1. Which of the following cannot be achieved by using multivariate methods?
- relationships between variables are specific to a single context
- assessing whether relationships between study variables vary depending on the level of other variables
- proving causal relationships between variables
- assessing influences of predictor variables which vary from one dependent variable to another
2. What is a latent variable?
- a variable which cannot be measured directly
- a variable whose impact is hidden
- a variable which is left out of a causal model
- a predictor variable which is not significant
3. Which of the following is NOT required for a multiple regression model?
- more than one predictor variable
- a category dependent variable
- a set of observed variables
- a predicted causal model about relationships among variables
4. The relative importance of a predictor variable in regression analysis is shown by ______.
- a product-moment correlation coefficient
- a multiple R coefficient
- a regression coefficient
- a factor loading
5. In hierarchical regression analysis, ______.
- the researcher chooses the ordering of predictor variables to be entered based on theory
- the order of entering predictor variables depends on the size of their correlations with the dependent variable
- the order of entering variables is determined by how well each one adds to predicting the dependent variable
- the most important variable is entered first
6. What is the difference between ANOVA and MANOVA models?
- MANOVA models have more than one category predictor variable.
- MANOVA models have more than one dependent variable.
- MANOVA models always include covariates.
- MANOVA models are for multiple groups and ANOVA models are for one group.
7. What is the difference between multiple regression analysis and logistic regression analysis?
- Logistic regression analysis includes a category dependent variable.
- Multiple regression analysis includes more than one dependent variable.
- Logistic regression models always contain covariates.
- Multiple regression analysis allows predictors to be added sequentially.
8. Which of the following is NOT required in a structural equation model?
- a statement about which relevant variables to include
- a hypothesis about causeeffect relationships among variables
- fixed and free parameters
- both latent variables and observed variables
9. In the context of SEMs, what is an identified model?
- a model which fits the assumptions of the method
- a model with enough information to obtain unique estimates of the components of a hypothesised model
- a model which confirms the researchers hypotheses
- a model which is the simplest possible
10. What is goodness of fit?
- how well a model fits the researchers expectations
- how well the predictions from a model are consistent with empirical data
- the significance level from a statistical test
- how well empirical data conform to the technical assumptions of a statistical method