CPHS EXAM 1, S99

PART A (Closed Book)

This part of the exam is closed-book. It is worth 15 points and should take approximately 15 minutes to complete. When you are done with this part of the exam, please turn it in and pick up Part B.

[Comment: This exam covered Study Design, Epi6Basics, and CONT-SEV. The content of your exam may differ.]

(1) Briefly, describe the main difference between experimental study designs and observational study designs.
(2) Differentiate between a "variable" and "value."
(3) Differentiate between a longitudinal study a cross-sectional study.
(4) What does a computer operating system do?
(5) What is the purpose of data documentation files
(6) What type of information would your want to include in a data documentation (DD) file?
(7) Before a data set can be analyzed in the ANALYSIS module, it must be opened. What command is used for this purpose?
(8) DOS filenames are limited to an 8 character filename and 3 character ___________. (Fill in the blank.)
(9) With respect to ANOVA, when we assume homoscedasticity, the estimate of the common (within-group) variance has many names. Provide one.
(10) With respect to the statistic cited in question 9, above, where does one find this statistic in an ANOVA table. (Circle the correct boxed location.)
Source of Variance Sum of Squares Degrees of freedom Mean Squares F statistic p value
Between _ _ _ _ _
Within _ _ _
Total _ _ _

(11) How many degrees of freedom are in the numerator of an ANOVA F statistic? [The numerator is the Between-Groups Variance statistic.]
(12) What statistics contributed to the drawing of a quartile plot?

ANSWERS:

(1) Experimental studies entail an intervention ("experimental variable") on the part of the researcher. The intervention is usually randomized, to maximize group comparability. Observational studies have intervention; groups are taken as they are. Observed differences are then compared.
(2) A variable is the generic thing being measured (e.g., AGE). A value is a realized measurement (e.g., "31").
(3) Longitudinal studies entail follow-up of study subjects over time. Cross-sectional studies take measures of variables with respect to approximately the same point in time.
(4) Operating systems control the input / output of information to and from the computer's CPU and peripherals.
(5) Data documentation files allow the users to decipher a data file by providing information about its content, variable names, variable types, data structure, coding practices, etc.
(6) Filename, file type, number of records, variable names, variable types, variable lengths, coding practices, and history of file manipulations, etc.
(7) READ
(8) extensions.
(9) Synonyms for within-group variance: pooled estimate of variance, Mean Square Within, "Variance," Mean Square Error
(10) You should circle the MS Within box.
(11) df(between) = k - 1
(12) 5 point summary: minimum, 25th percentile (Q1), median, 75th percentile (Q3), maximum

PART B (Procedure Book)

SOCPSYCH.REC: A social psychologist develops a scale that he calls the "sense of coherence" scale. This scale measures a general orientation that sees life as meaningful and manageable. The psychologist theorizes that one's sense of cohorence fosters coping and health, and therefore deserves additional study. To study his theory, sense of cohorence (SOC) is measured among three groups. Group 1 comprises concentration camp survivors, group 2 comprises concentration camp guards, and group 3 comprises a sample of the general population. Data are:

Group 1: 131, 167, 113, 134, 178

Group 2: 105, 52, 71, 56, 85

Group 3: 236, 168, 208, 175, 113

Tasks

  1. Create a data file with these data.
  2. Report summary statistics by group.
  3. On a separate piece of graph paper, draw a side-by-side quartile plot of the data. Interpret this graph.
  4. Calculate 95% confidence intervals for group means. Show all calculations.
  5. Test variances for inequality. Report all hypothesis testing steps, including the conclusion to the test. Interpret this result.
  6. Test for inequality of means. Report all hypothesis testing steps. Interpret the results.
  7. Conduct pair-wise comparisons, if warranted. Set up the test in each instance.
  8. Summarize your results in narrative form.
  9. Determine the study's power to uncover a minimal detectable difference of 10 units.