You are on page 1of 145

CDC EZ-Text

User’s Guide

Version 4.0

Released 7/14/08

Developed for:
Centers for Disease Control and Prevention
Atlanta, Georgia
CDC EZ-Text 4.0 User’s Guide

Table of Contents

Chapter 1: Introduction to CDC EZ-Text 1


1.1: Background 1
1.2: Summary of CDC EZ-Text Functions 2
1.3: Notice to Users 2
1.4: Suggested Citation for EZ-Text 3
1.5: Suggested Applications 3
1.6: Sources of Help for CDC EZ-Text 4
1.7: Citations and Suggestions for Further Reading 4

Chapter 2: Technical Features and Requirements 7


2.1: Program Language and Operating Platforms 7
2.2: Computer Hardware 7
2.3: Importance of Backing Up Databases 7
2.4: Other Technical Features 7

Chapter 3: Installation and Opening EZ-Text 8


3.1: How to Install EZ-Text for Windows® 98/2000/NT/XP/Vista Users 8
3.2: Opening EZ-Text 9

Chapter 4: Creating a New Database 11


4.1: Steps in Creating a Database 11
4.2: Administration Menu Screen Overview and Default Password 12
4.3: Initializing a New Database 13
4.4: Defining Database Structure 16
4.4.1: Modify Database Information 16
4.4.2: Define or Modify Structure 18
4.4.2(a): Title 23
4.4.2(b): Section Heading 24
4.4.2(c): Sub-section Heading 25
4.4.2(d): Short Text Field 26
4.4.2(e): Long Text Field 27
4.4.2(f): Date Field 28
4.4.2(g): Numeric Field 29
4.4.2(h): Drop-down List 30
4.4.2(i): Option List 32
4.4.2(j): Page Break 35
4.4.2(k): Blank Lines 35
4.4.2(l): Comments 36
4.4.3: Modify Database Defaults 36
4.4.4: Define/Modify Codebook 39
4.4.5: Lock Database Structure 39
4.4.6: Lock Data Entry 39
4.4.7: Exit Administration 39

i
Chapter 5: Using an Existing Database 40
5.1: Opening a Database 40
5.2: Performing Data Entry 43
5.2.1: Adding New Cases 43
5.2.2: Using Field Types 44
5.2.3: Saving a Record to the Database 45
5.2.4: Using Navigation Buttons 47
5.2.5: Deleting Records 49
5.2.6: Editing Records 49
5.3: Closing a Database 50
5.4: Restoring a Backup File 52
5.5: Repairing a Database 54
5.6: Exiting EZ-Text 56

Chapter 6: Data Coding and Data Coding Queries 57


6.1: Overview of Data Coding in EZ-Text 57
6.2: Defining Codes in the Codebook 59
6.3: Modifying the Codebook 62
6.4: Coding Responses 68
6.5: Viewing and Printing the Codebook 70
6.6: Exporting and Importing the Codebook 74
6.6.1: Creating a Codebook Export File 75
6.6.2: Importing a Codebook 76
6.7: Coding Queries and Query Reports from the Coding Form 79
6.7.1: Coding Queries from the Coding Form 79
6.7.2: Coding Query Reports from the Coding Form 83

Chapter 7: Database Queries and Query Reports from the Data Entry Form 85
7.1: Database Queries from the Data Entry Form 85
7.2: Database Query Reports from the Data Entry Form 89

Chapter 8: EZ-Text Agents 91


8.1: Copying a Subset of Records with the Subset Agent 91
8.2: Merging Files with the Merge Agent 98
8.3: Creating ASCII Text Files and Numeric Matrices
with the Data Export Agent 101
8.4: Checking Spelling with the Spell Check Agent 109

Chapter 9: Reliability Agents 113


9.1: Creating Comparison Files for Reliability Tests 114
9.2: Running the Two-coder Reliability Comparison 118
9.2.1: Comparing Sets of Codes Assigned by Two Coders 118
9.2.2: Comparing Assigned Codes Across Responses 124
9.3: Running the Multi-coder Reliability Comparison 131
9.3.1: Comparing Assigned Codes Across Responses 132
9.3.2: Comparing Sets of Codes Assigned by Multiple Coders 139

ii
Chapter 1. Introduction to CDC EZ-Text

1.1 Background
Many research projects call for collecting semi-structured qualitative data. These include
projects involving interview guides with a series of open-ended questions, consistently
administered to each person in the sample, with responses that may vary widely in
content, length, and complexity. Semi-structured qualitative studies vary greatly in size
and complexity. Some small studies might limit data collection to only a few open-ended
questions administered by a single interviewer to a small sample of respondents (e.g., 15
to 30 individuals). In contrast, other studies might be quite large. For example, Centers
for Disease Control and Prevention (CDC)-sponsored HIV/AIDS risk behavior studies
frequently use semi-structured interview instruments, typically with 20 to 50 open-ended
questions. Sample sizes may include several hundred individuals living in different
communities throughout the United States. Some projects use as many as 30 interviewers
(Carey and Gelaude 2008). Regardless of the size of the study, following each interview,
qualitative researchers need to write summaries or generate verbatim transcripts from
audio recordings made during the interview. The final computerized databases contain
hundreds and sometimes thousands of pages of text. After data entry is complete, further
coding and analysis of the themes present in the write-ups takes place in order to generate
final results and conclusions. Many methodological options have been described in the
qualitative research literature for guiding these steps (see Section 1.7 for suggested
further readings).

One challenge in this type of research is organizing and managing the data before
analysis. Interviewers may work on different Windows® or Macintosh® platforms, and
converting files between systems can be time-consuming. Even when all researchers use
the same platform, variations in margin, font, and other preferences affect the physical
organization of the data. Response segments relevant to a specific question may be
written up in different locations for each respondent. Further complications occur when
interviewers change the data collection protocol independently, such as altering the
number, question order, question wording, follow-up probes, or overall intention of the
questions in the interview guide.

These conditions make it difficult for study coordinators to monitor and ensure
consistency of data collection and write-up across the sample. Even when interviews are
conducted in a comparable manner, inconsistent organization of interview notes may
mean that study coordinators must reformat the data into a standard layout prior to
analysis. Not only does this waste financial resources and staff time, it delays generation
of research findings. Occasionally, these barriers are so large that the data are never fully
analyzed.

1
1.2 Summary of CDC EZ-Text Functions
CDC EZ-Text Version 4.04 (EZ-Text) is a qualitative software program developed to
help researchers create, manage, and analyze semi-structured qualitative databases. EZ-
Text helps solve the problem of consistency across interview write-ups by allowing
researchers to design a series of qualitative data entry templates tailored to their
questionnaire. Data can be typed directly into the templates or copied from word
processor documents and can accommodate non-English alphabet characters such as Æ,
Ö, é, à, í, ó, and ñ. Following data entry, investigators can create on-line codebooks,
apply codes to specific response passages, develop case studies or case series, conduct
database searches to identify text passages that meet user-specified conditions, and export
data in a wide array of formats for further analysis with other qualitative or statistical
analysis software programs. Project managers can merge data files generated by different
interviewers for combined cross-site analyses. The ability to export and import the
codebook helps coordinate the efforts of multiple coders working simultaneously with
copies of the same database file.

1.3 Notice to Users


The current version of "CDC EZ-Text" was jointly designed and developed by staff from
MANILA Consulting Group, Inc. and the CDC. Its primary purpose is to help CDC meet
its own public health research needs. Other investigators may find the program useful for
different purposes. However, neither CDC nor MANILA Consulting Group, Inc. makes
any expressed or implied guarantees that this software program will be an appropriate or
useful tool for addressing the needs of other users. No warranty is made or implied for
the use of the software for any particular purpose. CDC, MANILA Consulting Group,
Inc. and their staff or subcontractors are not responsible or liable in any way for any
consequences resulting from the use or misuse of the EZ-Text program or its
documentation.

The EZ-Text software program and its associated documentation are in the public
domain; they may be freely copied and distributed without restriction. However, potential
users should understand that there is no formal mechanism for technical support beyond
the help files included with the program installation CD or in the EZ-Text online help
files.

Trade names are used for identification purposes only or for examples; no endorsement
of particular products is intended or implied. The use of trade names or trademarks in the
EZ-Text documentation does not imply that such names, as understood by the Trade
Marks and Merchandise Marks Act, may be used freely by anyone.

2
1.4 Suggested Citation for EZ-Text
Carey, James W.; Wenzel, Patrick H.; Gelaude, Deborah; Sheridan, John; Reilly, Cindy;
and Burns, David (2008). CDC EZ-Text: Software for Collection, Management and
Analysis of Semi-structured Qualitative Databases (Version 4.04). Atlanta: Developed by
MANILA Consulting Group, Inc. for the Centers for Disease Control and Prevention.

1.5 Suggested Applications


A central assumption behind the design of the EZ-Text program is that users are
examining the same set of topics with each individual in their sample. As noted above,
this may entail using the same semi-structured interview instrument with each
respondent.

EZ-Text was specially designed for use by researchers who are collecting and analyzing
semi-structured qualitative data generated through face-to-face interviews with a sample
of individual respondents. Although EZ-Text can be used in small studies, the software
also includes a number of features intended to assist research teams conducting large-
scale qualitative studies (Guest and MacQueen 2008). In some limited circumstances, and
at the discretion of the user, EZ-Text might successfully be adapted for other purposes
(e.g., semi-structured focus group data, semi-structured behavioral observation field
notes, or perhaps analysis of pre-existing text documents).

However, EZ-Text is not likely to be very helpful to researchers administering highly


unstructured ethnographic studies where the discussion topics or behavioral observation
categories vary greatly among respondents in the sample. Similarly, EZ-Text is not a
substitute for statistical analysis software. EZ-Text helps address a specific set of needs
related to semi-structured qualitative data, especially when it is collected and coded by
different individuals following a common protocol on a multi-site research project.
Investigators who do not wish to generate and analyze a semi-structured qualitative
database should select a different software tool (Lewins and Silver 2007; Weitzman and
Miles 1995).

Researchers whose needs include use of this software tool will find the simple
instructions in this guide an easy introduction to the program.

3
1.6 Sources of Help for CDC EZ-Text
There are several ways that users can find out about EZ-Text features:

• Information provided in this User’s Guide. This guide is automatically included


with the installation files.

• An “example.ezt” database file is automatically installed with the program. For a


fast start, it is recommended that new users open this file, read its contents, and
experiment with the software functions.

• EZ-Text website:
http://www.cdc.gov/hiv/topics/surveillance/resources/software/ez-text/index.htm

• If users cannot find the answer to their question using these documents, emails
may be sent to: eztext@cdc.gov

1.7 Citations and Suggestions for Further Reading


Bernard, H. R.
2002 Social Research Methods: Qualitative and Quantitative Approaches. Thousand
Oaks, Calif.: Sage.

Bernard, H. R.
2006 Research Methods in Anthropology: Qualitative and Quantitative Approaches
(4th ed.). Walnut Creek, Calif.: AltaMira.

Boyatzis, R. E.
1998 Transforming Qualitative Information: Thematic Analysis and Code
Development. Thousand Oaks, Calif.: Sage.

CAQDAS
2008 Computer Assisted Qualitative Data analysis. CAQDAS Networking Project.
Internet discussion and resource website. http://caqdas.soc.surrey.ac.uk
Accessed July 3, 2008.

Carey, J. W., and D. Gelaude


2008 Systematic methods for collecting and analyzing multidisciplinary team-based
qualitative data. In, “Handbook for Team-based Qualitative Research,” G. Guest
and K. MacQueen, editors. Lanham, MD: Altamira.

Carey, J. W., M. Morgan, and M. J. Oxtoby


1996 Intercoder agreement in analysis of responses to open-ended interview questions:
Examples from tuberculosis research. Cultural Anthropology Methods Journal
8(3):1-5.

4
Fleiss, J. L.
1971 Measuring nominal scale agreement among many coders. Psychol.
Bull. 76: 378-382.

Fleiss, J. L., B. Levin, and M. C. Paik


2003 Statistical Methods for Rates and Proportions. Hoboken, NJ: John Wiley and
Sons.

Gorden, R. L.
1998 Basic Interviewing Skills (rev. ed.). Long Grove, Ill.: Waveland.

Guest, G. and K. M, editors


2008 Handbook for Team-based Qualitative Research. Lanham, MD: Altamira.

Hruschka, D., D. Schwartz, D. C. St. John, E. Picone-DeCaro, R. A. Jenkins, and J. W.


Carey
2004 Reliability in coding open-ended data: Lessons learned from HIV behavioral
research. Field Methods 16(3):307-331.

Lewins, A., and C. Silver


2007 Using Software in Qualitative Research: A Step-by-step Guide. London: Sage.

Mack, N., C. Woodsong, K. MacQueen, G. Guest, and E. Namey


2005 Qualitative Research Methods: A Data Collector's Field Guide. Research
Triangle Park, N.C.: Family Health International.

MacQueen, K., E. McLellan, K. Kay, and B. Milstein


1998 Code book development for team based qualitative analysis. Cultural
Anthropology Methods Journal 10(2):31-36.

MacQueen, K. M., and B. Milstein


1999 A systems approach to qualitative data management and analysis. Field Methods
11(1):27-39.

McLellan, E., K. M. MacQueen, and J. L. Neidig


2003 Beyond the qualitative interview: Data preparation and transcription. Field
Methods 15(1):63-84.

Miles, M. B., and A. M. Huberman


1994 Qualitative Data Analysis: An Expanded Sourcebook. Thousand Oaks, Calif.:
Sage.

Patton, M. Q.
2002 Qualitative Research and Evaluation Methods (3rd ed.). Thousand Oaks,
Calif.: Sage.

5
Qual-Software
2008 Archives of QUAL-SOFTWARE@JISCMAIL.AC.UK. Internet discussion
group. Available at www.jiscmail.ac.uk/lists/qual-software.html. Acessed July 3,
2008.

Ryan, G. W., and H. R. Bernard


2003 Techniques to identify themes. Field Methods 15:85-109.

Ulin, P. R., E. T. Robinson, and E. E. Tolley


2005 Qualitative Methods in Public Health: A Field Guide for Applied Research. San
Francisco: Jossey-Bass.

Ulin, P. R., E. T. Robinson, E. E. Tolley, and E. T. McNeill


2002 Qualitative Methods in Public Health: A Field Guide for Applied Research
in Sexual and Reproductive Health. Research Triangle Park, N.C.: Family
Health International.

Weber, R. P.
1990 Basic Content Analysis (2nd ed.). Newbury Park, Calif.: Sage.

Weitzman, E. A., and M. B. Miles


1995 Computer Programs for Qualitative Data Analysis. Thousand Oaks, Calif.: Sage.

6
Chapter 2. Technical Features and Requirements

2.1 Program Language and Operating Platforms


The program for EZ-Text was written in Visual Basic® and a proprietary database
format. EZ-Text includes features which allow users to export data from EZ-Text
databases for use with other software programs. All EZ-Text database file names must
have an ".ezt" extension (except for database backup files, which have a .bak extension).
EZ-Text runs on Windows® 98, Windows® NT, Windows® 2000, Windows® ME,
Windows® XP, and Windows® Vista operating systems. There are no plans to develop
versions for Macintosh® or other systems. Please note that Windows® 2000 users must
have Service Pack 4 or later installed for EZ-Text to install properly. Windows® XP
users need to have Service Pack 1 or later installed.

2.2 Computer Hardware


EZ-Text should be installed on Windows® compatible computers with at least 512 MB
of RAM. The program may work on less powerful computers, but users may exceed
system resources when working with large files. Qualitative databases can become very
large, and the computer must be able to work with large files. Use of EZ-Text on weak or
out-of-date computers may result in poor performance, system crashes, or data loss.

2.3 Importance of Backing Up Databases


As with other applications, users are always strongly recommended to copy their data
files to back-up locations regularly (for example, copy the database to a separate file
named for the date or day of the week after each work session). EZ-Text simplifies this
process by automatically asking users if they would like to make a back-up copy each
time a database is closed.

2.4 Other Technical Features


EZ-Text Version 4.04 is not designed for simultaneous use by multiple users on a LAN
or as a web-based application. Future versions of EZ-Text may address these limitations.
However, the program can be independently installed on local hard drives at different
work sites or multiple computers on a LAN. Files created on separate computers can be
merged, if each user uses identical data entry templates and respondent ID numbers do
not overlap. Once a template file is finalized and copies are distributed to research team
members, a password system prevents unauthorized personnel from making independent
modifications. This helps guarantee that templates used for data entry on different
computers remain comparable and can be merged later. A second password system
protects the content of the codebook, which helps ensure that multiple coders use
identical coding instructions during data analysis.

7
Chapter 3. Installation and Opening EZ-Text

3.1 How to Install EZ-Text for Windows® 98/2000*/NT/XP*/Vista Users


*Please note that Windows® 2000 users need to have Service Pack 4 or later installed
for EZ-Text to install properly. Windows® XP users need to have Service Pack 1 or
later installed.

If installing from a CD:


Step 1. Insert the CD labeled “CDC EZ-Text 4.04” in the CD Drive.
Step 2. Click on the Start button on the bottom taskbar, point to Settings, then click
Control Panel.
Step 3. Double-click on Add/Remove Programs.
Step 4. Click on the Install button and follow the instructions on the screen.

If installing from the Internet:

Follow the instructions provided on the website.

Note: It is strongly recommended that CDC EZ-Text be installed in the default folder
(C:\Program Files\eztext40). If CDC EZ-Text is installed in a different folder, the
EZText40.ini file will be placed in that folder and may need to be modified to reflect
the new paths.

8
3.2 Opening EZ-Text
To open EZ-Text, click Start in the lower left-hand corner of Windows®. Select
Programs, then select EzText40, and click the EzText40 icon.

When EZ-Text opens, a program information box appears (Figure 3.2-1).

Figure 3.2-1 Program Information Box

Wait a few seconds for the box to close automatically, or press any key on the keyboard
to close it faster. A blank home screen will remain (Figure 3.2-2).

9
Figure 3.2-2 Blank Home Screen

At this point, a new database can be created or an existing database can be opened for
viewing or for data entry. Chapter 4 discusses how to create a new database; Chapter 5
covers how to use an existing database.

10
Chapter 4. Creating a New Database

4.1 Steps in Creating a Database


A single researcher (for example, the study project manager) should create the new
templates and edit existing templates contained in a database to avoid confusion. This
requires eight steps:
1) Opening the Administration Screen
2) Initializing a New Database
3) Modifying Database Information (includes naming the study, setting passwords, and
choosing the Force Text Selection setting)
4) Defining or Modifying Structure
5) Modifying Database Defaults
6) Defining or Modifying Codebook
7) Locking Database Structure
8) Locking Data Entry

11
4.2 Administration Menu Screen Overview and Default Password
The Administration Menu is password-protected to prevent unauthorized users from
modifying a database. In research projects where many staff need to enter or work with
data, we strongly recommend that senior project managers create and limit access to these
passwords. To enter the Administration area, select Administration from the File menu on
the menu bar on the blank home screen (Figure 4.2-1).

Figure 4.2-1 Entering Administration

12
The program will prompt for the password to enter the Administration Menu (Figure 4.2-
2). The default password is Admin. (Note: Passwords are case sensitive.) Enter the
password in the space provided and press Enter on the keyboard or click the Continue
command button.

Figure 4.2-2 Administration Password Prompt

4.3 Initializing a New Database


The Administration Menu will open (Figure 4.3-1). The only available options initially
are Initialize New Database, Help, and Exit Administration. To create a new database,
click Initialize New Database.

Figure 4.3-1 Administration Menu Before Initializing a New Database

13
The Save As dialog box prompts for a file name for the database (Figure 4.3-2). All file
names must have an “.ezt” extension. Once the database file has been named, press Enter
on the keyboard or click Save. If the administrator enters the name of an existing
database, the program provides a warning prompt before overwriting the old database
with the new.

Figure 4.3-2 Naming a New Database

14
A progress meter will appear on the screen as the database is being created. Once
initialization is complete, the appearance of the Administration Menu will change. The
Initialize New Database button is disabled, while the other options are now available
(Figure 4.3-3).

Figure 4.3-3 Administration Menu after Database Creation

15
4.4 Defining Database Structure
Once the EZ-Text database has been initialized, the administrator may perform the
following functions using the steps defined in the sections indicated:
1) Provide a study name, change the administrative password, create passwords for the
codebook and the database, and change the Force Text Selection setting. (Section
4.4.1)
2) Design the data entry form by using titles or section headings and by defining and
placing variables. (Section 4.4.2)
3) Change the default EZ-Text settings for text alignment and fonts. (Section 4.4.3)
4) Create and maintain the codebook. (Section 4.4.4)
5) Lock the database structure. (Section 4.4.5)
6) Lock data entry. (Section 4.4.6)

To access these functions, click on the appropriate button on the Administration Menu.

4.4.1 Modify Database Information


Click the Modify Database Information button to provide a study name, set or change
passwords, or change the Force Text Selection setting (Figure 4.4-1).

Figure 4.4-1 Modify Database Information Screen

Naming the study: In the field labeled Study Name, the words “New Database” appear.
Use the mouse to highlight “New Database,” and then type the name of the study.

Setting or changing passwords: The administrator can assign unique passwords for
entering the Administrative Menu, for accessing the database, and for creating and
maintaining the codebook.
The default password for both the Administration Menu
and for the Codebook Maintenance screen is "Admin"
(do not type the quotation marks)
Remember: Passwords Are Case Sensitive!

16
Change the administrative password and codebook password by using the mouse to
highlight Admin in their respective password fields and typing the new password.

To assign a password to the database, click the Database Password field and type the
desired password. This will restrict access to the entire database by prompting for the
password before opening the database. This is a good feature to use for sensitive data that
needs to remain secure. If a database password is not assigned, there will be no
restrictions to opening the database.

These passwords apply only to the database file that the user has open. Users can assign
different sets of passwords to different EZ-Text database files.

Changing Force Text Selection setting: The administrator has the option of forcing
coders to select text when coding responses. The default setting for this feature is "off". If
users want to assign codes to specific text passages within a longer response (Section
6.4), change this check box value to "on".

To turn this feature on, simply click on the check box labeled Force Text Selection to
place a check in the box. When the Force Text Selection feature is on, users must identify
the exact text within the response that applies to the code when assigning the code.

Note that the fields for Database Version, Date Created, and Date Modified are disabled.
These values are automatically maintained by EZ-Text.

To save the study name, passwords, and Force Text Selection setting and return to the
Administration Menu, click Save Changes. To return to the Administration Menu without
saving changes, click Cancel.

17
4.4.2 Define or Modify Structure
Designing the data entry form and creating and maintaining variables are simple in EZ-
Text. To do so, click the Define/Modify Structure button on the Administration Menu.
This opens the interface for defining and placing titles or headers and variables (Figure
4.4-2).

Figure 4.4-2 Define/Modify Structure Screen

On this screen, the administrator can:

1) Add a new variable or header information at the end of the element list.
2) Insert a new element between other elements in the list.
3) Modify existing elements.
4) Delete an element.
5) Move an existing element up or down in the list.
6) Preview a data entry form without leaving the administration area.

Note that a numeric variable called respid (Respondent ID) already exists. This is a
required variable that EZ-Text creates when the database is initialized. It cannot be
deleted. It can, however, be modified. (See Modify button instructions below.)

18
Adding a new element (Add New button): To add a new element, first select the
Element Type from the drop-down list to the left of the Add New button (Figure 4.4-3),
and then click on the Add New button to open that element's Add New dialog box. The
element type options and specifics about defining each are described in detail in Sections
4.4.2(a) through 4.4.2(l).

Figure 4.4-3 Selecting an Element to Add To/Insert In the Database

Inserting an element (Insert button): The Add New function places elements at the end
of the element list and, therefore, at the end of the data entry form. To place an element
between elements that already exist, go to the Define/Modify Structure screen (Figure
4.4-2) and highlight the element that the new element will precede by clicking anywhere
in that element's row. Select the Element Type to be inserted from the drop-down list and
click Insert. Create the elements as described in Sections 4.4.2(a) through 4.4.2(l). The
screens for inserting the elements are similar to the Add New screens shown in those
sections. The only difference is in the title bar of each screen. Instead of Add New […]
Element, the title bar reads Insert New […] Element.

Modifying an element (Modify button): Once an element has been created, it can be
modified at any time. All elements except Page Break can be modified. To modify an
element, go to the Define/Modify Structure screen (Figure 4.4-2), highlight the element

19
that needs to be changed by clicking anywhere in that element's row, and then click
Modify. With the exception of the Drop-down List and Option List elements, the Modify
screens are the same as the Add New and Insert screens, except that the title bar reads
Modify […] Element.

The Modify screens for Drop-down List and Option List elements have an additional
command button, Add/Modify Values (Figure 4.4-4).

Figure 4.4-4 Modify Drop-down List Element Screen

Clicking the Add/Modify Values button on the Modify Drop-down List Element screen
opens the Drop-down List Maintenance screen (Figure 4.4-16). Clicking the Add/Modify
Values button on the Modify Option List Element screen opens the Option List
Maintenance screen (Figure 4.4-19). Instructions for these two maintenance screens are
in Sections 4.4.2(h) and 4.4.2(i), respectively. After clicking Return on the list
maintenance screens from here, the program returns to the element's Modify screen rather
than returning to the Define/Modify Structure screen (Figure 4.4-2) as it does in the Add
New mode.

Deleting an element (Delete button): To delete an element, simply highlight that


element on the Define/Modify Structure screen (Figure 4.4-2) by clicking anywhere in
that element's row, and then click Delete.

WARNING: Be careful when deleting an element. No prompt appears to verify the


deletion. When a variable is deleted, all data for that variable are also deleted.

20
Changing list placement (Up and Down buttons): The administrator can change an
existing element's placement in the element list and on the data entry form. Simply
highlight the element to be moved on the Define/Modify Structure screen (Figure 4.4-2)
by clicking anywhere in that element's row. Then, to move the element up in the list,
click the Up button. Conversely, to move the element down in the list, click the Down
button.

EZ-Text can automatically exchange elements' question numbers while moving an


element up and down the list. To activate this feature, click Options on the menu bar of
the Define/Modify Structure screen and select Swap question numbers while moving
elements (Figure 4.4-5). A check mark will appear next to the option to indicate that the
feature is on.

Figure 4.4-5 Swapping Question Numbers While Moving Elements

Now when an element is moved up or down the list, its number changes to the number of
the element with which it switches placement. To deactivate this feature, once again
select Options from the menu bar on the Define/Modify Structure screen and select Swap
question numbers while moving elements.

Previewing new elements (Preview Form button): At any time while creating elements,

21
the administrator can see how the elements will appear on the data entry form by clicking
the Preview Form button on the lower right of the Define/Modify Structure screen
(Figure 4.4-2). The Data Entry Form Preview screen will appear (Figure 4.4-6).

Figure 4.4-6 Data Entry Form Preview Screen

In preview mode, the administrator can see how the form will look and function without
leaving the administration area. The fields will accept text or numbers (depending on
field type) and the drop-down lists and option lists can be clicked and selected. Any
information typed into the fields on the Data Entry Preview Form screen will not be
saved anywhere in the EZ-Text program.

To return to the Define/Modify Structure screen, either click Return on the menu bar or
click the Return to Admin button on the lower left of the screen.

Returning to the menu (Return to Menu button): Once the elements have been created,
modified, and arranged, click either Return on the menu bar or Return to Menu on the
lower left of the Define/Modify Structure screen (Figure 4.4-2). The administrator can
always return to the Define/Modify Structure screen at any time to add more elements or
to make any changes by clicking Define/Modify Structure on the Administration Menu
(Figure 4.3-3).

22
Sections 4.4.2(a) through 4.4.2(l) below detail each of the element types that can be
created, modified, and ultimately used to enter data into an EZ-Text database.

4.4.2(a) Title
This is a simple text feature to name the data entry form. Select Title from the Element
Type drop-down list and click the Add New button. This will open the Add New Title
Element screen (Figure 4.4-7).

Figure 4.4-7 Add New Title Element Screen

The default values for Title are center alignment with a font of Arial 14-point bold. These
options can be changed during creation or modified later. (See Modify button
instructions.) The administrator can also specify the Number of Lines (up to five) to
display the title on the form. Enter the title in the Title field (maximum length--255
characters).

Click the Save button to return to the Define/Modify Structure screen (Figure 4.4-2). To
return to the Define/Modify Structure screen without saving the Title, click Cancel.

23
4.4.2(b) Section Heading
This element is similar to Title, but the default values are left-aligned Arial 12-point
normal font. Select this option from the Element Type drop-down list and click Add New
to define a Section Heading (Figure 4.4-8).

Figure 4.4-8 Add New Section Heading Element Screen

After typing the text in the Section Heading field and making any changes to the font
attributes, click the Save button to return to the Define/Modify Structure screen (Figure
4.4-2). To return to the Define/Modify Structure screen without saving, click Cancel.

24
4.4.2(c) Sub-section Heading
This element is similar to Title and Section Heading, with default values of left-aligned
Arial 10-point normal font. Select this option from the Element Type drop-down list and
click Add New to define a Sub-section Heading (Figure 4.4-9).

Figure 4.4-9 Add New Sub-section Heading Element Screen

After typing the text in the Sub-section Heading field and making any changes to the font
attributes, click the Save button to return to the Define/Modify Structure screen (Figure
4.4-2). To return to the Define/Modify Structure screen without saving the Sub-section
Heading element, click Cancel.

25
4.4.2(d) Short Text Field
This fixed-length element should be used when the response is expected to be relatively
short. Examples of short text fields would be respondents' first or last names and mailing
or e-mail addresses. Select this option from the Element Type drop-down list and click
the Add New button. This will open the Add New Short Text Element screen (Figure 4.4-
10).

Figure 4.4-10 Add New Short Text Element Screen

The administrator can change the Short Text Field's alignment and font. The default is
left-aligned Arial 10-point normal. Required fields are Question Number, Field Width,
and Caption.

The EZText Variable Name is assigned by EZ-Text and cannot be changed.

The Variable Description has a length limit of 30 characters and is used to describe the
Short Text Field in the administration area.

Question Number will appear on the form to the left of the caption. It can contain letters
and symbols in addition to numbers. Its limit is 5 characters. Please note this is a
required field.

Field Width has a maximum value of 60, but the administrator can restrict its length to
fewer than 60 characters. This determines the length of the response that can be entered
into the database. Please note that this is a required field.

The Caption appears on the form between the question number and the data entry field. It
describes what is to be entered into the field and has a limit of 20 characters. The

26
caption's default is Arial 10-point normal, and it is right aligned. The alignment and font
of the caption can be changed. Please note that this is a required field.

Click the Save button to return to the Define/Modify Structure screen (Figure 4.4-2). To
return to the Define/Modify Structure screen without saving the Short Text Field element,
click Cancel.

4.4.2(e) Long Text Field


Use this element for open-ended questions--when the length of a respondent's answer
cannot be reasonably anticipated. Select this option from the Element Type drop-down
list and click the Add New button. This will open the Add New Long Text Element screen
(Figure 4.4-11).

Figure 4.4-11 Add New Long Text Element Screen

This screen is similar to the Add New Short Text Element screen (Figure 4.4-10). The
font and alignment for both the element and the caption can be changed, and the EZText
Variable Name, Variable Description, and Question Number fields are identical.
Required fields are Question Number and Caption.

The administrator can set the No. of Caption Lines (1 through 10) based on the length of
the variable's description or question. Captions that exceed the number of lines will be
truncated, so it is important to set the number of caption lines carefully.

27
In the Caption field, type the variable's description or question as it is to be displayed on
the form.

Select the No. of Element Lines (2 through 20) that will be visible on the screen at any
given time. The respondent's answer can exceed this number of lines, at which point the
vertical scroll bar along the right side of the field will become active.

Click the Save button to return to the Define/Modify Structure screen (Figure 4.4-2). To
return to the Define/Modify Structure screen without saving the Long Text Field element,
click Cancel.

4.4.2(f) Date Field


This element captures dates, such as a respondent's birth date or the date the
questionnaire was administered. Select this option from the Element Type drop-down list
and click the Add New button. This will open the Add New Date Element screen (Figure
4.4-12).

Figure 4.4-12 Add New Date Element Screen

Essentially identical to creating a Short Text Field (Section 4.4.2(d)), the only exception
is the absence of Field Length on the Add New Date Element screen.

Click the Save button to return to the Define/Modify Structure screen (Figure 4.4-2). To
return to the Define/Modify Structure screen without saving the Date Field element, click
Cancel.

28
4.4.2(g) Numeric Field
Using this field for numeric responses cuts down on data entry errors. Numbers could
also be stored in Short Text Fields, but Short Text Fields have no way of limiting input
solely to numbers. Select this option from the Element Type drop-down list and click the
Add New button. This will open the Add New Numeric Element screen (Figure 4.4-13).

Figure 4.4-13 Add New Numeric Element Screen

This screen is very similar to the Add New Short Text Element screen (Figure 4.4-10).
One difference is the range for Field Width. For a numeric field, Field Width is limited to
nine digits.

There are also two fields unique to the Add New Numeric Element screen. Minimum
Value sets a lower limit for allowable data entry values, and Maximum Value sets the
upper limit. These fields can be left blank to allow for any number in the range up to the
Field Width (for example, a four-digit field width allows any number from 0 to 9999).
One or the other or both can be used when an upper or lower limit is known in advance. It
is recommended to set Minimum and Maximum Values whenever possible to reduce data
entry errors.

29
EZ-Text accommodates two types of numeric variables. First, if the user wants to only
enter whole numbers (with no numerals to the right of the decimal place), select the
"Whole Number" option from the Type of Number selection list. Second, if the user wants
this variable to accommodate decimal numbers, select the "Floating Point" option from
the selection list. Decimal numbers may have 1 to 6 numerals to the right of the decimal
place. A pop-up screen will appear if the "Floating Point" option is selected; this pop-up
screen allows the user to specify how many numerals they want to allow to the right of
the decimal for the variable.

Click the Save button to return to the Define/Modify Structure screen (Figure 4.4-2). To
return to the Define/Modify Structure screen without saving the Numeric Field element,
click Cancel.

4.4.2(h) Drop-down List


Use this element when the set of responses is known in advance. For example, a Drop-
down List could be used for a race or city field. Select this option from the Element Type
drop-down list and click the Add New button. This will open the Add New Drop-down
List Element screen (Figure 4.4-14).

Figure 4.4-14 Add New Drop-down List Element Screen

Creating a Drop-down List is a two-step process. The first step is identical to creating a
Short Text Field (Section 4.4.2(d)), except that there is no Element Alignment setting.
After filling out the fields for a Drop-down List and clicking Save, the administrator has
the option to define values for the Drop-down List (Figure 4.4-15).

30
Figure 4.4-15 Define Drop-down List Values Prompt

Clicking Yes brings up the Drop-down List Maintenance screen (Figure 4.4-16).
(Clicking No returns the program to the Define/Modify Structure screen (Figure 4.4-2).
The administrator can define the response options for the Drop-down List later. (See the
Modify button instructions.))

Figure 4.4-16 Drop-down List Maintenance Screen

Enter each response option in the New List Value field exactly as it is to appear in the
Drop-down List and click Add to List or just hit the Enter key. If a value is mistyped or
otherwise incorrect, highlight that value and click Remove from List. When all options
have been added, click Return to return to the Define/Modify Structure screen (Figure
4.4-2).

Note: Values are displayed alphabetically both on the Drop-down List Maintenance
screen and on the actual Drop-down List on the data entry form.

31
4.4.2(i) Option List
Like the Drop-down List, this element is used when the set of responses is known in
advance, such as a gender field. It can also be used to create a Likert scale. As a general
guideline, use the Drop-down List for more than five responses, and the Option List for
five or fewer answer choices. This rule is for space and aesthetic considerations. Either
element can be used regardless of the number of answer choices.

Select Option List from the Element Type drop-down list on the Define/Modify Structure
screen, and click the Add New button. This will open the Add New Option List Element
screen (Figure 4.4-17).

Figure 4.4-17 Add New Option List Element Screen

32
Like the Drop-down List, creating an Option List is a two-step process. The first step is
similar to creating a Long Text Field (Section 4.4.2(e)). The only difference is that there
is no No. of Element Lines setting. After filling out the fields for the Option List, click the
Save button. As with the Drop-down List, a prompt will appear asking if the
administrator would like to define the values for the Option List now (Figure 4.4-18).

Figure 4.4-18 Define Option List Values Prompt

To create the Option List now, click Yes. The Option List Maintenance Screen will open
(Figure 4.4-19). (Clicking No will return the program to the Define/Modify Structure
Screen (Figure 4.4-2). The administrator can add the Option List at a later time. (See the
Modify button instructions.))

Figure 4.4-19 Option List Maintenance Screen

Clicking Add New will allow the administrator to add an option value via the Add New
Option List Value prompt (Figure 4.4-20).

33
Figure 4.4-20 Add New Option List Value Prompt

In the Option Text field, enter the option as it will appear on the data entry form. Assign
the Corresponding numeric score for this option and click Save to return to the Option
List Maintenance screen. To return to the Option List Maintenance screen without saving,
click Cancel.

To modify an existing option, highlight the option on the Option List Maintenance screen
and click Modify. This will provide the Modify Option List Value prompt (Figure 4.4-
21).

Figure 4.4-21 Modify Option List Value Prompt

Note that the only field available to modify is the Option Text field. After making the
correction, click the Update button. To return to the Option List Maintenance screen
without making a modification, click Cancel.

To remove an option from the list, highlighting the desired option and clicking the
Remove button on the Option List Maintenance screen will provide the Delete Option
List Value prompt (Figure 4.4-22).

Figure 4.4-22 Delete Option List Value Prompt

34
To confirm removal, click Delete. To return to the Option List Maintenance screen
without removing the option, click Cancel.

When all options have been added, modified, or removed as needed, click Return on the
Option List Maintenance screen to return to the Define/Modify Structure screen (Figure
4.4-2).

4.4.2(j) Page Break


This element adds a page break to the data entry form. Select this option from the
Element Type drop-down list and click Add New. If the questionnaire is long, this feature
may need to be used to see all the questions on the screen.

4.4.2(k) Blank Lines


This element is a spacer. Select this option from the Element Type drop-down list and
click Add New. A pop-up window will appear prompting for the number of blank lines
(Figure 4.4-23).

Figure 4.4-23 Blank Lines Element Screen

After selecting the desired Number of Blank Lines (1 through 10) from the drop-down
list, click Save to return to the Define/Modify Structure screen (Figure 4.4-2). To return
to the Define/Modify Structure screen without creating any blank lines, click Cancel.

35
4.4.2(l) Comments
This element is very much like Title and Section and Sub-section Headings in the way it
is created. It can be used to provide on-screen information/instructions to data entry
personnel. Select this option from the Element Type drop-down list and click Add New.
This will open the Add New Comments Element screen (Figure 4.4-24.)

Figure 4.4-24 Add New Comments Element Screen

Note that a comment that is longer than the No. of Caption Lines will be truncated.
Therefore, the Comments Caption field text should be entered first and the No. of
Caption Lines chosen accordingly. A comment can be 10 caption lines long.

Click the Save button to return to the Define/Modify Structure screen (Figure 4.4-2). To
return to the Define/Modify Structure screen without saving the Comments element, click
Cancel.

4.4.3 Modify Database Defaults


If the administrator knows ahead of time precisely how certain elements (such as
alignment and font attributes) are to appear on the screen, the defaults for those elements
can be changed in EZ-Text to save time while creating the elements.

36
To change element defaults, click Modify Database Defaults on the Administration Menu
(Figure 4.3-3) to access the System Defaults Maintenance screen (Figure 4.4-25).

Figure 4.4-25 System Defaults Maintenance Screen

All element types have at least one default that can be changed here. The current settings
for the elements are displayed in a spreadsheet format. Any default options within an
element that cannot be changed are indicated by a value of “n.a.” in the column for that
option. In addition, those option fields will not appear on the Modify Database Defaults
screen.

To change the default values for an element, click anywhere in that element's row on the
System Defaults Maintenance screen and click Modify System Default in the lower right
corner of the screen. This opens the Modify Database Defaults window (Figure 4.4-26).

37
Figure 4.4-26 Modify Database Defaults Screen

After changing the settings to the desired values, click the Save button to return to the
System Defaults Maintenance screen. To return to the System Defaults Maintenance
screen without saving the changes, click Cancel.

While it appears by looking at the spreadsheet that the Blank Lines element has no
alterable default values, it can be seen on the Modify Database Defaults screen that the
default number of blank lines can be changed (Figure 4.4-27).

Figure 4.4-27 Modify Database Defaults Screen for Blank Lines

Once the elements have been set up with the desired default values, on the System
Defaults Maintenance screen click Return to Admin in the lower left corner of the screen
or click Return on the menu bar to return to the Administration Menu. The administrator
can always modify the database defaults at any time while creating a database.

38
4.4.4 Define/Modify Codebook
The Define/Modify Codebook button is used to create and edit the codebook. This will be
discussed in greater detail in Chapter 6: Data Coding and Data Coding Queries.

4.4.5 Lock Database Structure


To prevent anyone from modifying the data structure for a database, click Lock Database
Structure on the Administration Menu (Figure 4.3-3). This will place a check in the box
labeled “Database Structure Locked” to the right of this button, and the wording on the
button itself will change to Unlock Database Structure.

To restore database structure modification privileges, click Unlock Database Structure on


the Administration Menu. This un-checks the box to the right of the command button,
and the button again reads Lock Database Structure.

4.4.6 Lock Data Entry


To prevent anyone from modifying existing data or adding new cases on the data entry
form, click Lock Data Entry on the Administration Menu (Figure 4.3-3). This will place a
check in the box labeled “Data Entry Locked” to the right of this button, and the wording
on the button itself will change to Unlock Data Entry. This disables the following buttons
on the data entry form: Add New, Cancel, Delete, Edit, and Save. These buttons and the
data entry form in general will be covered in detail in Chapter 5: Using an Existing
Database. To restore data entry and data modification privileges, click Unlock Data Entry
on the Administration Menu. This un-checks the box to the right of the command button,
and the button again reads Lock Data Entry.

4.4.7 Exit Administration


To exit the Administration Menu (Figure 4.3-3) and return to data entry form mode,
simply click the Exit Administration button at the bottom of the Administration Menu.

39
Chapter 5. Using an Existing Database

5.1 Opening a Database


After opening EZ-Text (Section 3.2), click File on the menu bar on the blank Home
Screen and select Open Database (Figure 5.1-1).

Figure 5.1-1 Selecting Open Database on the Home Screen

The Open window will appear displaying any existing database file names in the folder
where EZ-Text was installed (Figure 5.1-2). If the desired database is in another folder on
the computer, navigate to the folder where the database file is stored. Open the database
by clicking on the file name and then clicking Open. Alternatively, open the database by
double-clicking the file name in the Open window.

40
Figure 5.1-2 Selecting and Opening the Database File

If the database is password protected, the “Enter Database Password” prompt will appear
(Figure 5.1-3). The data entry form for the database will load, and the first record will
appear on the screen. If no records have been entered into the data entry form, a blank
form will appear (Figure 5.1-4).

Figure 5.1-3 Database Password Prompt

41
Figure 5.1-4 Blank Data Entry Form

For databases that contain data, the data entry form has five buttons on the lower left and
a drop-down list in the lower center for navigating through existing cases. These buttons
and drop-down list are discussed further in Section 5.2.4. The five buttons on the lower
right of the data entry form are used for data entry and are discussed throughout Section
5.2.

42
5.2 Performing Data Entry

5.2.1 Adding New Cases


Upon first opening the data entry form, the data entry fields are disabled. This is to
prevent users from accidentally altering existing data. When opening a blank database,
the only enabled button is the Add New button.

To begin data entry, click the Add New button. Alternatively, select Add New from the
Record menu. The data entry fields will become yellow and the Cancel and Save buttons
will become enabled (Figure 5.2-1). One can now enter data in the appropriate fields.

Figure 5.2-1 Data Entry Form After Clicking Add New

Please note that Respondent ID is a required numeric field for all databases and must be
unique for each record in a database. This field typically will be the first field of a data
entry form. However, other fields can be placed above it.

To move to the next field on the form, use either the Tab key on the keyboard or click in
the next field with the mouse. To move back to a field that has already been entered or
was accidentally skipped over, the Shift-Tab key combination can be used to move

43
backward one field at a time. Alternatively, using the mouse, click directly in the desired
field.

5.2.2 Using Field Types


There are several field types that can be used for data entry. The following describes the
field types, how to enter data in these fields, and data entry restrictions for each field
type.

Numeric Fields: Depending upon how they are defined in Admin (Section 4.4.2(g)),
numeric fields will accept either whole numbers or decimal numbers. A pop-up window
will also appear if an entered value lies outside of a pre-determined range of acceptable
values (Section 5.2.3).

Date Fields: Dates can be entered into date fields in a number of ways. For example, a
date of May 12, 1984 can be entered into the data entry form as:

5/12/84 5-12-84 May 12, 1984 May 12, 84

Each of the above will be stored in the database (and will appear on the data entry form
once it is saved) as 05/12/1984. If an invalid date format is entered into a date field, a
pop-up window describing the error will appear when attempting to save the data
(Section 5.2.3).

Short Text Fields: Short text fields are fixed length. They can contain numbers, letters,
and characters. If the number of characters exceeds the pre-defined length, a pop-up
window describing the error will appear when attempting to save the data (Section 5.2.3).

Long Text Fields: Long text fields can contain numbers, letters, and characters. There is
no length limit for long text fields. Text may be typed directly into these fields after
turning on the EZ-Text "Edit" mode by clicking the command button on the lower right
of the Data Entry Form screen. Likewise, changes can be made directly to the text while
in EZ-Text "Edit" mode. Alternatively, users may wish to use their computer mouse
along with standard Windows® features to block, copy, and paste segments of text
selected from other computer files. For example, if an individual interview transcript was
typed into a Microsoft Word® file, the user may copy a specific text segment and paste it
into an appropriate location within their EZ-Text database.

Drop-down List Fields: Select a value from a drop-down list by clicking on the arrow
on the right of a drop-down list field and clicking the appropriate response on the list.

Option List Fields: Select the appropriate response in an option list by clicking the
button to the left of the correct response.

44
5.2.3 Saving a Record to the Database
Once all fields have been entered for a record, click the Save button or select Save from
the Record menu to save the record to the database. At this time, any data entry errors
will be flagged, and a pop-up window will appear identifying the errors. These data
checks are described below.

Data checks: Some field types have checks associated with them that are triggered when
the Save button is clicked. If a value fails one of these checks, a pop-up window appears
alerting the data entry personnel of the erroneous entry by providing the question number,
variable caption, and a description of the error (Figure 5.2-2).

Figure 5.2-2 Pop-up Window Alert

Clicking OK returns to the data entry form with the cursor in the affected field. Correct
the value on the data entry form window and click the Save button again. Error pop-up
windows will continue to appear until all checkable errors are resolved. When no more
EZ-Text-detectable data entry errors are found, the record is saved to the database.

45
After the record is successfully saved, the data entry fields become white again and are
disabled. The data for the saved record is displayed in the fields (Figure 5.2-3).

Figure 5.2-3 Data Entry Form with Data

To see fields not currently visible in the data entry form window, use the scroll bar along
the right side of the form window. Also note that the Delete and Edit buttons are now
enabled. These will be covered in Sections 5.2.5 and 5.2.6, respectively.

To continue entering new records, click the Add New button again, and follow the steps
in this section.

Exit without saving: At any time while entering a new record, the data entry person
may click the Cancel button to exit data entry mode for the current record without
saving the data. Alternatively, select Cancel from the Record menu.

46
5.2.4 Using Navigation Buttons
In addition to the Delete and Edit buttons, notice that once a case has been added to the
database, the navigation buttons on the bottom left of the data entry form are now
enabled. Use these buttons to maneuver between records in the database.

First button: Clicking the First button displays the first record in the database. If the user
is already on the first record and clicks this button, a window appears alerting the user
that the first record is already displayed (Figure 5.2-4).

Figure 5.2-4 Already at First Record Alert

Click OK to close the pop-up window.

Previous button: Click the Previous button to move backward through the database one
record at a time. If this button is clicked while the first record is displayed, a pop-up
window identical to the one displayed in Figure 5.2-4 appears. Click OK to close the
window.

Last button: Clicking the Last button displays the last record in the database. If this
button is clicked while the user is on the last record, a window appears alerting the user
that the last record is already displayed (Figure 5.2-5).

Figure 5.2-5 Already at Last Record Alert

Click OK to close the pop-up window.

Next button: Click the Next button to move forward through the database one record at a
time. If this button is clicked while the last record is displayed, a pop-up window
identical to the one displayed in Figure 5.2-5 appears. Click OK to close the window.

47
Go to button: The Go to button not only moves from one record to another but also
moves within a record from one data field to another. This is useful for long data entry
forms and databases with a lot of records. Clicking the Go to button opens the Go to
Menu (Figure 5.2-6).

Figure 5.2-6 Go to Menu

To move to a data field within the current record, click Go to Question Number. This
provides the Go to Question Number screen (Figure 5.2-7).

Figure 5.2-7 Go to Question Number Screen

Select the question number from the drop-down list and click Continue. The data entry
form will display that question number for the current record. To cancel the Go to
Question Number process, click Cancel.

To move directly to a record without having to browse through other records by using the
Previous and Next buttons, click Go to Respondent ID on the Go to Menu screen. This
provides the Go to Respondent ID screen (Figure 5.2-8).

Figure 5.2-8 Go to Respondent ID Screen

48
Select the respondent ID number from the drop-down list and click Continue. The
selected record will be displayed on the data entry form. To cancel the Go to Respondent
ID process, click Cancel.

Page Number drop-down list: Click the Page Number field to go to a particular page
number in a data entry form. Select the desired page number from the drop-down list.
The data entry form will display that page number for the current record.

Alternatively, these navigation features can be invoked from the menu bar. First,
Previous, Next, and Last can be found on the Record menu. Go to is its own menu option
on the menu bar and contains Question No, Respondent ID, and Page Number.

5.2.5 Deleting Records


Using the navigation buttons, locate the record to be deleted and click the Delete button
or select Delete from the Record menu. A prompt will appear to confirm deletion (Figure
5.2-9).

Figure 5.2-9 Confirm Deletion Screen

To delete the record, click Yes. This will permanently remove the record from the
database. To return to the data entry form without deleting the record, click No.

5.2.6 Editing Records


Records that have already been entered can be modified. Navigate to the desired record
and click Edit or select Edit from the Record menu. As with adding a new record, the
data fields will change color, signifying that they are now enabled for modification.
Either Tab to the desired field using the keyboard, click the field with the mouse, or use
the Go to Question Number navigation feature. After making the changes to the necessary
fields, click the Save button. As when saving a new record, EZ-Text will perform its
checks for data errors based on field constraints. Once the changes have been saved, the
data fields are disabled and appear white again. To exit edit mode without saving any of
the modifications, click the Cancel button.

49
5.3 Closing a Database
It is strongly recommended that a database be closed using the File menu, not by clicking
the X in the upper right hand corner of the screen. Using the File menu ensures that a
database is shut down properly and offers the option of creating a backup database. To
close a database, click the File menu and select Close Database (Figure 5.3-1).

Figure 5.3-1 Closing a Database Using the File Menu

50
The database will close and a prompt will appear asking whether to create a backup
database (Figure 5.3-2).

Figure 5.3-2 Create Backup File Screen

It is a good idea to create a backup file in case the working database becomes corrupted
or is accidentally deleted. Click Yes to create the backup database. A Save As dialogue
box appears (Figure 5.3-3).

Figure 5.3-3 Save As Dialogue Box for Creating Backup Files

The file extension for backup files is .bak. EZ-Text assigns a file name composed of the
database name and the current date. This file name can be modified if desired. To save
the backup database, click Save. To cancel the backup database process without saving
the backup file, click Cancel.

51
5.4 Restoring a Backup File
If it becomes necessary to use a backup file, the backup file must be restored. To do so,
select Restore Backup on the File menu (Figure 5.4-1).

Figure 5.4-1 Selecting Restore Back-up from File Menu

This invokes the Open dialogue box (Figure 5.4-2).

52
Figure 5.4-2 Open Dialogue Box for Restoring Backup Files

Select the backup file to restore and click Open. If the original file is still in the current
folder, a prompt will appear asking if the user wants to overwrite the existing file (Figure
5.4-3).

Figure 5.4-3 Overwrite Existing Database Prompt

To prevent the original file from being overwritten, click No. Otherwise, to finish
restoring the backup file, click Yes. A pop-up window appears informing the user that the
database was successfully restored (Figure 5.4-4).

Figure 5.4-4 Successful Restoration Pop-up

Click OK. The backup database is now ready for use.

53
5.5 Repairing a Database
Occasionally a database may become corrupt and need to be repaired. An example would
be when trying to open a database generates an error message. To repair a database,
select Repair and Compress Database on the File menu (Figure 5.5-1).

Figure 5.5-1 Repairing a Database

This invokes the Open dialogue box (Figure 5.5-2).

54
Figure 5.5-2 Open Dialogue Box for Repairing a Database

Select the file to repair and click Open. A pop-up window appears informing the user that
the database was successfully repaired (Figure 5.5-3).

Figure 5.5-3 Successful Database Repair Pop-up

Click OK. The database has been repaired and is now ready for use again.

55
5.6 Exiting EZ-Text
To exit EZ-Text, click Exit on the File menu, or click the X in the upper right hand corner
of the program (Figure 5.6-1).

Figure 5.6-1 Exiting EZ-Text

Exit EZ-Text only after first closing the database file as described in Section 5.3.

56
Chapter 6. Data Coding and Data Coding Queries

6.1 Overview of Data Coding in EZ-Text


After the qualitative data have been entered into an EZ-Text database, users must decide
on an approach to data coding. Qualitative data coding entails identifying the themes in
text passages or segments (Bernard 2006; Miles and Huberman 1994; Ryan and Bernard
2003). Themes may include beliefs, experiences, or opinions the respondent was trying to
communicate in response to the interviewer's questions. Different respondents may
address similar themes but state their ideas in different ways, or they may hold entirely
different views. Qualitative data coding requires coders to read accurately and
comprehend similarities and differences across text passages, regardless of how
respondents express themselves (Hruschka et al. 2004; MacQueen et al. 1998). Text
passages containing identical themes are coded the same way; passages containing
different themes receive different codes.

Regardless of the software package used, there are two major steps to qualitative data
coding: 1) text segmentation, and 2) code creation and assignment. Text segmentation
requires dividing the text into manageable segments or “chunks” (for example, a
paragraph or a few sentences, depending on the user's needs). Other qualitative software
programs require the user to place segmentation markers in the text manually before
assigning codes. However, EZ-Text can be a basis for organizing data, because it is
designed specifically for semi-structured data collection instruments. As discussed in
earlier sections, EZ-Text helps users create templates that correspond to each question on
a questionnaire. EZ-Text uses these templates to create a single text passage or segment
for each question asked of each respondent in the sample. For example, if there are 10
interview questions and 100 respondents, EZ-Text would organize and store the data in
1,000 text segments. In the EZ-Text system, the database structure closely corresponds to
the questions included in the original questionnaire. Data are "pre-segmented" during
data entry. No further effort is needed to segment the text. This can save an enormous
amount of time in database management, and greatly facilitate question-by-question
analysis.

Qualitative data analysts also typically develop a set of codes that correspond to each
theme in a given segment. This is an inductive task, based on what respondents have said.
Analysts compile a list of codes in a codebook and generally revise the codebook
numerous times until it accurately reflects the themes in the database. In general, users
should strive to develop codebooks with lists of carefully defined codes that are all
inclusive, yet mutually exclusive in terms of summarizing themes in the text (Boyatzis
1998; MacQueen et al. 1998). Users should also conduct formal intra- and inter-coder
reliability assessments to make sure that coders are using the final codebook consistently
and correctly (Carey et al. 1996; Hruschka et al. 2004).

EZ-Text has an array of features to help users develop a sophisticated codebook they can
use to assign sets of codes to responses, reflecting the corresponding themes. Users can
create as many codes as they need. For example, a moderately large database (say 70
respondents, 30 questions each) may require 200 to 400 separate codes. This depends,

57
however, on the desired level of coding detail and the degree of thematic diversity in the
original data. As described in Section 6.4 of this chapter, EZ-Text users can determine
which sets of these codes reflect the themes in each text segment.

For each separate code, the EZ-Text codebook contains seven major elements (Figure
6.2-2). These permit creation of highly detailed codebooks:
1. A unique seven-character mnemonic code name.
2. A 50-character brief definition of the code.
3. A full definition of the code (unlimited length).
4. Instructions to the coder on when to use the code (unlimited length).
5. Instructions to the coder on when not to use the code (unlimited length).
6. An example of a text passage copied from the database showing one or more
situations where the text contains the theme corresponding to the code.
7. The option of assigning a parent code.
"Parent codes" typically are codes that have conceptually broad and inclusive definitions.
A parent code may have one or more "child codes," each of which has a more specific
and less inclusive definition. Child code definitions should be logically subsumed as
subtopics within their corresponding parent code.

When developing names for codes, it is recommended that users assign names that will
assist coders in easily recognizing their intended use. For example, in a study of cancer
prevention, users might create a code called "NoSmoke" to refer to the idea that a
respondent believes that not smoking tobacco will help reduce their chance of developing
lung cancer. In some circumstances, users may wish to give related codes similar
prefixes. For example, suppose that a researcher wants to create a set of codes that are
only supposed to be used for responses to Question 1 and Question 2 on their
questionnaire. Code names for the first question all might start with "Q1" (e.g.,
"Q1_abc," "Q1_def," etc.), while question 2 codes might be "Q2_ghi," "Q2_jkl," etc.
Because EZ-Text displays the codebook in alphabetical order by the code name, a coder
could quickly find all the question-specific codes listed together, rather than needing to
search through the entire codebook.

Users are not required to use EZ-Text to code the data in EZ-Text database files. As with
all software tools, EZ-Text has its limitations. Other software packages may better suit a
user's particular coding needs. To use a different software package for coding and
analysis, use the EZ-Text Data Export Agent (Section 8.3) to export the text data into a
new ASCII file, and then import the data into another qualitative data analysis software
package. In this approach, the researcher could use EZ-Text merely as a qualitative data
entry and database management tool. In circumstances where EZ-Text's analysis features
do not match the user's needs and research goals, this might be the best choice.

58
6.2 Defining Codes in the Codebook
Any combination of codes can be assigned to a specific response. As mentioned
previously, for each entry in the codebook, the following information can be defined: a
seven-character name, a short definition, a long definition, further instructions for when
to apply or not apply the code, a text passage illustrating an example where the code
should be used, and a parent code.

Step 1. Open the database, then go to the Administration Menu.

Step 2. Click on the Define/Modify Codebook button.

Step 3. Click on the Add New Code command button (Figure 6.2-1).

Figure 6.2-1 Add New Code Command Button

59
Step 4. The Add New Code screen will appear (Figure 6.2-2). From this screen, enter the
desired code, a brief description of the code, a full description of the code, a use for the
code, a way in which the code should not be used, and an example of the code in use, and
select a parent code from the drop-down list of codes that have already been created. EZ-
Text does not require that all seven pieces of information be entered into the Add New
Code screen, but it is recommended (Boyatzis 1998; Hruschka et al. 2004; MacQueen et
al. 1998). Providing all of the information known about a specific code makes it easier
for multiple coders to understand its meaning. The Brief Description field is required.

Figure 6.2-2 Add New Code

60
Step 5. After entering a code and the rest of the information on the Add New Code
screen, click on the Save command button (Figure 6.2-3). Repeat steps 3, 4, and 5 for
each code to be entered.

Figure 6.2-3 Saving a Code in the Codebook

61
6.3 Modifying the Codebook
There are two ways to modify the codebook.

Method 1. Modifying the Codebook from the Administration Menu

Step 1. Open the database, then go to the Administration Menu.

Step 2. Click on the Define/Modify Codebook command button.

Step 3. To modify an existing code, click on the code to be modified. This will highlight
the code, parent, and brief description of the code (Figure 6.3-1).

Figure 6.3-1 Highlighting a Code

62
Step 4. Click on the Modify Code command button (Figure 6.3-1). This allows the user to
edit a code, the definitions of a code, the uses or examples of a code, and the parent code
designation on the Modify Code screen (Figure 6.3-2).

Figure 6.3-2 Modify Code

Step 5. Once the editing of the code is complete, click on the Update command button
(Figure 6.3-2). Repeat this process to modify as many codes as needed.

63
Deleting a code: Follow steps 1 through 3 in the instructions for modifying a code above
to highlight the code to be deleted. Click on the Delete Code command button. This will
open a deletion confirmation screen (Figure 6.3-3).

Figure 6.3-3 Delete Code

Click on the Delete command button to delete the code displayed on the delete code
screen. Caution should be used when deleting codes. If a code has been used and is
subsequently deleted from the codebook, all previous points of assignment to response
passages are also removed from the database.

64
Viewing the Parent/Child Codes Hierarchically Using Tree View: Follow steps 1 and
2 in the instructions for modifying a code above. Click on the Tree View option on the top
menu bar to access the Tree View screen (Figure 6.3-4).

Figure 6.3-4 Codebook Tree View Screen

Note that it is also possible to add new codes and edit or delete existing codes while in
Tree View. Click on the Return command button to return to the Codebook Maintenance
screen.

Method 2. Modifying the Codebook from the Coding Form

Step 1. Open the database.

Step 2. On the Data Entry Form screen, click on the Coding option on the top menu bar.

Step 3. Click on the Modify Codebook option on the top menu bar on the Coding Form
screen (Figure 6.3-5).

65
Figure 6.3-5 Modifying Codebook in Coding Form

Step 4. A pop-up screen will appear prompting the user to enter a Codebook Password.
Enter the codebook password and click on the Continue command button (Figure 6.3-6).

Figure 6.3-6 Codebook Password

See Section 4.4.1 for information on the default codebook password and changing the
codebook password.

66
Step 5. The Codebook Maintenance screen will appear (Figure 6.3-7).

Figure 6.3-7 Codebook Maintenance

Note that this screen is nearly identical to the screen for adding new codes and modifying
and deleting existing codes, topics discussed in Section 6.2 and this section. Click the
appropriate command buttons to add new codes, delete existing codes, or modify the
current codebook definitions. Caution should be used when deleting codes. If a code has
been used and is subsequently deleted from the codebook, all previous points of
assignment to response passages are also removed from the database. After changes to
the codebook are made, click on the Return to Coding Form command button to go back
to the Coding Form and continue with the coding process.

67
6.4 Coding Responses
Step 1. Open the database.

Step 2. On the Data Entry Form screen, click on the Coding option on the top menu bar
to open the Coding Form screen (Figure 6.4-1).

Figure 6.4-1 Coding Form

Step 3. Using the drop-down lists, select the question number and respondent ID to code.

Step 4. Boxes that display the original Question and corresponding Response text
passage for the selected question number and respondent are also present. Below these
boxes are two more boxes that list Available Codes and Assigned Code(s). The Available
Codes list is used to scroll through the code names and short definitions in the codebook.
After reading and identifying themes in the response passage, highlight the appropriate
code in the Available Codes list and click on the top arrow button to move the code to the
Assigned Code box. If the forced text selection is checked in the Administration Menu
(see Figure 4.3-3 and Section 4.4.1), the section of text relating to the code must be
highlighted before the code is selected. Repeat this process until the Assigned Code box
contains a list of codes corresponding to all the themes in the response. Similar steps,

68
invoking the bottom arrow button after selecting the code to unassign, are used to remove
a code from the Assigned Code box.

Viewing details of a code: To view a specific code along with the code's definitions and
uses, double-click on the code from the Available Codes box. The Code Definition screen
will appear (Figure 6.4-2).

Figure 6.4-2 Code Definition

To return to the Coding Form, click on the Return command button on the Code
Definition screen.

Note: The Available Codes list is displayed in alphabetical and numerical order based on
the seven-character code names. Users can have as many codes in the codebook as they
need (just a few codes or even hundreds of codes). A scrollbar automatically appears on
the right side of the Available Codes list when there are more codes than fit in the display
box on the screen. To assist in accessing known codes quickly, there is also a Seek Code
field below the Available Codes list. As the code name is typed into this field, the first
available code that matches the text being entered is highlighted. When more characters
are entered, the highlight will move to the next code that matches. Note that the code
name is case sensitive.

69
To return to the Data Entry Form, click on the Return command button on the Coding
Form screen.

6.5 Viewing and Printing the Codebook


The codebook can be viewed or printed in various formats.

Step 1. Go to the Coding Form as instructed in Section 6.4.

Step 2. Click on Reports on the top menu bar and select the type of report to be printed
(Figure 6.5-1).

Figure 6.5-1 Report Selection

Codebook Summary Report (sorted by code) displays all available codes and their short
definitions, alphabetized by code name.

Codebook Summary Report (sorted by brief description) displays all available codes and
their short definitions, alphabetized by brief description.

70
Codebook Summary Report (sorted by tree view) displays all available codes and their
short definitions, alphabetized by parent codes, then their respective child codes.

Full Codebook Report displays the entire codebook.

Coding Frequency Report Agent displays a coding frequency report based on options
including confidence interval, respondent IDs, long text elements, and codes to compare
chosen by the user. Selecting this agent from the Reports menu opens a series of screens
that prompt for the desired options (Figures 6.5-2 through 6.5-5).

Figure 6.5-2 Coding Frequency Report Agent Confidence Interval

71
Figure 6.5-3 Coding Frequency Report Agent Respondent IDs

Figure 6.5-4 Coding Frequency Report Agent Long Text Elements

72
Figure 6.5-5 Coding Frequency Report Agent Codes to Compare

Step 3. Once the report type is selected, it is displayed on the Report Preview screen
(Figure 6.5-6). To print the report, click on either the Print Page command button to print
the current page or Print All command button to print the entire report. In addition, the
report can be saved in .rtf (Rich Text File) format by selecting the Save to File option on
the top menu bar of the Report Preview screen.

73
Figure 6.5-6 Report Preview

6.6 Exporting and Importing the Codebook


EZ-Text has the capability of exporting and importing codebooks. This feature allows
multi-site projects to create one universal codebook for use at all sites. The researcher
responsible for the codebook can add, modify, or delete codes in the codebook, export the
codebook, and then send it to the sites. Once the codebook is received, the sites can
import the codebook into their database.
Sometimes researchers conduct a series of related studies over a period of time, and they
might use similar questionnaires in these different projects. Although it is unlikely that
the same codebook could be used without revision across all their studies, a codebook
developed for one study might be a useful starting point in subsequent studies. Rather
than entirely re-creating the codebook for each new study, in some instances researchers
might use this EZ-Text feature to import a copy of an old codebook from a previous
study into their new study's database. Appropriate edits could be made to reflect the
needs of the new study.

74
6.6.1 Creating a Codebook Export File
Step 1. Go to the Coding Form as instructed in Section 6.4.

Step 2. Select the Modify Codebook option on the top menu bar (Figure 6.3-5). The
Codebook Maintenance screen will appear (Figure 6.3-7).

Step 3. Click on the Export Codebook option on the top menu bar.

Step 4. A Save As dialog box will appear. Enter the file name (which always has the .cdb
file name extension) and click Save (Figure 6.6-1).

Figure 6.6-1 Naming a Codebook Export File

75
6.6.2 Importing a Codebook
An existing codebook associated with a database can be replaced using the Import
Codebook function available on the File Menu on the Data Entry Form. This feature
helps project managers of multi-site studies ensure that all sites use the same codebook.

To do this, the researcher responsible for the codebook should:


(1) Be the only person who knows the Administration and Codebook passwords
(2) Facilitate cross-site consensus on codebook revisions (by conference call or e-
mail)
(3) Distribute revised versions of the codebook file to each coder as needed.
Coders at each site then replace their old codebook with the revised version by using the
Import Codebook feature. To minimize confusion about the current version of the
codebook, EZ-Text automatically provides a date and time field. To view, click on the
About option on the top menu bar of the Data Entry Form screen to access the About
CDC EZ-Text screen (Figure 6.6-2).

Figure 6.6-2 About CDC EZ-Text

Note that the EZ-Text program version number and the date that the database template
was last modified are also indicated on this screen.

To import the codebook:

Step 1. Open the database.

Step 2. On the Data Entry Form screen, click on the File option on the top menu bar and
select Import Codebook (Figure 6.6-3).

76
Figure 6.6-3 Importing the Codebook

Step 3. An Open dialog box will appear. Navigate to the location of the codebook to
import. Highlight the desired codebook and click Open (Figure 6.6-4).

77
Figure 6.6-4 Selecting a File for Codebook Import

Caution: A pop-up box will appear informing the user that any codes already assigned
that do not appear in the codebook being imported will be permanently unassigned and
erased from the database (Figure 6.6-5).

Figure 6.6-5 Import Codebook Warning Pop-up

Click Yes to continue or No to cancel importation. If Yes is selected, another pop-up box
will appear informing that the codebook has been imported successfully.

78
6.7 Coding Queries and Query Reports from the Coding Form
The Query option on the Coding Form's top menu bar helps conduct Boolean searches of
the coded responses and their relationship to the respondent ID and the questions. Coding
queries are designed using combinations of ID numbers, open-ended question numbers,
and assigned codes. Query results can be browsed on screen or viewed in a report format
that can be printed or saved. The search feature may be used to find and replace
inappropriately assigned codes. After coding is complete, it can be helpful for selecting
text passages to use in ethnographic case studies.

Note: A coding query is different than a database query, which conducts Boolean
searches of the database on user-defined data fields. Database queries are discussed
in further detail in Chapter 7.

6.7.1 Coding Queries from the Coding Form


To perform a simple query:

Step 1. Go to the Coding Form as instructed in Section 6.4.

Step 2. Select the Query option on the top menu bar (Figure 6.4-1). The Coding Query
Screen will appear (Figure 6.7-1).

Figure 6.7-1 Coding Query Screen

79
Step 3. Click on the desired Boolean operator (And/Or) on the top left side of the screen.

Step 4. Click on the Add Criteria command button on the top right side of the screen to
define the query using the Add Query Criteria pop-up box (Figure 6.7-2).

Figure 6.7-2 Add Query Criteria

Step 5. Choose Field to Query (Assigned Code, Question Number, or Respondent ID),
How to Compare, and Value to Query by using the drop-down lists, and then click on the
Continue command button.

Step 6. Repeat steps 4 and 5 to define the desired number of queries for the Boolean
operator chosen in step 3.

Additional steps to follow to perform a complex query:

Step 7. Click on the desired Boolean operator (And/Or/Not) in the middle of the screen.

Step 8. Click on the desired Boolean operator (And/Or) on the bottom left side of the
screen.

Step 9. Click on the Add Criteria command button on the bottom right side of the screen
to define the query using the Add Query Criteria pop-up box (Figure 6.7-2).

Step 10. Choose Field to Query (Assigned Code, Question Number, or Respondent ID),
How to Compare, and Value to Query by using the drop-down lists, and then click on the
Continue command button.

Step 11. Repeat steps 9 and 10 to define the desired number of queries for the Boolean
operator chosen in step 8. The Coding Query Screen will now display the complex query
defined (Figure 6.7-3).

80
Figure 6.7-3 Coding Query Screen for Complex Query

Step 12. To view the query results, click on the Browse Records command button. This
returns to the Coding Form screen with only the records that meet the query criteria
available to view (Figure 6.7-4).

81
Figure 6.7-4 Viewing Coding Query Results

While browsing, use the command buttons -First, Previous, Next, Last- to navigate
through the results.

Step 13. To return to the full, unfiltered database, click on the Full Database command
button at the bottom of the screen.

82
6.7.2 Coding Query Reports from the Coding Form
Step 1. Perform steps 1 through 11 in Section 6.7.1.

Step 2. Click on the Create Report command button on the Coding Query Screen (Figure
6.7-1) to open the Report Options screen (Figure 6.7-5).

Figure 6.7-5 Report Options for Coding Query Results

Step 3. Select the fields to list in the report from the Available Fields list (Respondent ID,
Question Number, Assigned Code, and/or Full Response). To select all available fields,
click on the double-arrow; to select one or more fields, highlight the field, then click on
the single-arrow. To deselect fields, highlight a field in the Selected Fields list and click
on either the double-arrow to deselect all fields or the single-arrow to deselect the
highlighted field.

Step 4. After selecting the fields to report, identify the Report Options to be included in
the header of the report (Report Title, Page Numbers, Today's Date, and Query Logic).

83
Step 5. Click on the Preview command button to view the report on the Report Preview
screen (Figure 6.7-6) before printing or saving.

Note: If an extremely large report is requested, EZ-Text will generate a message saying
that the report cannot be previewed on screen. If that happens, users will only be allowed
to send the report directly to the printer.

Figure 6.7-6 Coding Query Report Preview

Step 6. Click on either the Print Page command button to print the current page or Print
All command button to print the entire report. In addition, the report can be saved in .rtf
(Rich Text File) format by selecting the Save to File option on the top menu bar of the
Report Preview screen.

84
Chapter 7. Database Queries and Query Reports from the Data Entry Form

7.1 Database Queries from the Data Entry Form


The Query option on the Data Entry Form menu bar assists in conducting Boolean
searches of the database. Database queries are designed using combinations of fields in
the database. Query results can be browsed on screen or viewed in a report format that
can be printed or saved.

Note: A database query is different from a coding query, which conducts Boolean
searches of coded responses and their relationship to the respondent ID and the
questions. Coding queries are discussed in further detail in Section 6.7.

To perform a simple query:

Step 1. On the main Data Entry Form, select the Query option on the top menu bar. The
Record Query Screen will appear (Figure 7.1-1).

Figure 7.1-1 Record Query Screen

85
Step 2. Click on the desired Boolean operator (And/Or) on the top left side of the screen.

Step 3. Click on the Add Criteria command button on the top right side of the screen to
define the query using the Add Query Criteria pop-up box (Figure 7.1-2).

Figure 7.1-2 Add Query Criteria

Step 4. Choose database Field to Query, How to Compare, and Value to Query by using
the drop-down lists, then click on the Continue command button.

Step 5. Repeat steps 3 and 4 to define the desired number of queries for the Boolean
operator chosen in step 2.

Additional steps to follow to perform a complex query:

Step 6. Click on the desired Boolean Operator (And/Or/Not) in the middle of the screen.

Step 7. Click on the desired Boolean operator (And/Or) on the bottom left side of the
screen.

Step 8. Click on the Add Criteria command button on the bottom right side of the screen
to define the query using the Add Query Criteria pop-up box (Figure 7.1-2).

Step 9. Choose database Field to Query, How to Compare, and Value to Query by using
the drop-down lists, then click on the Continue command button.

Step 10. Repeat steps 8 and 9 to define the desired number of queries for the Boolean
operator chosen in step 7. The Record Query Screen will now display the complex query
defined (Figure 7.1-3).

86
Figure 7.1-3 Record Query Screen for Complex Query

Step 11. To view the query results, click on the Browse Records command button. This
returns to the Data Entry Form screen with only the records that meet the query criteria
available to view (Figure 7.1-4).

87
Figure 7.1-4 Viewing Data Query Results

While browsing, use the command buttons -First, Previous, Next, Last- to navigate
through the results.

Step 12. To return to the full, unfiltered database, click on the Full Database command
button at the bottom of the screen.

88
7.2 Database Query Reports from the Data Entry Form
Step 1. Perform steps 1 through 10 in Section 7.1.

Step 2. Click on the Create Report command button on the Record Query Screen (Figure
7.1-3) to open the Report Options screen (Figure 7.2-1).

Figure 7.2-1 Report Options for Data Query Results

Step 3. Select the fields to list in the report from the Available Fields list. To select all
available fields, click on the double-arrow; to select one or more fields, highlight the
field, then click on the single-arrow. To deselect fields, highlight a field in the Selected
Fields list and click on either the double-arrow to deselect all fields or the single-arrow to
deselect the highlighted field.

Step 4. After selecting the fields to report, identify the Report Options to be included in
the header of the report (Report Title, Page Numbers, Today's Date, and Query Logic).
To view the Respondent IDs of records that will appear in the report, click on the drop-
down list below the Report Options.

89
Step 5. Click on the Preview command button to view the report on the Report Preview
screen (Figure 7.2-2) before printing.

Note: If an extremely large report is requested, EZ-Text will generate a message saying
that the report cannot be previewed on screen. If that happens, users will only be allowed
to send the report directly to the printer.

Figure 7.2-2 Data Query Report Preview

Step 6. Click on either the Print Page command button to print the current page or Print
All command button to print the entire report. In addition, the report can be saved in .rtf
(Rich Text File) format by selecting the Save to File option on the top menu bar of the
Report Preview screen.

90
Chapter 8. EZ-Text Agents

The Agents Menu, located on the top menu bar on the Data Entry Form, guides the user
through the following tasks:

(1) Subset Agent (copying a subset of records to a new file)


(2) Merge Agent (merging files)
(3) Data Export Agent (creating ASCII text files and numeric matrices)
(4) Reliability Agents (checking inter- and intra-coder reliability, see Chapter 9)
(5) Spell Check Agent (checking spelling in data file)

8.1 Copying a Subset of Records with the Subset Agent


The Subset Agent allows selection of a subset of the records and copies these records to a
new EZ-Text database file. The database can be subset by Respondent ID. For a more
specific subset, such as a subset of the database based on values in a field, use the
Database Query function (see Section 7.1), and then use the subset agent on the current
result set (see Example of using the Subset Agent to subset by Current Result Set which
can be found later in this section).

For example, the Database Query could be used to select only males from a larger data
set that contains records for both men and women, and gender is identified for each
record with a drop-down list variable. When viewing the Database Query results, the
Current Result Set feature is enabled. Users would then be able to save all male records
to an entirely new EZ-Text database file with a new filename. As another example,
suppose data had been collected in three cities and that interview location is recorded in a
drop-down list variable. These EZ-Text features could be used to create a new database
file containing records from just one city.

91
Example of using the Subset Agent to subset by Respondent ID:

Step 1. On the Data Entry Form screen, click on the Agents option on the menu bar and
then click on Subset Agent (Figure 8.1-1) to begin the subset agent process (Figure 8.1-2).

Figure 8.1-1 Selecting Subset Agent

92
Figure 8.1-2 Subset Agent Step 1

Select the Respondent IDs option and click on the Next command button.

93
Step 2. On the second screen, select the desired respondent ID numbers from the
Available Respondent IDs and click on the Next command button (Figure 8.1-3).

Figure 8.1-3 Subset Agent Step 2

94
Step 3. The third screen lists subset options (Figure 8.1-4).

Figure 8.1-4 Subset Agent Step 3

Select whether to Copy Codebook to New Database or to Copy Coding Data (and the
codebook) to New Database (only enabled after the database has been coded). Continue
by clicking on the Next command button.

95
Step 4. On the fourth screen, enter a file name for the subset database (different from the
original database), including the .ezt extension, and then click on the Finish command
button (Figure 8.1-5).

Figure 8.1-5 Subset Agent Step 4

Step 5. A pop-up box will indicate that the database subset was successful. Click on the
OK command button to return to the Data Entry Form.

Example of using the Subset Agent to subset by Current Result Set:

After a database query is executed, the data can be subset to a new database by using the
Current Result Set option in the subset agent. Note: The Current Result Set feature is
only enabled while viewing the search results after conducting a database query.

Step 1. Follow steps 1 through 11 in Section 7.1 to access the query result Data Entry
Form screen (Figure 7.1-4).

96
Step 2. Click on the Agents option on the menu bar and select Subset Agent to begin the
subset agent process (Figure 8.1-6).

Figure 8.1-6 Subset Agent from Query Results Screen

Select the Current Result Set option and click on the Next command button.

Step 3. Follow steps 3 through 5 from the Example of using the Subset Agent to subset by
Respondent ID instructions found earlier in this section. Note that step 2 (Selecting
Respondent IDs) from that section is omitted in the Current Result Set procedure since
the subset selection was completed using the query function.

97
8.2 Merging Files with the Merge Agent
The Merge Agent is used to append another EZ-Text file into an open database (for
example, merging files containing two sets of interviews collected in two cities in a
multi-site research project). The merge process is successful only if two conditions are
met:

1. There is no overlap in the ID numbers used in the two files, and


2. Identical template versions have been used to create the two files.

To ensure that data files created on separate computers can be merged, project managers
should assign a different ID number series to each interviewer. In addition, all
interviewers should check the date information for their copy of the database, available
on the About CDC EZ-Text pop-up box (Figure 6.6-2) accessed by clicking on the About
option on the top menu bar of the Data Entry Form screen. By comparing the date of the
last template modification, project managers can verify that all interviewers have exactly
the same revision on their computers before data entry begins.

To merge data:

Step 1. Open one of the databases to be merged. On the Data Entry Form top menu bar,
click the Agents option, and then select Merge Agent (Figure 8.2-1).

Figure 8.2-1 Selecting Merge Agent

98
Step 2. On the first screen, select a different database file to merge into the currently
open database and click on the Next command button (Figure 8.2-2).

Figure 8.2-2 Select Database to be Merged

Note: Even though the two files being merged were created from the same template, they
should have different file names, such as file1.ezt and file2.ezt. It is necessary for both
files to have the same file structure.

99
Step 3. On the next screen, select to merge coding data into the existing database (only
enabled after the database has been coded) if desired by clicking the selection box (Figure
8.2-3).

Figure 8.2-3 Set Merge Database Options

Click on the Finish command button.

Note: If the Merge Coding Data option is chosen and there is any discrepancy between
the codebooks, EZ-Text generates an error message and the two files will not be merged.
Users can verify that the codebooks are the same by comparing the date information for
the codebooks in the two files on the About CDC EZ-Text pop-up mentioned earlier in
this section.

Step 4. A pop-up box will indicate that the database merge was successful. Click on the
OK command button to return to the Data Entry Form.

100
8.3 Creating ASCII Text Files and Numeric Matrices with the Data Export Agent
EZ-Text provides numerous options for identifying and exporting data into various
formats useful for different purposes. Several of these options are discussed in earlier
parts of this User’s Guide. For example, refer to Section 6.7 to conduct Coding Queries
(allows a user to search for combinations of codes assigned to the open-ended text
passages in Long Text fields and then view the results on screen, create reports for
printing, or save the results to word-processor readable files). See Chapter 7 for Database
Queries (allows one to search contents of structured response variables or text passages in
Long Text fields within the database and then view the results on screen, create reports
for printing, or save the results to word-processor readable files). Also refer to the Subset
Agent (see Section 8.1), which allows the user to select specific individual records, either
by ID number or by user-specified search criteria, and create new database files readable
by EZ-Text.

In addition to these previously described methods, EZ-Text also has a separate Data
Export Agent. This function is accessed by clicking on the Agents option on the top menu
bar of the Data Entry Form. The Data Export Agent can be used to export information
from the database in one of four additional formats:
1. Export short data fields and coding data to a comma-delimited text file for use
with a statistical package such as SAS®
2. Export short data fields and coding data to a tab-delimited text file for use with
SPSS®
3. Export questions, responses, and assigned codes to a text file
4. Export short data fields (without coding data) to a comma-delimited text file
The first, second, and fourth data export format options produced by the Data Export
Agent create files containing ASCII comma-delimited or tab-separated matrices. The
matrix files can be imported later into a spreadsheet or statistical software for further
analysis. By use of a common ID number, researchers might link EZ-Text matrices with
other quantitative data generated for the same respondents with another program such as
SPSS® (e.g., merging an EZ-Text matrix with a second file containing socioeconomic,
medical, psychometric, or other similar variables). In creating the matrices, users request
combinations of ID numbers, interview information variables, and, for the first two
format options, codes assigned to open-ended question responses. The comma-delimited
output file contains one row for each requested ID, and one column for each of the
selected short data field variables and codes (Figure 8.3-1).

101
Figure 8.3-1 Example of ASCII Comma-delimited Matrix as Viewed in Notepad

Matrix cell values either contain the data that were entered for the short data field
variables, or they contain ones and zeroes which indicate whether or not a code was
assigned to an open-ended question response. Each column is labeled with the original
short data field variable name or a slightly modified version of the code name used in the
codebook. When requesting coding data for more than one open-ended question, EZ-Text
places single letters (A, B, C) as a prefix to the code name. For example, if an EZ-Text
codebook contains three codes named CODE1, CODE2, and CODE3, each code might
be assigned to responses for either or both Question 1 and Question 2. Matrix column
labels for the two questions in this example would be ACODE1, ACODE2, ACODE3,
BCODE1, BCODE2, and BCODE3. Because codes in the codebook can be no longer
than seven characters, insertion of the single letter prefixes yields column labels no
longer than eight characters. Since the matrices may be imported into statistical software,
this convention corresponds to the eight character variable name limit within some
statistical programs.

The third format option copies some or all of the data to an ASCII text file. This file can
be viewed in a word processing program or imported into other qualitative data analysis
software. The agent's dialog screens help users specify the desired content of the file by
requesting combinations of ID numbers, open-ended question responses, and short data
field variables.

102
To export data:

Note: Not all steps apply to all data export format options. Pay special attention to
any notes associated with each step below.

Step 1. Open the database to be exported. On the Data Entry Form top menu bar, click
the Agents option, and then select Data Export Agent (Figure 8.3-2).

Figure 8.3-2 Selecting Data Export Agent

103
Step 2. Select the type of data to export and click on the Next command button. (Figure
8.3-3)

Figure 8.3-3 Select Type of Data to Export

104
Step 3. Select the respondent IDs to export from the Available Respondent IDs list and
click on the Next command button (Figure 8.3-4).

Figure 8.3-4 Select Respondent IDs

Note: When option four is selected in step 1, the next step in the data export process
is selecting data fields. Skip to step 5 below.

105
Step 4. (Only applies when option one, two, or three is selected in step 1.) Select the
long text questions of interest from the Available Long Text Questions list and click on
the Next command button (Figure 8.3-5).

Figure 8.3-5 Select Long Text Questions

106
Step 5. Select the data fields to export from the Available Data Fields list and click on
the Next command button (Figure 8.3-6).

Figure 8.3-6 Select Data Fields

Note: When either option three or four is selected in step 1, this is the last step in the
data export process. Click on the Finish command button at this point and skip to
step 7 below.

107
Step 6. (Only applies when option one or two is selected in step 1.) Select the codes to
export from the Available Codes list and click on the Finish command button (Figure 8.3-
7).

Figure 8.3-7 Select Codes

Step 7. A Save As dialog box will pop-up (Figure 8.3-8) prompting for a filename for the
text file.

Figure 8.3-8 Data Export Agent Save As

108
Enter a filename (the default is export.txt), and click on the Save command button. A
pop-up box will appear to indicate that the data export was successful. Click on the OK
command button to return to the Data Export Agent.

8.4 Checking Spelling with the Spell Check Agent


The Spell Check Agent checks the spelling within the text of one or more questions for
one or more Respondent IDs.

To check spelling:

Step 1. Open the database to be spell checked. On the Data Entry Form top menu bar,
click the Agents option, and then select Spell Check Agent (Figure 8.4-1).

Figure 8.4-1 Selecting Spell Check Agent

109
Step 2. Select the respondent IDs to spell check from the Available Respondent IDs list
and click on the Next command button (Figure 8.4-2).

Figure 8.4-2 Spell Check Agent Step 1 - Respondent IDs

110
Step 3. Select the long text elements to spell check from the Available Questions list and
click on the Finish command button to begin checking the selected respondent IDs and
questions (Figure 8.4-3).

Figure 8.4-3 Spell Check Agent Step 2 - Long Text Elements

Step 4. When the Spell Check Agent finds a misspelled word or a word not in the Spell
Check dictionary, the Spelling Check screen will pop up, displaying the misspelled word
in the Not in Dictionary field (Figure 8.4-4).

Figure 8.4-4 Spelling Check

111
The Spelling Check screen contains six command buttons:

Ignore - ignores the misspelled word once, and the spell check continues.

Ignore All - ignores the misspelled word for that entire spell check session, and the spell
check continues.

Add - adds the selected word to the Spell Check dictionary, and the spell check continues.
Use caution when adding words into the Spell Check dictionary, because words cannot be
removed from the dictionary.

Change - changes the misspelled word in the Not in Dictionary field. To correct the
spelling, either type the correct spelling directly into the Not in Dictionary field or select
the correct spelling from the Suggestions box by clicking on it. Note that once the
misspelled word is changed, the label changes from Not in Dictionary to Change To
(Figure 8.4-5).

Figure 8.4-5 Correcting Misspellings using the Spell Check Agent

After clicking on the Change command button, the spell check continues.

Change All - similar to the Change command button, but changes all of the words spelled
like the misspelled word for the entire spell check session, and the spell check continues.

Cancel - cancels the spell check session.

A pop-up box will appear to indicate when the spell check session is complete. Click on
the OK command button to return to the Data Entry Form screen.

112
Chapter 9. Reliability Agents

For many reasons, initial intra- or inter-coder reliability in coding qualitative data often is
quite poor (Miles and Huberman, 1994). To accurately code the themes present in the
responses to open-ended questions, it is essential to refine codebooks and train coders to
properly assign codes (Carey and Gelaude, 2008). Fortunately, it is possible to overcome
reliability problems in coding text obtained from semi-structured interviews (Carey et al.,
1996; Hruschka et al., 2004). EZ-Text provides Reliability Agents, which help users
assess and improve inter- and intra-coder reliability. Provision of reliability tools is an
unusual feature in qualitative software programs.

With the help of the Reliability Agents, EZ-Text users can select a random subset of
respondents from a database for up to four coders. The data subset is copied to a file for
each coder (up to four files for four coders), which the coders then independently code
using the same codebook. (For intra-coder reliability, the same coder would code two
files.) After the coders have completed their work, the coder Reliability Agent (Two-
coder or Multi-coder) compares the files and generates reports. The report generated by
using one comparison method (comparing a set of codes assigned by the coders to
responses) helps investigators identify which open-ended questions or respondents have
poor coding reliability. The other comparison method (comparing how the coders
assigned each code across responses) generates a kappa statistic for how the coders use
each code in the codebook. This helps identify which codes have poor reliability.

Researchers can use the reports to improve their codebook, focus additional attention on
hard-to-code portions of the database, or address further coder training needs. After
several iterations of reliability testing and codebook improvement, researchers can use
the reports to estimate the final level of inter- or intra-coder reliability. Using the Multi-
coder Reliability Agent, it is possible to compare coding by two or more coders.

113
9.1 Creating Comparison Files for Reliability Tests
The first step in using the Reliability Agents is to create a comparison file for each coder
to be compared (or two comparison files for a single coder for intra-coder reliability).

Step 1. Open the database to be checked for reliability. On the Data Entry Form top menu
bar, click on the Agents option, then highlight the Reliability Agents option on that menu,
and finally select Create Comparison Files (Figure 9.1-1).

Figure 9.1-1 Selecting Create Comparison Files from the Reliability Agents Menu

114
Step 2. A dialog screen is opened which prompts the user to specify the number of
respondents for EZ-Text to randomly select and copy to the comparison files (Figure 9.1-
2).

Figure 9.1-2 Selecting Number of Records for Comparison Files

This number is constrained by the number of records in the database. After entering the
desired number, click the Next command button.

115
Step 3. The next step prompts the user to specify the number of comparison files to
create (Figure 9.1-3).

Figure 9.1-3 Selecting Number of Comparison Files

EZ-Text can compare from two to four files to determine reliability. For inter-coder
reliability, create one comparison file for each coder under consideration. Up to four
coders can be compared at once. For intra-coder reliability, create two comparison files.
Once the desired number of files is selected from the drop-down list, click the Next
command button.

116
Step 4. The final step prompts the user to specify filenames for each of the comparison
files (Figure 9.1-4).

Figure 9.1-4 File Names for Comparison Files

When the names and paths for the files have been selected, click on the Finish command
button to create the files. A pop-up box will indicate that the comparison files have been
successfully created. Click on the OK command button and then the Cancel command
button to return to the Data Entry Form.

Note: In order to make meaningful reliability comparisons, coders should not alter the
data or the codebook copied into each of their comparison files. The only difference
between the files should be the way in which the coders have assigned the codes to the
identical set of responses using the same codebook. EZ-Text includes several internal
checks to help ensure that these conditions are met. This helps prevent meaningless or
invalid comparisons (for example, the coders use different codebook instructions, or they
use responses from different sets of respondents).

117
9.2 Running the Two-coder Reliability Comparison
If using the Two-coder Reliability Agent, after the two coders have independently coded
their files, the files can be compared to assess the degree of agreement. EZ-Text provides
two different statistical methods for assessing reliability.

Note: In projects where there is only one coder, we suggest conducting intra-coder
comparisons (i.e., can the single coder replicate his or her prior coding decisions using
the same codebook and the same data?). In this situation, the user may wish to create two
identical comparison files. The coder would code the first file, wait for an appropriate
period, and then re-code the second file without looking at the coding assignments in the
first file. Running the comparison reports with the two coded files may provide insights
into intra-coder consistency problems.

9.2.1 Comparison of Sets of Codes Assigned by Both Coders to the Same Set of
Responses
This method pinpoints which questions or respondents in the database have poor coding
reliability. To do this, EZ-Text compares the sets of codes each coder has assigned to
each open-ended question response in the comparison files (see Carey et al., 1996 for an
example of an application using this technique). For example, suppose that a coder
named Mary assigned CODE1, CODE2, and CODE3 to the response for open-ended
question 1 for respondent ID 1000. A coder named John also assigned CODE1, CODE 2,
and CODE3 to the same response for the same respondent. EZ-Text counts this situation
as an "agreement.” However, if John had only assigned CODE1 and CODE3, or assigned
an additional CODE4 to this response, EZ-Text would count these situations as
disagreements. For the purposes of computing reliability statistics, EZ-Text treats any
discrepancy in the way two coders assign sets of codes to a response--no matter how
small--as a disagreement. This procedure is intended to yield a statistically conservative
estimate of reliability.

The comparison process is repeated for all responses selected by the user (that is, the
number of selected respondents multiplied by the number of selected questions). EZ-Text
generates an output summarized by each respondent ID number and each open-ended
question number. In addition, the EZ-Text reliability output provides an estimate of
overall reliability by combining the results from all the selected questions and
respondents.

The output can be used to identify which open-ended questions and which respondents
have the lowest coding reliability levels. Subsequent coding improvement efforts can
focus on the data from these difficult-to-code questions and respondents. After
developing a final version of the codebook, both coders should recode their files.
Generating another reliability report will provide the final level of agreement achieved in
this study.

118
Step 1. From the EZ-Text Home Screen (before a database is opened) click on the Agents
option on the top menu bar, then highlight the Reliability Agents option on that menu, and
finally click on Two-coder Reliability Agent (Figure 9.2-1).

Figure 9.2-1 Activating the Two-coder Reliability Agent

119
Step 2. Select the name and path of the two files to compare by either typing the
information into the text fields or browsing to the path and file by clicking on the Browse
command button (Figure 9.2-2).

Figure 9.2-2 Select Comparison Files

Once the desired files have been selected, click on the Next command button.

120
Step 3. Click the selection button to the left of Compare set of codes assigned by the two
coders to responses, and click on the Next command button (Figure 9.2-3).

Figure 9.2-3 Select Comparison Method 1

121
Step 4. Select the long text elements to compare from the Available Questions list and
click on the Finish command button (Figure 9.2-4).

Figure 9.2-4 Select Long Text Elements

122
Step 5. The Reliability Report Preview displays the results of the comparison (Figure
9.2-5).

Figure 9.2-5 Reliability Report Preview

Step 6. Click on either the Print Page command button to print the current page or Print
All command button to print the entire report. In addition, the report can be saved in .rtf
(Rich Text File) format by selecting the Save to File option on the top menu bar of the
Report Preview screen.

123
9.2.2 Comparison of How Two Coders Use Each Code Across the Same Set of
Responses
The second reliability method compares how each coder has used each code across the
responses. EZ-Text constructs a 2 x 2 contingency table containing four cells showing
whether each coder assigned or did not assign a specific code to the set of responses. For
example, suppose a user wants to see how coders Mary and John used CODE1 across 200
response passages (for example, responses to 20 open-ended questions collected from 10
respondents). There are two agreement, or concordant cells: the upper left cell shows the
number of times both coders assigned the code, and the lower right cell contains the
number of times both coders did not assign CODE1 to the responses. The two discordant
cells indicate the number of times CODE1 was assigned by one coder but not the other.
The sum of all four cells is equal to the total number of response passages. (In this
example, the sum is 200).

A deceptively simple measure of agreement for how CODE1 is assigned would be the
proportion of times the two coders agreed (in this example, the sum of the two
concordant or agreement cells divided by 200). However, simple proportions do not take
into account that some of this agreement may occur by chance (Gorden, 1992: 184-185;
Fleiss et al., 2003: 598-626). Therefore, the proportion of agreement is a biased
overestimate of the agreement of how two coders have assigned CODE1.

To correct for this problem, the kappa statistic is a more appropriate measure. (See Carey
et al., 1996 for an example of an application using this technique.) In general, kappa can
be thought of as the amount of agreement between two coders after statistically adjusting
for agreement owing to chance. A kappa value of 1.00 indicates perfect agreement. A
value of 0.00 indicates no agreement beyond chance, and kappa takes on negative values
as low as -1.00 when there is less agreement than would be expected by chance (Fleiss et
al., 2003).

Ideally, researchers will want the kappa scores calculated for all the codes in their final
codebook to be as close as possible to 1.00. In the study conducted by Carey et al. (1996),
final inter-coder reliability kappa score values for 135 of 152 codes present in the final
codebook fell between 0.90 and 1.00. Of these 135 codes, 126 had a kappa value exactly
equal to 1.00, indicating complete agreement between the two coders. See Hruschka et al.
(2004) for additional examples.

Users may obtain kappa statistics for any combination of codes in the codebook in the
two comparison files. After instructing EZ-Text which codes, respondents, and questions
to include in the calculations, EZ-Text generates an on-screen report that shows the
following results for each code: the simple percent agreement between the two coders,
the amount of agreement estimated to be caused by chance, the kappa statistic, and the
standard error for kappa. If desired, the 2 x 2 table can be viewed on screen. A report
summarizing these results can also be viewed and printed or saved.

124
The results can be used to identify which codes have the poorest reliability. In many
cases, simple changes in the codebook can ameliorate the problem (clarifying ambiguous
code definitions, removing duplicate or overlapping codes, clarifying coder training
misunderstandings, etc.).

Step 1. From the EZ-Text Home Screen (before a database is opened) click on the Agents
option on the top menu bar, then highlight the Reliability Agents option on that menu, and
finally click on Two-coder Reliability Agent (Figure 9.2-1).

Step 2. Select the name and path of the two files to compare by either typing the
information into the text fields or browsing to the path and file by clicking on the Browse
command button (Figure 9.2-2). Once the desired files have been selected, click on the
Next command button.

Step 3. Click the selection button to the left of Compare how the two coders assigned
each code across responses, and click on the Next command button (Figure 9.2-6).

Figure 9.2-6 Select Comparison Method 2

125
Step 4. Select the long text elements to compare from the Available Questions list and
click on the Next command button (Figure 9.2-7).

Figure 9.2-7 Select Long Text Elements

126
Step 5. Select the codes to compare from the Available Codes list and click on the Finish
command button (Figure 9.2-8).

Figure 9.2-8 Select Codes to Compare

127
Step 6. Once the kappa values have been calculated, the results are displayed in a
summary table (Figure 9.2-9).

Figure 9.2-9 Two Coder Kappa Summary Table

128
Step 7. Click on Print Report to preview the Two Coder Kappa Summary Report (Figure
9.2-10).

Figure 9.2-10 Two Coder Kappa Summary Report Preview

Step 8. Click on either the Print Page command button to print the current page or Print
All command button to print the entire report. In addition, the report can be saved in .rtf
(Rich Text File) format by selecting the Save to File option on the top menu bar of the
Report Preview screen. Click on the Return command button to return to the Kappa
Summary Table (Figure 9.2-9).

129
Step 9. On the Two Coder Kappa Summary screen (Figure 9.2-9), click on the row
containing a particular code, and then click on the Details command button to view the 2
x 2 contingency table discussed at the beginning of this section (Figure 9.2-11).

Figure 9.2-11 Viewing the 2 x 2 Contingency Table

Note: The distribution of cell values in the 2 x 2 contingency table sometimes falls into
one cell. When this occurs, kappa cannot be calculated using the formula. Instead, since
this is a boundary condition, the occurrence must be handled by definition of the
particular condition. Specifically:

If all the cell values fall into either of the two concordant cells (that is, all in the "Yes-
Yes" or the "No-No" cells in Figure 9.2-11), the two coders are in complete agreement.
Kappa is 1.0 by definition. (Fleiss et al., 2003)
Or
If all the cell values fall into either of the two discordant cells (that is, all in the "Yes-No"
or the "No-Yes" cells in Figure 9.2-11), the two coders are in complete disagreement.
Kappa is -1.0 by definition. (Fleiss et al., 2003)

When either of the two foregoing cell value patterns occurs, other results cannot be
calculated in addition to kappa. These are the "% Due to Chance" and the "Standard
Error" of kappa. In this circumstance, "Not Applic." is entered in the appropriate column
of the Kappa Summary Report (Figure 9.2-9).

130
If all of the cell values fall into the two cells in a single row of the 2 x 2 table or into the
two cells in a single column of the 2 x 2 table, the value of kappa is zero (0) and the
standard error is zero (0).

9.3 Running the Multi-coder Reliability Comparison


As was mentioned earlier, it may be useful to assess coding reliability when there are
more than two coders. This could be the case when one is attempting to train three or
more coders at the same time. When this is the circumstance, it is possible to use a form
of the coefficient kappa developed by Fleiss (1971). Fleiss' kappa is a generalization of
Cohen's kappa statistic. Whereas Cohen's kappa works for only two raters, Fleiss' kappa
works for any number of coders giving categorical ratings to a fixed number of items. It
can be interpreted as expressing the extent to which the observed amount of agreement
among raters exceeds what would be expected if all raters made their ratings completely
randomly.

The calculation of the Fleiss statistic has some marginal similarity to that of Cohen's
kappa. It differs, due to the increased number of coders, in how the rate of agreement and
the expected random agreement are calculated. This is based on the arrangement of the
data which must accommodate the number of coders. The section that follows addresses
this arrangement and how EZ-Text is used to calculate this form of kappa.

Like two-coder reliability, there are two methods EZ-Text employs for multi-coder
reliability as well. The first method compares how each coder assigned each code across
responses (coder accuracy). The second method compares sets of codes assigned by all
coders to the same set of responses (coding reliability).

131
9.3.1 Comparison of How Multiple Coders Assign Each Code Across the Same Set
of Responses
After two or more coders have independently coded their files, the files can be compared
to assess degree of agreement. The process for comparing multiple coders is similar to
that for two coders.

Step 1. From the EZ-Text Home Screen (before a database is opened) click on the Agents
option on the top menu bar, then highlight the Reliability Agents option on that menu, and
finally click on Multi-coder Reliability Agent (Figure 9.3-1).

Figure 9.3-1 Activating the Multi-coder Reliability Agent

132
Step 2. Select the name and path of the files to compare by either typing the information
into the text fields or browsing to the path and file by clicking on the Browse command
button (Figure 9.3-2).

Figure 9.3-2 Select Comparison Files

Once the desired files have been selected, click on the Next command button.

133
Step 3. Click the selection button to the left of Compare how the coders assigned each
code across responses, and click on the Next command button (Figure 9.3-3).

Figure 9.3-3 Select Comparison Method 1

134
Step 4. Select the long text elements to compare from the Available Questions list and
click on the Next command button (Figure 9.3-4).

Figure 9.3-4 Select Long Text Elements

135
Step 5. Select the codes to compare from the Available Codes list and click on the Finish
command button (Figure 9.3-5).

Figure 9.3-5 Select Codes to Compare

136
Step 6. Once the kappa values have been calculated, the results are displayed in a
summary table (Figure 9.3-6).

Figure 9.3-6 Multiple Coder Kappa Summary Table

137
Step 7. Click on Print Report to preview the Multiple Coder Kappa Summary Report
(Figure 9.3-7).

Figure 9.3-7 Multiple Coder Kappa Summary Report Preview

Step 8. Click on either the Print Page command button to print the current page or Print
All command button to print the entire report. In addition, the report can be saved in .rtf
(Rich Text File) format by selecting the Save to File option on the top menu bar of the
Report Preview screen.

138
9.3.2 Comparison of Sets of Codes Assigned by Multiple Coders to the Same Set of
Responses
Step 1. From the EZ-Text Home Screen (before a database is opened) click on the Agents
option on the top menu bar, then highlight the Reliability Agents option on that menu, and
finally click on Multi-coder Reliability Agent (Figure 9.3-1).

Step 2. Select the name and path of the files to compare by either typing the information
into the text fields or browsing to the path and file by clicking on the Browse command
button (Figure 9.3-2). Once the desired files have been selected, click on the Next
command button.

Step 3. Click the selection button to the left of Generate a Coding Comparison Report,
and click on the Next command button (Figure 9.3-8).

Figure 9.3-8 Select Comparison Method 2

Step 4. Select the long text elements to compare from the Available Questions list and
click on the Next command button (Figure 9.3-4).

139
Step 5. Select the codes to compare from the Available Codes list and click on the Next
command button (Figure 9.3-9).

Figure 9.3-9 Select Codes to Compare

140
Step 6. Select the respondent IDs to compare from the Available Respondent IDs list and
click on the Finish command button (Figure 9.3-10).

Figure 9.3-10 Select Respondent IDs to Compare

141
Step 7. The Coding Comparison Report Preview displays the results of the comparison
(Figure 9.3-11).

Figure 9.3-11 Coding Comparison Report Preview

Step 8. Click on either the Print Page command button to print the current page or Print
All command button to print the entire report. In addition, the report can be saved in .rtf
(Rich Text File) format by selecting the Save to File option on the top menu bar of the
Report Preview screen.

142

You might also like