The SAS Workbook

144 The SAS Workbook

D. For all patients with a primary diagnosis (PRIM_DX) of 01, 02, or 04, compute frequencies

for the variables GENDER, VITAMINS, and PREGNANT. Do not create a separate data

set for this part.

E. Create a data set (PREG_35) containing all observations in CLINICAL for women who are

pregnant and over 35 years of age (as of May 1, 1996). List the ID, AGE (as of the last

birthday), and vitamin status of these women.

SECTION 4 Statistics Problems

CHAPTERS

22 Basic Descriptive Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .147

23 Tests of Proportions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .151

24 Comparing Means: Two Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . .157

25 Analysis of Variance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161

26 Parametric and Nonparametric Correlations . . . . . . . . . . . . . . . . .173

27 Simple, Multiple, and Logistic Regression . . . . . . . . . . . . . . . . . . . . .177

28 Random Assignment and Random Selection of Subjects . . . . . . . .181

29 More Advanced Statistics Projects . . . . . . . . . . . . . . . . . . . . . . . . . .185

145

146

147

Chapter 22

PROBLEMS

1 Producing Basic Descriptive Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .147

2 Computing Descriptive Statistics, Broken Down by More than One Variable . . . . . .148

3 Generating Frequency Distributions and Histograms for Categorical Variables . . . .148

4 Generating Histograms for Numeric Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .149

5 Computing Means where There Is More than One Observation per Subject . . . . . .149

INTRODUCTION

The problems in this section cover topics ranging from simple descriptive statistics to more

advanced repeated measures analysis of variance. You can use them to sharpen your statistical

programming skills, and you can use the solutions as model programs for similar statistical

problems.

You may notice that the Tools lists do not go into as much detail (such as statements and options)

as those in previous chapters. The reason for this is that too much detail in the Tools list would

give too many hints on how to solve the problems and ruin your fun.

These problems test your ability to generate basic descriptive statistics such as means and

standard deviations; test distributions for normality; and produce frequency distributions,

histograms, and two-way tables.

PROBLEM 1

Producing Basic Descriptive Statistics

Tools

PROC MEANS

BY or CLASS statements

PROC UNIVARIATE

Data

SAS data set CLINICAL

Directions

Compute the following

Basic Descriptive Statistics

148 The SAS Workbook

A. The number of non-missing observations (n), the mean, standard deviation, and standard

error (all to three decimal places) for the variables HR (heart rate), SBP (systolic blood

pressure), and DBP (diastolic blood pressure) for all patients.

B. The same statistics listed in A, except compute them separately for males and females.

C. The median HR, SBP, and DBP.

D. A stem-and-leaf plot for SBP. Test if the distribution is significantly different from a normal

distribution (use α = .05).

PROBLEM 2

Computing Descriptive Statistics, Broken Down by More than One Variable

Tools

PROC MEANS

CLASS or BY statement

Data

SAS data set CLINICAL

Directions

Compute the N, mean, and standard deviation of HR, SBP, and DBP, for each level of GENDER

and age group, where two age groups are defined as

1 = less than 65

2 = greater than or equal to 65

Compute age as the age at the time of the visit, dropping any fractional part of a year. Express

all statistics to three decimal places.

PROBLEM 3

Generating Frequency Distributions and Histograms for Categorical Variables

Tools

PROC FREQ

PROC CHART

Data

SAS data set CLINICAL