Lecture 12: Multilevel Modelling (Alternative) Flashcards
(36 cards)
What is the main goal of hypothesis testing in psychology?
To explain variation in behavioural data.
What are the two sources of behavioural variation?
Systematic and random variation.
What does NHST stand for?
Null Hypothesis Significance Testing.
What is the p-value in NHST?
The probability of the data assuming H₀ is true.
What p-value threshold is commonly used to reject H₀?
p < .05.
Does NHST test the probability that H₀ is true given the data?
No.
What does NHST actually tell us?
P(Data | H₀).
What is a common misinterpretation of a low p-value?
That H₀ is false.
In NHST, is H₁ ever formally tested?
No.
What statistical analogy shows the importance of base rates?
HIV testing.
What does Bayes’ Theorem calculate?
Posterior probability of a hypothesis.
What two components does Bayes’ Theorem combine?
Prior probability and data likelihood.
What does Bayesian updating involve?
Revising prior beliefs based on new evidence.
What do Bayes Factors compare?
H₀ and H₁.
What is a credible interval?
The range of parameter values most plausible given the data.
What must researchers justify in Bayesian analysis?
Their choice of priors.
Why are Bayesian methods more common now?
Advances in computing and analytic tools.
What does MLM stand for in statistics?
Multi-Level Modelling.
When is MLM used?
When data have a nested or hierarchical structure.
What is a common example of nested data in psychology?
Students within classes or schools.
Why can’t traditional regression be used with nested data?
It assumes observations are independent.
What statistical paradox illustrates why MLM is needed?
Simpson’s paradox.
What does MLM do with variance?
Splits it across levels of the data structure.
What are random intercepts?
Different baseline outcomes for each group.