Practice for First Comp. Public Health Stats Midterm

Research Problem

MONKEY.REC. Twenty-three monkeys are randomly assigned to one of four groups. Group 1 gets a preparation of Monkey tonic, Group 2 gets a course in Applied Monkey Training (ATM, patent pending), Group 3 gets a course in advanced antics, and Group 4 gets a sham treatment. Each Monkey's Intelligence Quotient (MIQ) is recorded at the beginning of the study (BASELINE) and after 12 weeks after treatment (WEEK12). Data are:

REC  GROUP      BASELINE   WEEK12
---  ---------- ---------- ------
  1           1        298    190
  2           1        124    262
  3           1        176    334
  4           1        116    226
  5           1        152    268
  6           1        106    356
  7           2        114    210
  8           2        106    104
  9           2        100    142
 10           2         92    112
 11           2        212    382
 12           2         92    170
 13           3        180    472
 14           3        186    336
 15           3        202    416
 16           3        192    350
 17           3         80    226
 18           4        174    232
 19           4        252    566
 20           4        210    444
 21           4        162    180
 22           4         68    216
 23           4         74    184
 

Exercises

(1) Study Design: Consider the study's design. Is it experimental or observational? Are measurements paired? Are groups independent? What types of descriptions and comparisons would you like to pursue? Lay out an analysis plan.
(2) Data Entry & Validation: Create a file with these data, validating the data as needed.
(3) Data Documentation & Exchange: Export the data file to a fixed-length field (CAR) file. Document the card file in the form of a DD file. Export the data to a Lotus spreadsheet (WKS) format, and read it into an Excel(tm) session.
(4) Groups at Baseline: Perform an analysis of variance to determine whether groups differ at BASELINE. In so doing, report summary statistics and perform parametric and non-parametric tests. Also, for practice, estimate group standard errors.
(5) Change Within Group 1: Determine whether there has been a significant change in MIQ scores in Group 1. Here's the program you'll need for this analysis:

READ MONKEY
* CREATE DELTA VARIABLE
DEFINE DELTA ######
DELTA = WEEK12 - BASELINE
* SELECT OUT GROUP 1 FOR ANALYSIS
SELECT GROUP = 1
* CHECK DATA
LIST
* COMPUTE STATISTICS
MEANS DELTA

(6) Change Within Group 2, Group 3, and Group 4: Repeat the paired analysis to see if there has been a significant change within the other three groups.
(7) Comparing the Change Between Groups: Determine whether changes in MIQ scores in the four groups has changed significantly. Here's the complete computational program:

READ MONKEY
* CREATE DELTA VARIABLE
DEFINE DELTA ######
DELTA = WEEK12 - BASELINE
* CHECK DATA
LIST
* CALCULATE STATISTICS BY GROUP AND ANOVA-RELATED STATS
MEANS DELTA GROUP /N



Click for Partial Answer Key