Political Science

Homework 4
1. Use the dataset: For Homework4.sav. This hypothetical dataset is based on published research pertaining to the fallibility of eyewitness testimony. For this type of research, participants are shown a video of a collision between two cars. After the participant watches the video, some time elapses and then the participant is asked to estimate how fast the car was going when it collided with the other car. For example, “How fast was the car going when it tapped the other car?”. Participants’ estimation of speed is influenced by which word is used to describe the collision (i.e., tap, bumped, hit, smashed, crashed). For this example, suppose each participant was given videos to watch under each of the descriptor words for the collision. Thus, the study involves repeated measurements from the same individuals.
a. Conduct a repeated measures ANOVA. Label the repeated measures independent variable “type_of_verb”. Provide SPSS output showing the significance level for the within participants IV. (1 point)
b. State the null and alternative hypotheses for the main effect of type of verb on participants’ estimates of car speeds. (1 point)
c. For this particular dataset, do the results suggest that type of verb has a small, medium, or large effect on participants’ estimates of car speed? Be sure to justify your response by mentioning the effect size for the main effect. (1 point).
d. Produce a line graph that shows each of the means. Provide SPSS output showing for this line graph. (1 Point).
e. Does the line graph in 1d suggest that there are significant differences between each of the means? Justify your answer by explaining which means seem to be different and which means do not seem to be different.
f. Conduct pairwise comparisons between each of the means using the Bonferroni correct for family-wise error rate. Provide SPSS output showing the pairwise contrasts. (1 point).
g. Which of the means are significantly different from each other based on the output provided in 1f. (1 point)
h. Suppose you wished to conduct this study again. Using the effect size based on the current dataset, what would the sample size have to be to conduct this study again with power = .80? Provide a picture of the G*Power results with the number of participants required circled in that picture. (1 point)
i. Suppose you wished to conduct the study with experienced highway patrol officers. You decide that the effect size will probably be small. Assuming a d value of .28, what would the sample size need to be to obtain power at .80?
i. Show the calculation of f using the formula: d = 2*f (1 point)
ii. Provide a picture of the G*Power results with the number of participants required circled in that picture. (1 point)
j. Suppose the effects of type of verb occurs because of weaknesses in participants working memory; those with poorer working memory are unable to remember how fast the car was travelling due to weak representations of the viewed events. To explore this possibility, the researcher decides to use working memory skills as measured by the digit span to control for the effects of individual differences in working memory on the effects of type of verb.
i. Provide an SPSS printout that shows the within subjects effects (i.e., main effect for type of verb and interaction between the main effect of type of verb and working memory). (1 point)
ii. State the null and alternative hypotheses for the main effect of type of verb when the covariate of working memory is included in the analysis. (1 point)
iii. Report the F-test for the main effects of type of verb while the covariate is in the model. Use the standard way of reporting the F-test result (this includes the interpretation of the effect). (1 point).
iv. Did the effect of type of verb depend on working memory? Justify your answer using the appropriate F-test. (1 point)