Chapter 16 – Data Analysis: Preliminary Steps
I. Learning Objectives:
Upon completing this chapter, the student should be able to:
1. Explain the purpose of the editing process.
2. Define what coding is.
It is the technical procedure by which data are categorized.
3. Describe the kinds of information contained in a codebook.
• General instructions on how each item was coded
4. Describe common methods for cleaning the data file.
• Examining frequency distributions
5. Discuss options for dealing with missing data in analyses.
• Reporting missing information as a separate category
• Eliminating the case with missing information from all analyses
II. Chapter Outline:
A. Editing
B. Coding
Exhibit 16.1: Primary tasks in the Editing Process
2. Coding Open-Ended Items
a. Coding Factual Open-Ended Items
Chapter 16 – Data Analysis: Preliminary Steps
2
C. Aggregate Data
1. Building the Data File
D. Cleaning the Data
Manager’s Focus
E. Handling Missing Data
F. Avery Fitness Center Project
Exhibit 16.3: Avery Fitness Center Application Form
Exhibit 16.4: Data Sources on Avery Fitness Center Project
G. Summary
H. Key Terms
I. Review Questions
III. Answers to Review Questions:
1. Unless you are using an online survey that forces respondents to complete each
item before they can move on in the survey (and we don’t usually recommend
2. When a question allows multiple responses, assign separate variables for each
3. The two kinds of open-ended questions are factual open-ended items and
exploratory open-ended items. A factual open-ended question is easy to code by
converting the responses to numbers). An exploratory open-ended question is
more difficult and expensive to code. For many open-ended questions, there are
multiple legitimate responses, some of which you might not anticipate in
advance.
4. Multiple coders help reduce bias in interpreting different responses, a form of
5. There are numerous methods, including creating text data files in word
processing software, using spreadsheet software, using database software,
6. The codebook contains general instructions indicating how each item of data
7. An office error that arises during editing, coding, or data entry.
8. Data are entered into two separate data files and then compared for
discrepancies.
9. Several options exist, including (a) eliminating the case with missing information
from all analyses, (b) eliminating the case with missing information from only
IV. Instruction Suggestions:
1. Begin by again presenting the total research process so as to point out where we
are in the process. With this diagram, you can then make several points:
(a) The particular analyses one employs will depend on the decisions
study, several initial aspects of data analysis are common across studies,
including…
(b) the preliminary steps of editing, coding, developing a codebook, inputting
data, “cleaning” data after input, and dealing with missing data. These
issues are relevant to practically all studies.
2. Then the discussion can proceed to the specifics of these preliminary data
analysis steps by focusing on such questions as:
(a) Editing
— what to do with incomplete answers, obviously wrong answers,
and answers that reflect a lack of interest and how one goes
about detecting such problems.
(b) Coding
— the establishment of categories.
— the coding of open-ended versus fixed-alternative/closed-ended
questions.
— the coding of data for subsequent computer analysis. Here it is
helpful to distribute a completed sample questionnaire for a study
with clearly stated objectives. Class discussion can then be
productively directed at how each student would code the
information on this questionnaire and why.
(c) Developing the Codebook
— decisions about what type of data file to use.
(d) Cleaning the Data
— techniques for identifying blunders, including frequencies and
double-entry.
— new techniques for reducing blunders, including optical scanning.
(e) Handling Missing Data
— decisions about how to handle missing data. It is useful to remind
students that the goal should normally be to try to retain and use
It may not be possible to address all of these topics adequately in one period,
and instructors could be expected to pick and choose topics for further
elaboration on the basis of student backgrounds and their own experiences and
interests. Our experience suggests that students sometimes have difficulty
grasping what the codebook represents and how it is developed, but that once
they have begun working with a codebook (either the one from the text or, even
better, one for a project they are pursuing for the course), they quickly
understand.