Using Stata for Survey Data Analysis

Annex 1: Quick reference guide to Stata commands ....................................................................... 58 
Annex 2: Comparison of SPSS and Stata commands ...................................................................... 59 

---

---

Using Stata for Survey Data Analysis 
 
Minot 
 
Page 1
SECTION 1: INTRODUCTION TO TRAINING GUIDE 
Background
This manual was prepared to be used as part of a one-week training course. Earlier versions of the 
manual have been used in training courses in various countries. This manual describes how to use 
Stata to store, describe, and analyze data. The emphasis is on the analysis of household survey data, 
but Stata can be used with any database.
It should be noted that this course is not a lecture course, but rather it is a semi-structured hands-on 
workshop in which trainees will use computers to learn different methods of analyzing data. Thus, 
active participation of the trainees is expected and necessary to maximize the benefit from the 
training.
The training modules focus on how to use computer software to implement a wide range of topics and 
analytical methods. In order to cover this range of methods, the course cannot provide detailed 
explanations of the all statistical methods themselves, so it is assumed that trainees have some 
familiarity with statistical concepts.
At the end of the session, we will issue Certificates of Completion to trainees who have attended all 
the sessions and mastered the concepts taught in the course.
Objectives
The objective of this training module is to improve the ability of the trainees to use Stata to generate 
descriptive statistics and tables from survey data, as well as carry out multiple linear regression 
analysis of those data. In particular, the course aims to train the participants in the following methods: 
basic file management such as opening, modifying, and saving files 
advanced file management such as merging, appending, and aggregating files 
documenting data files with variable labels and value labels 
generating new variables using various functions and operations 
creating tables to describe the distribution of continuous and discrete variables 
creating tables to describe the relationships between two or more variables
using regression analysis to study the impact of various variables on a dependent variable 
testing hypotheses using statistical methods