Lecture 12: Multilevel Modelling (Alternative)

Question 1

What is the main goal of hypothesis testing in psychology?

Answer 1

To explain variation in behavioural data.

Question 2

What are the two sources of behavioural variation?

Answer 2

Systematic and random variation.

Question 3

What does NHST stand for?

Answer 3

Null Hypothesis Significance Testing.

Question 4

What is the p-value in NHST?

Answer 4

The probability of the data assuming H₀ is true.

Question 5

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

Answer 5

p < .05.

Question 6

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

Answer 6

No.

Question 7

What does NHST actually tell us?

Answer 7

P(Data | H₀).

Question 8

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

Answer 8

That H₀ is false.

Question 9

In NHST, is H₁ ever formally tested?

Answer 9

No.

Question 10

What statistical analogy shows the importance of base rates?

Answer 10

HIV testing.

Question 11

What does Bayes’ Theorem calculate?

Answer 11

Posterior probability of a hypothesis.

Question 12

What two components does Bayes’ Theorem combine?

Answer 12

Prior probability and data likelihood.

Question 13

What does Bayesian updating involve?

Answer 13

Revising prior beliefs based on new evidence.

Question 14

What do Bayes Factors compare?

Answer 14

H₀ and H₁.

Question 15

What is a credible interval?

Answer 15

The range of parameter values most plausible given the data.

Question 16

What must researchers justify in Bayesian analysis?

Answer 16

Their choice of priors.

Question 17

Why are Bayesian methods more common now?

Answer 17

Advances in computing and analytic tools.

Question 18

What does MLM stand for in statistics?

Answer 18

Multi-Level Modelling.

Question 19

When is MLM used?

Answer 19

When data have a nested or hierarchical structure.

Question 20

What is a common example of nested data in psychology?

Answer 20

Students within classes or schools.

Question 21

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

Answer 21

It assumes observations are independent.

Question 22

What statistical paradox illustrates why MLM is needed?

Answer 22

Simpson’s paradox.

Question 23

What does MLM do with variance?

Answer 23

Splits it across levels of the data structure.

Question 24

What are random intercepts?

Answer 24

Different baseline outcomes for each group.

Question 25

What are random slopes?

Answer 25

Predictor effects that vary across groups.

Question 26

What does the null model in MLM check for?

Answer 26

Whether there’s meaningful variation between groups.

Question 27

What is added in Stage 2 of MLM model building?

Answer 27

Level-1 predictors.

Question 28

What is added in Stage 3 of MLM model building?

Answer 28

Level-2 predictors and cross-level interactions.

Question 29

What is a key benefit of MLM over aggregating data?

Answer 29

It preserves within-group variance and statistical power.

Question 30

Can MLM handle repeated measures data?

Answer 30

Yes.

Question 31

What are fixed effects in MLM?

Answer 31

Effects assumed to be constant across all groups.

Question 32

What are random effects in MLM?

Answer 32

Effects allowed to vary across groups.

Question 33

In the speed dating example, what required MLM?

Answer 33

Predicting outcomes using both interaction- and individual-level variables.

Question 34

What is the main risk of using standard regression on nested data?

Answer 34

Misleading or incorrect conclusions.

Question 35

Why is MLM increasingly important in psychology?

Answer 35

Because much psychological data is clustered.