Research Methods. Ch. 12. Experiments with more than 1 Independent Variable

Describe at least two cues indicating that a popular press article is probably describing a factorial design

) the phrase "it depends" or "only when" b) when articles discuss a participant variable (e.g., age, gender, or ethnicity)

What are two common reasons to use a factorial design?

1. Factorial designs can test limits; to test whether an independent variable effects different kinds of people, or people in different situations, the same way. 2. Factorial designs can test theories; can test generalizability of a causal variable and also test theories.

participant variable

A variable such as age, gender, or ethnicity whose levels are selected (i.e., measured), not manipulated. (IV)

What is a factor?

An independent variable

What question to you ask to see if there's a correlation?

If I remove this from the situation, does the situation still exist?

Marginal Means

In a factorial design, the arithmetic means for each level of an independent variable, averaging over the levels of another independent variable. inspect the main effects in a factorial design and they use statistics to find out whether the difference is statistically significant 551 559 difference is -11 562 552 difference is 7 -11<7. The differences are statistically significant and there is an interaction

Why was the word association study on alcohol and thoughts of aggression a factorial design?

It's a factorial design because they used all possible combinations of two independent variables

Using a factorial design to test limits is called testing for

Moderators. An independent variable that changes the relationship between another IV and the DV the effect of one IV depends on (is moderated by) the level of another IV driver age did not moderate the impact of cell phone use on braking onset time.

What is an example of a spreading interaction?

Probability that my dog will sit is the dependent variable. Independent variables are whether I have a treat or not. <

How can you detect an interaction from a table of means?

Start by computing two differences. Begin with one level of the first independent variable. Then go to the 2nd level of the first independent variable. Be sure to compute the difference in the same direction both times. If the differences are different, you can conclude there is an interaction in this factorial study.

Interaction Effect

The difference in the levels of one independent variable (cell phone use) changes depending on the level of the other independent variable (driver age). Doe the effect of cell phone use depend on age? A difference in differences Does the Effect hold while another IV is being changed?

When a study shows both a main effect and an interaction, what is more important?

The interaction is almost always more important

Why might it be better to call a main effect an overall effect?

The term main effect is usually misleading, because it seems to suggest that it is the most important effect in a study. It is not. The interaction itself is the most important effect.

Spreading Interaction

There is a spreading effect on the data based off the amount of IV *only when*. My dog sits when I say "Sit" but only when I have a treat there is zero difference between the treat and no treat condition When I say "Sit" there is a large difference between the treat and no treat condition

Crossover Interaction

There is no overall effect, but there is a cross over interaction. The effect of B depending on the dependent variable is opposite depending on the value of Factor A.

Mediation

Third variable that causes the correlation

Moderation

Third variable that influences the correlation

Factorial Variations

What if the IV has more than 2 levels? 3 IVs? One IV within groups? Independent groups factorial design- small people vs. big people within group- every subject experiences every condition mixed factorial- independent within groups design

Factorial Design

a study when there are more than 2 IVs/ factors You're looking for two factors with 2 levels (2x2 factorial design) You can use manipulated variables and/or participant variables May not always find an interaction They're a form of external validity to see if the effect is generalizable

In an empirical journal article, in what section will you find the independent and dependent variables of the design? In what section will you find whether the main effects and interactions are statistically significant?

a) Independent and dependent variables: Method section b) Main effects and interactions: Results section

6. Suppose these researchers ran their study again, but this time they compared how high or low self-esteem people responded to three kinds of statements: Positive self-statements, negative self-statements, and no statements. What kind of design would this be?

a. 2 x 2 b. 2 x 3 CORRECT ANSWER c. 2 x 2 x 2 d. 6 x 1

This study is an example of a(n)

a. Independent-groups factorial design CORRECT ANSWER b. Within-groups factorial design c. Mixed factorial design

4. Which of the following sentences describes the main effect for self-esteem in the Wood study?

a. Overall, the moods of high self-esteem people are more positive than the moods of low self-esteem people. CORRECT ANSWER b. Mood depended on both the participants' level of self-esteem and what self-statement condition they were in. c. Overall, the moods of people in the "I am a lovable person" condition are about the same as the moods of people in the "No statement" condition. d. Positive self-statements made the moods of high self-esteem people go up, but they made the moods of low self-esteem people go down.

2. What are the factors in this study?

a. Self-esteem level: high versus low b. Self-statement instructions: positive versus none c. Self-esteem level: high versus low; and self- statement instructions: positive versus none CORRECT ANSWER d. Positive mood

why might the Wood team have conducted their study as a factorial design?

a. To test how well positive self-statements work, compared to no statements. b. To compare the moods of high self-esteem and low self-esteem people. c. To test whether the effect of positive self- statements would depend on people's level of self-esteem. CORRECT ANSWER

3. There is an interaction in the results. How do you know?

a. When the cell means are graphed, the lines are not parallel. b. The difference between low and high self- esteem is large for the positive self- statement condition, and smaller for the neutral statement condition. c. Positive self-statements made the mood of high self-esteem people go up, but they made the mood of low self-esteem people go down. d. All of the above. CORRECT ANSWER

Do you like food hot or cold? is an example of what

an interaction. It depends. Two IV's food you are judging and the temperature

Factorial Designs can test

generalizability of causal variables and theories Many theories and hypotheses make predictions about how variables interact with each other

Main Effect

the *overall effect* of an IV on a DV, averaging over the levels of the other IV - a simple difference - is there a correlation or relationship? If we want to combined the two we look at the interaction but the main effects are analyzed independently

What happens to the marginal means if the sample sizes are equal

the marginal means are a simple average

What happens to the marginal means if the sample sizes are unequal

the marginal means will be computed using the weighted average, counting the larger sample more

To interpret an interaction you need to inspect

the two main effects the interaction