Exam 2 - 1 of 5 Exam 2 PS 217, Fall 2010 1. Liljenquist et al. (2010) published a paper (The smell of virtue: Clean scents promote reciprocity and charity) that included a study described as follows: Experiment 1 tested the impact of clean scents on reciprocation of trust. Twenty-eight participants were individually assigned to either a clean-scented room or a baseline (regular-smelling) room. The only difference between these rooms was a spray of citrus-scented Windex in the clean-scented rooms. In both conditions, participants engaged in a one-shot anonymous trust game involving two parties: a sender and a receiver. In a typical trust game, the sender is given money that he or she can either keep or “invest” with an anonymous receiver. Any money sent is tripled, and the receiver then decides how to split the tripled money. For example, if the sender invests all of the money and the receiver reciprocates this trust by returning half of the tripled amount, both parties would be better off. However, sending money can be risky if the receiver chooses to exploit the sender and keep all the invested money. All participants in Experiment 1 were told that they had been randomly assigned to play the role of the receiver and that their ostensible counterpart had decided to send them the full amount ($4), which was tripled to $12. They had to decide how much money to return to the sender. Participants could exploit the sender by keeping all the money, or they could honor the trust by returning some portion to the sender. Below are the data from the experiment. Analyze the data as completely as you can (e.g., creating a source table for the ANOVA) and interpret the results. (You have sufficient information below, but to make your computations easier, I’ll also tell you that FObtained = 7.44.) [15 pts] Baseline (Regular) Room Clean-Scented Room M $2.81 $5.33 s2 7.90 4.04 n 14 14 H0: µB = µC H1: Not H0 Source SS df MS F Between 44.45 1 44.45 7.44 Within 155.22 26 5.97 Total 199.67 27 !!!"#$""%=!". !"!!"+!". !"!!"−!!". !"!!"= !!. !" !!"#=!. !!. !"= !. !" and FMax Critical is approximately equal to 3.0, so there’s no concern about violating the homogeneity of variance assumption and you’d proceed with α = .05. Thus, FCritical(1,26) = 4.22. Decision: Reject H0, because FObtained ≥ FCritical. Conclusion: People in the clean-scented room sent back significantly more money (M = $5.33) than those in the regular room (M = $2.81). [N.B. No need for post hoc test, because there are only two conditions.] !!=!!. !"!"". !"=. !!Exam 2 - 2 of 5 2. Not only are repeated measures designs more powerful than independent groups designs, they are also more efficient (“more bang for the buck”). Suppose that you have a study with four conditions (k = 4). If you want to have at least 20 scores per condition, how many people would you need in a repeated measures study? In an independent groups study? [3 pts] In an independent groups design, the total number of subjects would be 4 x 20 = 80. In a repeated measures design, you’d want to completely counterbalance, which would lead to 4! = 24 orders and therefore 24 participants. 3. Power is an important concept, with people like Jacob Cohen suggesting that we should conduct experiments with power of at least .80. What are they saying about the level of Type II error that they are willing to tolerate? [2 pts] Power and Type II error are complimentary, so the probability would be .20. 4. Recently Simone Schnall and her colleagues published a paper in Psychological Science: Schnall, S., Roper, J., & Fessler, D. M. T. (2010). Elevation leads to altruistic behavior. From their abstract: Feelings of elevation, elicited by witnessing another person perform a good deed, have been hypothesized to motivate a desire to help others. However, despite growing interest in the determinants of prosocial behavior, there is only limited evidence that elevation leads to increases in altruistic behavior…Feelings of elevation, but not feelings of amusement or happiness, predicted the amount of helping. Together, these results provide evidence that witnessing another person’s altruistic behavior elicits elevation, a discrete emotion that, in turn, leads to tangible increases in altruism. From their Procedure section: Participants were informed that they were taking part in a 1-hr experiment on episodic memory in which they would watch a film clip, write about it, and complete a 30-min computer task. Tested individually, participants were randomly assigned to watch the elevation film clip from the Oprah Winfrey Show (elevation condition), the control film clip (the first 7 min of “The Open Ocean,” David Attenborough’s (1984) nature documentary describing a journey through the deepest part of the ocean), or a clip from a British comedy (“Fawlty Towers”) intended to induce mirth (mirth condition). The experimenter then feigned three unsuccessful attempts to open the computer file that ostensibly needed to be completed by the participant. She then told the participant that, because it was impossible to complete the next part of the study, the participant was free to leave, but would still receive the full hour’s worth of course credit. Following the procedure outlined in Bartlett and DeSteno (2006), when the participant got up to leave, the experimenter asked, apparently as an afterthought, whether she would be willing to complete another questionnaire, ostensibly from another study for which the experimenter needed to establish norms. The experimenter noted that the questionnaire was, unfortunately, rather boring, emphasizing that the participant was under no obligation, and was free to stop whenever she wanted, but that completing any number of the items would greatly assist the experimenter. If the participant agreed to help, she was seated at a desk, reminded that she was free to stop whenever she wished, and given 85 elementary math problems. The participant’s work on the problems was secretly timed (the dependent variable in the experiment, time spent on the task). The participant was then probed for suspicions regarding the purpose of the study and debriefed.Exam 2 - 3 of 5 The results from the study were analyzed as illustrated in the incomplete source table below. First, complete the source table below, then analyze the results as completely as you can. [15 pts] ANOVA Time Sum of Squares df Mean Square F Sig.