clustered.
36