Deck 21: Multilevel Linear Models

Question 1

Take a given variable,mark each score and subtract the mean of all scores for that variable.This is an example of?

A)Grand mean centring
B)Group mean centring
C)Mean analysis
D)None of the above

Answer 1

A

Question 2

In the Sig.column of the table of Type III Tests of Fixed Effects,what is the maximum value for the effect to be considered significant?

A)1.00
B).75
C).10
D).05

Answer 2

D

Question 3

What is the intraclass correlation coefficient (ICC)a good gauge of?

A)used when quantitative measurements are made on groups of units
B)used when quantitative measurements are made on single units
C)used when qualitative measurements are made on groups of units
D)used when qualitative measurements are made on single units

Answer 3

A

Question 4

An effect is random if the experiment contains what?

A)All possible treatment conditions that a researcher is interested in
B)None of the possible treatment conditions that a researcher is interested in
C)A random sample of possible treatment conditions that a researcher is interested in
D)A fixed number and percentage of possible treatment conditions that a researcher is interested in

Question 5

How can the Information Criteria table be used in multilevel modelling?

A)To test the significance of the covariant
B)To assess the overall fit of the model
C)To reduce the random effects of the model
D)To determine the regression coefficient for each effect

Question 6

A cubic trend growth curve is also known as?

A)A first-order polynomial
B)A second-order polynomial
C)A third-order polynomial
D)None of the above

Question 7

AICC is the acronym used for which version of the log-likelihood value?

A)Akaike's information criterion
B)Hurvich and Tsai's criterion
C)Bozdogan's criterion
D)Schwarz's Bayesian criterion

Question 8

Intraclass correlation can be used to estimate what?

A)Dependency between scores
B)Error scores
C)Standard deviation
D)Significance of the data

Question 9

SPSS produces four adjusted versions of the log-likelihood value.True or false?

Question 10

Multilevel models where changes over time in an outcome are modelled using potential growth patterns,such as linear or quadratic,are called what?

A)Simple models
B)Regression models
C)High models
D)Growth models

Question 11

When data are organized in hierarchies,what are the terms used to describe variables within other variables?

A)Nestled or bunched
B)Caught or grouped
C)Wedged or collected
D)Clustered or nested

Question 12

The table of Estimates of Fixed Effects tells us about any random effect in the model.

Question 13

A polynomial can be defined by a simple power function.

Question 14

In multilevel models,why can centring be useful?

A)To combat multicollinearity between predictor variables
B)To reject the null hypothesis early on
C)To improve the strength of the analysis
D)To hide errors

Question 15

When data are measured over time,an autoregressive structure (AR1)cannot be assumed.

Question 16

What does it mean if the BIC value is small?

A)That the model is ill-fitting
B)That the model used is incorrect
C)That the model is better-fitting
D)None of the above

Question 17

A fixed effect is when all possible treatment conditions that you are interested in as a researcher are present in the experiment.

Question 18

Given that different teachers have different personalities and ways of teaching their students,the classroom in which an experiment is conducted is what type of variable?

A)Standard variable
B)Random variable
C)Contextual variable
D)Simple variable

Question 19

The overall fit of a multilevel model can be tested using a chi-square likelihood ratio test.

Question 20

Which of the following is an advantage to multilevel models over linear models?

A)The assumption of independence.
B)There is no need to run ANCOVAs.
C)They do not require complete data sets.
D)All of the above.

Question 21

The participants in multilevel models are usually at what level of analysis?

A)Low
B)High
C)It makes no difference.
D)Random levels

Question 22

A model in which intercepts can vary,so individual observation scores on the dependent variable can be predicted by the intercept that varies across groups,is called what?

A)Random slopes model
B)Random intercepts model
C)Developing slopes model
D)Developing intercepts model

Question 23

Homoscedasticity is an assumption of multilevel modelling.What is assumed?

A)That there is a rectilinear relationship between variables
B)That the error terms of each level of the model are normally distributed
C)Independence
D)Equality of population variances

Question 24

In SPSS,how are growth curves entered?

A)Manually
B)Automatically
C)It is not possible to put growth curves in SPSS.
D)In secondary analysis

Question 25

Homoscedasticity is also known as what?

A)Linearity
B)Normality
C)Effect size
D)Homogeneity of variance