Saying Becomes Believing: Consider Emily. You do not know her, but you read a brief summary about her (focusing on her personality and likes), and you now need to describe Emily to Helen. Helen does know Emily, and Helen really dislikes Emily. Knowing that Helen dislikes Emily, how do you describe Emily? More importantly, how do you feel about Emily?

There is a lot of research on the attitudebehavior relationship. While some research shows that attitudes can lead to behavior, other studies show that it can go in the other direction. Rather than forming an attitude first and then seeing how it impacts behavior, people behave first and then form an attitude based on their behavior. Think about Emily again. Although you may not know Emily, knowing that you will describe her to a person who does not like her might impact how you describe her. Thus, you might alter your description (and thus alter your behavior) to please your audience. If describing Emily to a person who dislikes Emily, your description may be negative. Given your negative description, you may form a negative attitude about Emily yourself, even if you had no attitude about Emily prior to providing the description.

Imagine you do a study looking at the "Emily" study. You ask participants to read a personality description of Emily, which is rather bland, boring, and broad. You then have them summarize that information and present it to "Helen", though you tell the participants that Helen either likes, dislikes, or has no opinion about Emily. After describing Emily to Helen, participants then rate their own feelings about Emily.

Attitude Condition: Helen likes Emily vs. Helen dislikes Emily vs. Helen has no opinion about Emily.

Measures: First, you code the description that participants write about Emily, breaking them into three categories: "Positive Description", "Negative Description", or "Neutral Description". Second, you ask participants rate to their own feelings about Emily ("How likeable is Emily?") using a scale ranging from 1 (Emily seems very unlikeable) to 7 (Emily seems very likeable).

Predictions: You first predict that participants will write a positive description of Emily if they think Helen likes Emily, that participants will write a negative description of Emily if they think Helen dislikes Emily, and that participants will write a neutral description of Emily if they think Helen has no opinion of Emily. You further predict that participants will develop attitudes about Emily that are consistent with Helen's "attitude". That is, participants will rate Emily as more likeable if they thought Helen liked Emily and less likeable if they thought Helen disliked Emily, with participant likeability ratings falling in the middle when Helen is thought to have no opinion about Emily.

1). What is the main independent variable in this study, and how many levels are there for this independent variable? Choose the BEST option (.5 points)

A. IV: Participant Descriptions of Emily, two levels (Positive Description, Negative Description)

B. IV: Participant Feelings about Emily (1 = Emily seems very unlikeable to 7 = Emily seems very likeable)

C. IV: Attitude Condition, with two levels (Helen likes Emily, Helen dislikes Emily)

D. IV: Attitude Condition, with three levels (Helen likes Emily, Helen dislikes Emily, Helen has no opinion about Emily)

2). What are the dependent variables in this study, and what scale of measurement are they based on (using NOIR)? Choose the BEST option (.5 points)

A. DV #1: Participant descriptions of Emily: Nominal scale - DV #2: Participant ratings of Emily's likeability: Nominal scale

B. DV #1: Participant descriptions of Helen: Nominal scale - DV #2: Participant ratings of Helen's likeability: Interval scale

C. DV #1: Participant descriptions of Emily: Nominal scale - DV #2: Participant ratings of Emily's likeability: Interval scale

D. DV #1: Participant ratings of Emily's likeability: Nominal scale - DV #2: Participant descriptions of Emily: Interval scale

3). You are going to run some data analyses using the DataAnalysisFIU#1EmilyFall.sav SPSS file in Canvas. Use your independent variable and your nominal dependent variable in the SPSS analysis. (Hint: Your scale of measurement for the nominal dependent variable should let you know which specific statistical test to use!). After running the test, choose the correct analysis, write-up, and conclusion from the options below (1.5 points)

A. We ran a chi square using Attitude Condition as the independent variable (Helen likes Emily, Helen dislikes Emily, Helen has no opinion about Emily) and participants' description of Emily as the dependent variable. A significant effect emerged, ²(2) = 137.49, p < .001. As expected, 33 out of 40 participants (or 82.5%) in the "Helen likes Emily" condition wrote a positive description of Emily. Similarly, 32 out of 40 participants (or 80%) in the "Helen dislikes Emily" condition wrote a negative description of Emily. Finally, 34 out of 40 participants (or 85%) in the "Helen has no opinion about Emily" condition wrote a neutral description of Emily. Cramer's V, which is appropriate for this design, was very strong. Participants seemed to adapt their descriptions of Emily to align with whether Helen liked, disliked, or had no opinion about Emily.

B. We ran a chi square using Attitude Condition as the independent variable (Helen likes Emily, Helen dislikes Emily, Helen has no opinion about Emily) and participants' description of Emily as the dependent variable. A significant effect emerged, ²(6) = 137.49, p < .001. As expected, 33 out of 40 participants (or 82.5%) in the "Helen likes Emily" condition wrote a positive description of Emily. Similarly, 32 out of 40 participants (or 80%) in the "Helen dislikes Emily" condition wrote a negative description of Emily. Finally, 34 out of 40 participants (or 85%) in the "Helen has no opinion about Emily" condition wrote a neutral description of Emily. Cramer's V, which is appropriate for this design, was very strong. Participants seemed to adapt their descriptions of Emily to align with whether Helen liked, disliked, or had no opinion about Emily.

C. We ran a chi square using Attitude Condition as the independent variable (Helen likes Emily, Helen dislikes Emily, Helen has no opinion about Emily) and participants' description of Emily as the dependent variable. A significant effect did not emerge, ²(6) = 137.49, p =.10. There was no difference in Emily descriptions between the three conditions, indicating that Helen's opinion of Emily did not impact how participants described Emily. Cramer's V, which is appropriate for this design, was very weak. Participants did not adapt their descriptions of Emily to align with whether Helen liked, disliked, or had no opinion of Emily.

D. We ran a One-Way ANOVA using Helen Attitude Condition as the independent variable (Helen likes Emily, Helen dislikes Emily, Halen has no opinion about Emily) and participants' description of Emily as the dependent variable. A significant effect emerged, F(2, 117) = 57.10, p < .001. Tukey post hoc tests showed that participants had fewer positive descriptions of Emily in the "Helen likes Emily" condition (M = 1.35, SD = 0.80) than in the "Helen dislikes Emily" condition (M = 2.10, SD = 0.44) and the "Helen no opinion about Emily" condition (M = 2.80, SD = 0.52), with more positive descriptions in the "no opinion" condition than in the "dislikes" condition. This indicates that participants actually had more negative opinions of Emily when Helen liked Emily, which is contrary to the predictions.

4). For the main analysis, you predicted that participants would develop attitudes about Emily that were consistent with Helen's "attitude". That is, participants would find Emily more likeable if they thought Helen liked Emily and less likeable if they thought Helen disliked Emily, with participant likeability ratings falling in the middle when Helen was thought to have no opinion about Emily. Run the correct analysis to see if you confirmed your predictions, and choose the correct conclusion from the options below (1.5 points)

A. We ran a One-Way ANOVA using Attitude Condition as the independent variable (Helen likes Emily, Helen dislikes Emily, Helen has no opinion about Emily) and participants' ratings of whether Emily is likeable as the dependent variable. A significant effect emerged, F(2, 119) = 11.03, p < .001. In support of the hypotheses, Tukey post hoc tests showed that participants rated Emily as more likeable in the "Helen likes Emily" condition (M = 5.00, SD = 1.04) than in both the "Helen has no opinion about Emily" condition (M = 4.47, SD = 0.93) and the "Helen dislikes Emily" condition (M = 3.95, SD = 0.88), with participants also liking Emily more in the "Helen has no opinion about Emily" condition than in the "Helen dislikes Emily" condition.

B. We ran a One-Way ANOVA using Attitude Condition as the independent variable (Helen likes Emily, Helen dislikes Emily, Helen has no opinion about Emily) and participants' ratings of whether Emily is likeable as the dependent variable. A significant effect emerged, F(2, 117) = 12.18, p < .001. In support of the hypotheses, Tukey post hoc tests showed that participants rated Emily as more likeable in the "Helen likes Emily" condition (M = 5.00, SD = 1.04) than in both the "Helen has no opinion about Emily" condition (M = 4.47, SD = 0.93) and the "Helen dislikes Emily" condition (M = 3.95, SD = 0.88), with participants also liking Emily more in the "Helen has no opinion about Emily" condition than in the "Helen dislikes Emily" condition.

C. We ran a One-Way ANOVA using Attitude Condition as the independent variable (Helen likes Emily, Helen dislikes Emily, Helen has no opinion about Emily) and participants' ratings of whether Emily is likeable as the dependent variable. A significant effect emerged, F(2, 117) = 12.18, p < .001. Contrary to the hypotheses, Tukey post hoc tests showed that participants rated Emily as less likeable in the "Helen dislikes Emily" condition (M = 3.95, SD = 1.04) than in both the "Helen has no opinion about Emily" condition (M = 4.47, SD = 0.93) and the "Helen likes Emily" condition (M = 5.00, SD = 0.88), with participants also liking Emily less in the "Helen has no opinion about Emily" condition than in the "Helen likes Emily" condition.

D. We ran a One-Way ANOVA using Attitude Condition as the independent variable (Helen likes Emily, Helen dislikes Emily, Helen has no opinion about Emily) and participants' ratings of whether Emily is likeable as the dependent variable. A significant effect did not emerge, F(2, 117) = 12.18, p = .19. Contrary to the hypotheses, participants did not differ in their likeability ratings of Emily in the "Helen likes Emily" condition (M = 5.00, SD = 1.04), the "Helen has no opinion about Emily" condition (M = 4.47, SD = 0.93), and the "Helen dislikes Emily" condition (M = 3.95, SD = 0.88).

Imagine we alter the design a bit. First, in terms of the attitude condition, we retain only the "Helen dislikes Emily" and the "Helen likes Emily" conditions (dropping the "No opinion" category). Second, while all participants read the same bland summary of Emily, for half of the participants, we stop the study right after they write their summary about Emily but before they rate Emily's likeability. We tell these participants that we made a mistake about whether Helen liked Emily and we wanted to correct that statement. Thus, participants are assigned to either a "mistake" or "no mistake" condition. This results in four conditions: 1). Participants are told that Helen likes Emily with no mistake. 2) Participants are told that Helen dislikes Emily with no mistake. 3). Participants are told that Helen likes Emily, but this is a mistake and "Actually, Helen dislikes Emily". 4). Participants are told that Helen dislikes Emily, but this is a mistake and "Actually, Helen likes Emily". The dependent variables remain the same, though we add a new manipulation check question that asks participants about Helen's true feelings about Emily ("Does Helen really like Emily?" Yes versus No). Using this new design, answer the following questions.

5). What is/are the independent variable(s) in this study, and how many levels are there to each? (.5 points)

A. IV#1: Attitude Condition, two levels (Helen likes Emily versus Helen dislikes Emily) - IV #2: Whether participants rated Emily as likeable, two levels (Yes versus No).

B. IV#1: Attitude Condition, two levels (Helen likes Emily versus Helen has no opinion about Emily) - IV #2: Mistake condition, two levels (Mistake versus No Mistake).

C. IV#1: Attitude Condition, four levels (Helen likes Emily versus Helen dislikes Emily versus Emily likes Helen versus Emily dislikes Helen) - IV #2: Mistake Condition, two levels (Mistake versus No mistake).

D. IV#1: Attitude Condition, two levels (Helen likes Emily versus Helen dislikes Emily)- IV #2: Mistake condition, two levels (Mistake versus No mistake).

6). Consider all of the possible main effects and interactions for this study. Choose the option below that best describes the outcome. (.5 points)

A. There are two significant main effects and a significant interaction

B. There is one significant main effect, one non-significant main effect, and a significant interaction

C. There is one significant main effect, one non-significant main effect, and no significant interaction

D. There are two significant main effects but there is no significant interaction