Deck 9: Ancova

Question 1

When do you use an analysis of covariance (ANCOVA)?

A) When you compare means from two conditions
B) When you analyse experimental data while controlling for a measured confounding variable
C) When you compare means from more than two conditions
D) When you assess a correlation between two variables

Answer 1

B

Question 2

What is a confounding variable?

A) A variable the experimenter manipulates
B) A variable being tested and measured in an experiment
C) An independent variable
D) An unmeasured variable that might affect the dependent variable

Answer 2

D

Question 3

Which of the following statements is not correct?

A) Including a confounding variable will change the analysis by decreasing the significance if the confound explains mainly random variance.
B) Including a confounding variable will change the analysis by increasing the significance if the confound explains mainly random variance.
C) Including a confounding variable will change the analysis by very little or not at all if the confound explains both experimental and random variance.
D) Including a confounding variable will not change the analysis if the confound explains mainly random variance.

Answer 3

B

Question 4

How can you remove any effects of the control variable in the descriptive statistics while interpreting ANCOVA to have an accurate reflection of the findings?

A) Using the estimated marginal means
B) Using the corrected/adjusted p value
C) Using the adjusted medians
D) Using the unadjusted means

Question 5

Which steps do you need to take to run a one-way independent measures ANCOVA in SPSS?

A) Analyze > Compare Means > Univariate…
B) Analyze > Compare Means > Multivariate…
C) Analyze > General Linear Model > Multivariate…
D) Analyze > General Linear Model > Univariate…

Question 6

How do you get the effect sizes in SPSS?

A) Go into the "Options" box and select the "Observed power" when running the ANCOVA.
B) Go into the "Post Hoc" box and select the "Estimates of effect size" when running the ANCOVA.
C) Go into the "Options" box and select the "Estimates of effect size" when running the ANCOVA.
D) Go into the "Options" box and select the "Parameter estimates" when running the ANCOVA.

Question 7

You must use either planned or post hoc contrasts to determine which groups are different if you find a statistically significant ANCOVA. What determines whether you use planned or post hoc contrasts?

A) Whether the hypothesis is one-tailed (directional) or two-tailed (no direction)
B) Whether the data is normally distributed
C) The sample size
D) Whether the assumption of homogeneity of variance has been met or not

Question 8

Which of the following statistics do you not need to report when writing up a one-way independent measures ANCOVA?

A) Levene's test of homogeneity of variance
B) The effect of the co-variate
C) Mann-Whitney U test statistic
D) The main effect and the planned or post hoc contrasts if the main effect is significant

Question 9

What is Dummy Variable Modelling used for?

A) To transform a non-normal set of data into a normal distribution
B) To remove outliers
C) To include continues variables as a covariate in an ANCOVA
D) To include variables with three of more categories as a covariate in an ANCOVA

Question 10

What is the non-parametric equivalent of ANCOVA in SPSS?

A) There is not a non-parametric equivalent of ANCOVA in SPSS.
B) Mann-Whitney test
C) Kruskal-Wallis test
D) Chi-squared test

Question 11

In the worksheet exercise, what was the significance of Levene's test of homogeneity of variance?

A) p = 2.661
B) p = .052
C) p = .048
D) p = .000

Question 12

In the worksheet exercise, what is the F ratio for the confounding variable?

A) 7.236
B) 56.956
C) 15.633
D) 2.737

Question 13

In the worksheet exercise, which of the following is the best interpretation of the main effect?

A) Before controlling for the confound, the main effect is not significant.
B) After controlling for the confound, the main effect is not significant.
C) Before controlling for the confound, the main effect is significant.
D) After controlling for the confound, the main effect is significant.

Question 14

In the worksheet exercise, what is the estimated marginal mean for the "low" condition?

A) 12.5161
B) 12.008
C) 10.509
D) 8.572

Question 15

In the worksheet exercise, looking at the adjusted statistics, in which condition is there the largest amount of dispersion?

A) Frequently
B) Sometimes
C) Low
D) Never