Exam 3 - 1 of 6 Final Exam PS 217, Fall 2009 1. What’s the difference between a parametric statistic (e.g., ANOVA, Pearsons’ product-moment correlation coefficient) and a nonparametric statistic (e.g., Spearman’s rho)? [1 pt] A nonparametric statistic makes fewer assumptions about the nature of the data. Thus, ordinal or nominal data would be interpretable with certain nonparametric statistics. Parametric statistics, however, may presume that the data are interval or ordinal. 2. Place, et al. (2008) wrote an article that appeared in Psychological Science titled: “The ability to judge the romantic interest of others.” The abstract read: The ability to judge another individual's romantic interest level—both toward oneself and toward others—is an adaptively important skill when choosing a suitable mate to pursue. We tested this ability using videos of individuals on speed dates as stimuli. Male and female observers were equally good at predicting interest levels, but they were more accurate when predicting male interest: Predictions of female interest were just above chance. Observers predicted interest successfully using stimuli as short as 10 s, and they performed best when watching clips of the middle or end of the speed date. There was considerable variability between daters, with some being very easy to read and others apparently masking their true intentions. Variability between observers was also found. The results suggest that the ability to read nonverbal behavior quickly in mate choice is present not only for individuals in the interaction, but also for third-party observers. To simplify the study and place it in a format that is consistent with your knowledge, let’s imagine the study as a 2x3 independent groups design. Forty-eight participants (24 male and 24 female) viewed a short video clip of a speed dating couple. They were shown only a portion of the speed dating interaction from the beginning, middle, or end of the interaction. Half the participants rated only the males in the interaction (Male’s interest in female) and the other half of the participants rated only the females in the interaction (Female’s interest in male). The dependent variable is accuracy in correctly judging the romantic interest of the target person in the video (on a scale of 0 = totally inaccurate to 1 = totally accurate). Below is a source table that analyzes data similar to those found by Place et al. Complete the source table and interpret the results as completely as you can. [15 pts]Exam 3 - 2 of 6 Tests of Between-Subjects Effects Dependent Variable:Accuracy Source Sum of Squares df Mean Square F Sig. Partial Eta Squared Noncent. Parameter Observed Powerb Gender .026 1 .026 26 .000 .406 28.739 .999 Time .136 2 .068 68 .000 .783 151.798 1.000 Gender * Time .010 2 .005 5 .006 .217 11.630 .846 Error .042 42 .001 Corrected Total .214 47 Homogeneity of Variance: Could either compute Hartley’s FMax (4.21) and compare to FMax Crit (10.8) and determine that there was no reason to be concerned about heterogeneity of variance (and use α = .05) OR note that all the Fs were significant with ps less than .01, so they would be significant even in the face of heterogeneity of variance. € HSD = 4.2.0018= .047 There was a significant main effect of Gender, F(1,42) = 26, MSE = .001, p < .001, η2 = .406. There was also a significant main effect of Time, F(2,42) = 68, p < .001, η2 = .783. Finally, there was a significant interaction of Gender and Time, F(2,42) = 5, p = .006, η2 = .217. Post hoc tests using Tukey’s HSD revealed that at the beginning and end of the clip, males were significantly more accurate (M = .550 and M = .660, respectively) than females (M = .498 and M = .581, respectively). However, at the middle of the clip, males (M = .651) and females (M = .644) did not differ significantly. 0.450.50.550.60.650.7Beginning Middle EndPlace et al. StudyAccuracyTime of ClipMale Interest in FemaleFemale Interest in MaleExam 3 - 3 of 6 3. In an article by Shackman et al. (2009) in Psychological Science entitled “Right dorsolateral prefrontal cortical activity and behavioral intervention,” the abstract reads: Individuals show marked variation in their responses to threat. Such individual differences in behavioral inhibition play a profound role in mental and physical well-being. Behavioral inhibition is thought to reflect variation in the sensitivity of a distributed neural system responsible for generating anxiety and organizing defensive responses to threat and punishment. Although progress has been made in identifying the key constituents of this behavioral inhibition system in humans, the involvement of dorsolateral prefrontal cortex (DLPFC) remains unclear. Here, we acquired self-reported Behavioral Inhibition System Sensitivity scores and high-resolution electroencephalography from a large sample. Using the enhanced spatial resolution afforded by source modeling techniques, we show that individuals with greater tonic (resting) activity in right-posterior DLPFC rate themselves as more behaviorally inhibited. This observation provides novel support for recent conceptualizations of behavioral inhibition and clues to the mechanisms that might underlie variation in threat-induced negative affect. From the PASW analysis below (of data that mimic those found in the Shackman et al. study), interpret the results as completely as you can. How many participants are in the study? If a person received a score of 20 on the Behavioral Inhibition Scale (BIS), what would you predict that person’s EEG asymmetry score to be? If a person received a score of 12 on the Behavioral Inhibition Scale (BIS), what would you predict that person’s EEG asymmetry score to be? What is the function of the standard error of estimate? [10 pts] There was a significant negative linear relationship between EEG and BIS, r(48) = -.448, p = .001. The sample size, n, was 50. The coefficient of determination, r2, was .201. The standard error of estimate, SEE, was .137. The SEE allows one to determine the accuracy of predictions, with a smaller SEE indicating that the points are closer to the line of best fit, so the accuracy would be higher. Using the regression equation (Y = -.023X + .427), one would predict that a BIS score of 20 would yield an EEG score of -.033. Because you didn’t observe a BIS score of 12, you could either decline to make a prediction,