Question 1

If we were to run a four-way between-groups ANOVA, how many sources of variance would there be?

Accepted Answer

A) 16 
B) 4 
C) 8 
D) 15 
A) 16 
B) 4 
C) 8 
D) 15

Question 2

A new product tester (working for 'Fizzy Sugary Delight') wanted to test and compare the consumer rating of two new fizzy drinks; his own company's drink and a rival's drink. He randomly selected 12 participants and then randomly divided this group into three subgroups: a 'Placebo' group who drank fizzy sugary water; a 'new Improved Formula' group who drank the new Fizzy Sugary Delight drink; and a 'Rival fizz' group who drank the new fizzy drink belonging to a rival company. Each group consumed one glass each of their designated drink and then rated its taste. What sort of research design is this?

Accepted Answer

A) Independent factorial design. 
B) Repeated-measures factorial design. 
C) ANOVA. 
D) Multiple linear regression. 
A) Independent factorial design. 
B) Repeated-measures factorial design. 
C) ANOVA. 
D) Multiple linear regression.

Question 3

What would be the point of bootstrapping post hoc tests in a factorial design?

Accepted Answer

A) To make the post hoc tests more robust particularly against potential bias caused by heteroscedasticity. 
B) To make the post hoc tests more robust particularly against potential bias caused by homoscedasticity. 
C) To make the post hoc tests less robust particularly against potential bias caused by heteroscedasticity. 
D) To make the post hoc tests less powerful particularly against potential bias caused by homoscedasticity. 
A) To make the post hoc tests more robust particularly against potential bias caused by heteroscedasticity. 
B) To make the post hoc tests more robust particularly against potential bias caused by homoscedasticity. 
C) To make the post hoc tests less robust particularly against potential bias caused by heteroscedasticity. 
D) To make the post hoc tests less powerful particularly against potential bias caused by homoscedasticity.

Question 4

How can you resolve the problem identified in the previous question?

Accepted Answer

A) By manually creating dummy variables in SPSS. 
B) By automatically creating dummy variables in SPSS. 
C) By manually calculating Bayes factors. 
D) By manually bootstrapping variables in SPSS. 
A) By manually creating dummy variables in SPSS. 
B) By automatically creating dummy variables in SPSS. 
C) By manually calculating Bayes factors. 
D) By manually bootstrapping variables in SPSS.

Question 5

Which of the following is calculated or presented in the same way for both repeated-measures and between-subjects designs?

Accepted Answer

A) The simple effects 
B) The model error 
C) The degrees of freedom 
D) Partial eta squared 
A) The simple effects 
B) The model error 
C) The degrees of freedom 
D) Partial eta squared

Question 6

A sports researcher wanted to test and compare the impact of three different energy drinks as a means to examine to what extent energy drinks consumed during a workout increases workout rates. He randomly selected 36 gym goers and then randomly divided this group into three subgroups: a 'Placebo' group who drank water; a 'Caffeine' group who took a caffeine-rich energy drink; and a 'Sugar' group who took a sugar-rich energy drink. The study involved the groups drinking their energy drink prior to a thirty-minute workout. At the end or their workouts, their workout rates were recorded. What sort of research design is this?

Accepted Answer

A) Repeated-measures factorial design. 
B) Independent factorial design. 
C) ANOVA. 
D) Multiple linear regression. 
A) Repeated-measures factorial design. 
B) Independent factorial design. 
C) ANOVA. 
D) Multiple linear regression.

Question 7

What do non-parallel lines on an interaction graph indicate?

Accepted Answer

A) Some degree of interaction between the variables and a definitive interaction effect. 
B) No interaction between the variables and a therefore no interaction effect. 
C) Some degree of interaction between the variables and a possible interaction effect. 
D) No interaction between the variables but a possible interaction effect. 
A) Some degree of interaction between the variables and a definitive interaction effect. 
B) No interaction between the variables and a therefore no interaction effect. 
C) Some degree of interaction between the variables and a possible interaction effect. 
D) No interaction between the variables but a possible interaction effect.

Question 8

Two-way ANOVA is basically the same as one-way ANOVA, except that:

Accepted Answer

A) We calculate the model sum of squares by looking at the difference between each group mean and the overall mean. 
B) The model sum of squares is partitioned into three parts. 
C) The model sum of squares is partitioned into two parts. 
D) The residual sum of squares represents individual differences in performance. 
A) We calculate the model sum of squares by looking at the difference between each group mean and the overall mean. 
B) The model sum of squares is partitioned into three parts. 
C) The model sum of squares is partitioned into two parts. 
D) The residual sum of squares represents individual differences in performance.

Question 9

Why might it be useful to use a stricter criterion than 0.05 for accepting an F statistic as significant in a factorial design?

Accepted Answer

A) Because the F statistics are independent and thus do control the Type I error rate. 
B) Because the F statistics are independent and thus do not control the Type I error rate. 
C) Because the F statistics are not independent and thus do not control the Type I error rate. 
D) Because the F statistics are not independent and thus do control the Type I error rate. 
A) Because the F statistics are independent and thus do control the Type I error rate. 
B) Because the F statistics are independent and thus do not control the Type I error rate. 
C) Because the F statistics are not independent and thus do not control the Type I error rate. 
D) Because the F statistics are not independent and thus do control the Type I error rate.

Question 10

How can you counter the problem of heteroscedasticity in factorial designs?

Accepted Answer

A) It is impossible to counter this problem. 
B) Run robust models using parameter estimates with robust standard errors. 
C) Run post hoc tests. 
D) Run tests for significance. 
A) It is impossible to counter this problem. 
B) Run robust models using parameter estimates with robust standard errors. 
C) Run post hoc tests. 
D) Run tests for significance.

If we were to run a four-way between-groups ANOVA, how many sources of variance would there be?

What would be the point of bootstrapping post hoc tests in a factorial design?

How can you resolve the problem identified in the previous question?

Which of the following is calculated or presented in the same way for both repeated-measures and between-subjects designs?

What do non-parallel lines on an interaction graph indicate?

Two-way ANOVA is basically the same as one-way ANOVA, except that:

Why might it be useful to use a stricter criterion than 0.05 for accepting an F statistic as significant in a factorial design?

How can you counter the problem of heteroscedasticity in factorial designs?

Why Is My Evil Lecturer Forcing Me to Learn Statistics

The Spine of Statistics

The Phoenix of Statistics

The IBM Spss Statistics Environment

Exploring Data With Graphs

The Beast of Bias

Non-Parametric Models

Correlation

Regression

Comparing Two Means

Moderation, Mediation and More Regression

GlM 1: Comparing Several Independent Means

GlM 2: Comparing Means Adjusted for Other Predictors Analysis of Covariance

GlM 4: Repeated-Measures Designs

GlM 5: Mixed Designs

Multivariate Analysis of Variance Manova

Exploratory Factor Analysis

Categorical Outcomes: Chi-Square and Loglinear Analysis

Categorical Outcomes: Logistic Regression

Multilevel Linear Models

Filters

Exam 14: GlM 3: Factorial Designs

If we were to run a four-way between-groups ANOVA, how many sources of variance would there be?

What would be the point of bootstrapping post hoc tests in a factorial design?

How can you resolve the problem identified in the previous question?

Which of the following is calculated or presented in the same way for both repeated-measures and between-subjects designs?

What do non-parallel lines on an interaction graph indicate?

Two-way ANOVA is basically the same as one-way ANOVA, except that:

Why might it be useful to use a stricter criterion than 0.05 for accepting an F statistic as significant in a factorial design?

How can you counter the problem of heteroscedasticity in factorial designs?

Why Is My Evil Lecturer Forcing Me to Learn Statistics

The Spine of Statistics

The Phoenix of Statistics

The IBM Spss Statistics Environment

Exploring Data With Graphs

The Beast of Bias

Non-Parametric Models

Correlation

Regression

Comparing Two Means

Moderation, Mediation and More Regression

GlM 1: Comparing Several Independent Means

GlM 2: Comparing Means Adjusted for Other Predictors Analysis of Covariance

GlM 4: Repeated-Measures Designs

GlM 5: Mixed Designs

Multivariate Analysis of Variance Manova

Exploratory Factor Analysis

Categorical Outcomes: Chi-Square and Loglinear Analysis

Categorical Outcomes: Logistic Regression

Multilevel Linear Models

Filters