4 Basic Data Analysis
4.1. Overview
This chapter covers the steps for preparing secondary data for statistical analysis and how to run common statistical tests. The reader will be introduced to some basic data analysis procedures using SAS 9.4. It is important to follow the steps for preparing secondary data for statistical analysis to ensure accuracy, particularly when using numerous years or combining multiple data files within each year. The examples in this chapter will use data from the 2018 National Health Interview Survey (NHIS) Sample Adult file. The examples will demonstrate ways to answer the following two research questions among adults ages 18 and older in the United States (US):
- Research Question 4.1: What are the associations between region (geographic location where participant lives) and health information technology (HIT) usage?
- Research Question 4.2: What is the association between sex and HIT usage?
The NHIS includes five questions to measure individuals’ use of different aspects of HIT usage during the past 12 months, including whether or not adults use computers to fill prescriptions, schedule appointments, communicate with others through chat groups, look up health information online, and communicate with health care providers by e-mail.1 The primary author of this textbook (Kindratt) and colleagues have used NHIS data to explore how HIT usage influences vaccination and cancer screening uptake.2,3 The examples in this chapter will focus on differences by sociodemographic factors, such as place of residence and sex. SAS 9.4 procedures will be used to demonstrate how to meet the following research objectives using common statistical tests.
- Objective 4.1: To determine the association between region and looking up health information on the internet
- Objective 4.2: To determine the association between region and filling prescriptions on the internet
- Objective 4.3: To determine the association between region and scheduling appointments on the internet
- Objective 4.4: To determine the association between region and communicating with health care provider by e-mail
- Objective 4.5: To determine the association between sex and the number of HIT uses (scale from 0 to 4)
- Objective 4.6: To determine the associations between sex and each HIT use (looking up health information online, filling prescriptions, scheduling appointments, and communicating with a health care provider by e-mail) before and after controlling for other contributing factors
It is important to note that the examples provided in this chapter are not weighted and do not include procedures for adjusting the results based on the complex survey design. Therefore, estimates you obtain as the results are not representative of the true findings. The complex sample design features that should be used when analyzing national health data will be described in Chapter 5.
4.2 Four-Step Process for Secondary Data Analysis
When preparing secondary data for statistical an