Quantitative Methods 3 (31684)

Spring 2007

Exam 3

This examination is due on Monday May7thby 12:30pm. Drop off your exams with either myself or the secretary of the Psychological Clinic, Room 176, Kent Hall. NO extensions will be considered without a written note from a physician or a similar administrative authority.

YOU MAY NOT COLLABORATE WITH ANYONE ELSE IN ORDER TO COMPLETE THIS EXAM. This rule includes both individuals taking this class and those who are not. Those caught collaborating on this exam will be subject to the University’s policy regarding cheating. Remember: I reserve the right to change exams to an in-class, closed-book, and closed-note format if students collaborate on exams.

This exam is comprised of 12 questions. Each of the 10 multiple choice questions are worth 1 pt. for a total of 10 pts. Each of the final 3 essay-type questions are worth30 pts for a total of 90 pts. The final question is extra credit. You cannot lose points for attempting this question. Answer the following questions as completely as possible using class notes, texts, and assigned readings. When you turn in this exam, you must submit PRINTED copies of your output and answers to exam questions with this exam assignment sheet stapled to the front of your answers. No electronic submissions will be considered for exams.

To complete this exam, you must use the “Exam 3 SPSS Data” file found on the course homepage at Refer to the following key in order to identify variables found within the data file.

drelat–dadjustScore on a variety of maladaptive behavioral outcomes

(16 items) of divorce

arelat-amain2Score on a variety of adaptive behavioral outcomes of

(13 items)divorce

incomeCurrent family income (7 levels: 0=”DTS”; 1=”<$10,000”; 2=”$10-20,000”; 3=”$20-30,000”; 4=”$30-40,000”; 5=”$40-50,000”; 6=”$60,000+”

lsatisfyPre-divorce life satisfaction

traumaSeverity of pre-divorce trauma suffered

cogcopePre-divorce use of adaptive cognitive coping techniques

behcopePre-divorce use of adaptive behavioral coping techniques

avoicopePre-divorce use of maladaptive avoidance coping techniques

locusPre-divorce locus of control-higher scores indicate a tendency to assign blame to others, lower scores indicate a tendency to take responsibility

spirituaPre-divorce degree of spirituality

socsuppPre-divorce degree of social support

Please note: the strength of your argument, as supported by the concepts we’ve discussed in class, is extremely important. Your answers should reflect an understanding of the fundamental concepts underlying the statistical procedures covered in the questions below.

Good luck! Questions begin on the back of this page.

  1. The correlation between two variables is a measure of the degree to which
  2. points cluster together around some best-fitting straight line
  3. differences in one variable can be predicted from differences in the other variable
  4. one variable varies with the other variable
  5. all of the above
  1. When we restrict the range of X or Y, we may
  2. lower the correlation from what it would otherwise be
  3. raise the correlation from what it might be
  4. leave the correlation the same as it would otherwise be
  5. all of the above are possible
  1. When we use heterogenous subsamples of data, such as older and younger subjects, the resulting correlation between intelligence and education could
  2. tell you more about the relationship between Age and Education than between Intelligence and Education
  3. be very misleading
  4. represent the relationship between Education and Intelligence accurately
  5. all of the above
  1. Which of the following represents a stronger relationship between two variables?
  2. r = .00
  3. r = .50
  4. r = -.30
  5. r = -.65
  1. If we find that the correlation between pessimism and depression is equal to r = .25, what can we conclude?
  2. 25% of the variance in pessimism is accounted for by depression
  3. 5% of the variance in pessimism is accounted for by depression
  4. 6.25% of the variance in pessimism is accounted for by depression
  5. pessimism is not reliably related to depression
  1. If we have a regression line predicting the amount of improvementin your performance as a function of the amount of tutoring your receive, an intercept of 12 would mean that
  2. You need to have 12 hours of tutoring to get an A
  3. If you don’t have any tutoring, the best you can do is a grade of 12
  4. Even without tutoring you will improve
  5. Tutoring helps
  1. In the equation
  2. A difference of one unit in X will lead to a 5 point difference in the prediction
  3. will decrease as X increases
  4. The correlation is certain to be significant
  5. A difference of one unit in X will lead to a 12.6 point difference in the prediction
  1. When we think in terms of standardized data, the slope represents
  2. The change in X for a one unit change in Y
  3. The number of standard deviations Y will differ for a one standard deviation difference in X
  4. The height of the regression line
  5. 0
  1. If data with only one predictor variable were standardized, the slope would equal
  2. r
  3. b
  4. a
  1. We can think of the standard error of the estimate as
  2. The standard deviation of the errors that we make when using the regression equation
  3. The standard deviation of Y
  4. The variance of the errors that we would make when using the regression equation
  5. The variance of X

Answer the following question using the exam data file and SPSS.

11.You are a researcher interested in a number of psychological and behavioral variables associated with recovery from divorce. In order to explore these relationships between variables, you have collected data-from 229 individuals who have recently divorced. You begin with a raw data set. You decide to explore how destructive behaviors (desbehav) and adaptive behaviors (adpbehav) relate to other relevant variables in your data set. Conduct bivariate correlations for desbehav and adpbehav on lsatisfy, trauma, cogcope, behcope, avoicope, locus, spiritua, and socsupp.Combine all of the various individual destructive behavior variables into one composite variable. Do the same for the individual adaptive behavior variables. Report the results of the correlations between desbehav and the other variables using APA style. Report the results of the correlations between adpbehav and the other variables generally. Identify any assumptions of correlation that have been violated and discuss what corrections, if any, should be made. Don’t forget to include and interpret relevant effect size statistics. (30 points)

12.You hypothesize that a large portion of the variance in the relationship between adpbehavand socsuppis accounted for by two other factors: income and cogcope.Run two separate analyses to evaluate whether the relationship between adpbehavand socsuppexists when income and cogcope are statistically controlled for. Identify any assumptions of correlation that have been violated and discuss what corrections, if any, should be made. Don’t forget to include and interpret relevant effect size statistics.(30 points)

13.You want to predict desbehav using lsatisfy, cogcope, behcope, avoicope, locus, spiritua, and socsupp because you believe that a variety of coping and psychosocial variables impact destructive behavior in response to divorce. Conduct the analysis. Which type of variable entry should you probably use and why? What other forms of variable entry might you have used? What are the results of your analysis? Describe the results in general, but be sure to answer the following questions: How much variance in the criterion variable does your model account for? Which variables are statistically predictive of the criterion? Identify any assumptions of linear regression that have been violated and discuss what corrections, if any, should be made. Don’t forget to include and interpret relevant effect size statistics. Report the results of your analysis using APA style. (30 points)

E.C.Conduct an a priori power analysis for either question 11, 12, or 13 (i.e. not all 3) using either G*Power or hand calculations. (5 points)