QCA 3.

Qualitative Comparative Analysis

(C) Copyright 1992-98, Kriss A. Drass

 THE BASICS

QCA uses a fairly consistent set of commands for gathering

information from you. Information requests are presented on the
screen as boxed windows. A window may contain any of the following
elements: text fields, where you type text information (file name,
variables, etc.), list fields, where you make choices from a list
by highlighting an option and pressing the space bar or <─┘, or by
clicking on an option with the left mouse button, and button fields,
where you select an immediate action by highlighting the button and
pressing <─┘, by pressing Alt-Highlighted Letter, by pressing a hot
key, or by clicking on the button with the left mouse button.

You can move among window fields and buttons using the TAB (next
field/button) and Shift-TAB (previous field/button) keys. The
cursor control keys   <- -> can also be used, although not all of
the cursor keys work with all fields.

The commands for editing text in a text field are:

<- Move the Cursor One Character to the Left

-> Move the Cursor One Character to the Right
Home Move the Cursor to the Beginning of the Field
End Move the Cursor to the End of the Field
Back Space Delete the Character to the Left of the Cursor
Del Delete the Character at the Current Cursor
Position
Alt-D Delete All Characters in the Field
Ins Switch between Insert (Block Cursor) and
Overwrite (Line Cursor) Modes

Some text fields scroll horizontally to accomodate lines longer than

the size of the field. QCA marks these scrolling text fields with
a -> at the far right of the field.

The standard button hot keys are:

Esc - Cancel and return to the next higher level of the program,
which is usually the Log;

F10 - "Ok", which means all requested information has been

entered, so you are ready to continue to the next step.

F1 - Help, which will display help information about the

-2-
 current window.

Alt-X is the standard Exit key for quitting editors, and the program
itself.

In addition to these basics, each window may have other commands

for performing actions specific to it. The keys used to execute
these commands will be listed across the top or bottom line of the
screen. If a key is highlighted, i.e., it appears in red on color
systems or normal video (as opposed to reverse video) on monochrome
systems, then you can also execute the command by clicking on the
highlighted portion of the screen with the left mouse button. Some
commands access menus from which other commands can be selected.

Menus

QCA's Log and editors (all discussed in detail below) allow you to
execute commands either by pressing a "hot key" or by selecting
options from a menu. Most of the hot keys involve holding down the
Alt key and then pressing a letter. Menu options can be selected
by highlighting your choice using the   cursor keys and then
pressing <─┘, or by pressing a highlighted letter key (do not hold
down the Alt key when making selections within a menu). The <- and
-> cursor keys allow you to move across menus. You can access QCA's
menus by pressing the / key, or by pressing Alt-F for the File menu,
Alt-E for the Edit Menu, or (if available) Alt-M for the Methods
menu.

Objects

A lot of software these days uses some type of object-oriented

design philosophy. Generally, this philosophy is explained to the
user through the use of the "desk top" metaphor. The computer
screen, we are told, is like a desk top. Objects appear on the
screen as icons and/or text labels. When we want to do something,
we simply select an object (with the cursor or mouse) and then
perform an action with it, much like we do when we move our hand
toward our desk calendar and then turn a page.

This object-oriented approach fits the data analysis process very

nicely. If you think about data analysis from an abstract
perspective, it is nothing more than a process of manipulating
objects (data sets) using actions (methods) consistent with the
characteristics of the objects (e.g., using regression techniques

-3-
to analyze continuous data, applying classification rules to the
coding of qualitative data). QCA employs an object-oriented
approach to the analysis of qualitative data using techniques from
Boolean algebra. For QCA, objects are different types of Boolean
functions. Once created, these objects appear on your computer
screen as headers and, possibly, equations. To manipulate one of
these objects, you point to it with the cursor and then select the
action that you want to perform. The actions that you direct toward
these objects are analytic techniques (some are based upon Boolean
algebra, others are extensions of traditional social science
research methods) designed to help you better understand the nature
of your data.

This object-oriented approach is more than just an elaborate ruse

to justify the purchase of a mouse. An important part of this
philosophy is the idea that objects can be related to one another
in a hierarchical manner. Objects at a lower level of the hierarchy
"inherit" features from objects at a higher level of abstraction.
These objects are "descendants" of the higher level "ancestor"
objects. Thus, objects constitute a kind of family. Like members
of any family, descendants inherit features from their ancestors.
QCA relies upon this notion of inheritance to structure
relationships among Boolean functions. Therefore, you need to
understand a little bit about QCA's object hierarchy in order to
get the most out of the program.

Boolean Functions: The Object Hierarchy

At the top level of QCA's object hierarchy is the truth table. Truth
tables list the various combinations of independent variables that
appear in a data set along with their corresponding values on the
dependent variable. Configurations appear only once in a truth
table, regardless of their frequency of occurrence in the data.
Truth table objects combine a truth table along with all of the
information necessary to create it from a data set; i.e., the data
file's name, directory, type, independent and dependent variables,
rules for processing missing values, rules for selecting cases, etc.
Truth table objects can appear explicitly in the Log (as a header),
or they may appear implicitly as the ancestor of another object.
Regardless, since truth tables "know" how to re-create themselves,
you can analyze them, or their descendants, without having to re-
enter a lot of file and data definition information. From a user's
perspective, this is where you realize the power of objects.
Therefore, if you would like to edit a truth table, or edit the raw

-4-
data associated with a truth table, simply point to an object (either
a truth table or one of its descendants) with the cursor (i.e.,
position the cursor anywhere on the function), select your action,
and go! QCA keeps track of all of the other details for you.

The major descendant of a truth table is the minimized truth table.

As its name implies, a minimized truth table object combines a
minimized truth table function, produced by applying a minimization
algorithm to a truth table, along with all of the file and data
definition information inherited from the truth table object. Once
created, QCA provides procedures for computing the complement and
assumptions of a minimized function, creating hypotheses for a
minimized function, computing and viewing frequencies for a
minimized function, and computing intersections between minimized
functions and/or their hypotheses. To perform any of these actions,
you just point and select -- QCA gets the necessary details from
the object.

Once a minimized truth table has been created, a number of other

objects can be linked to it through inheritance. These descendant
objects inherit the truth table in its complete and minimized form
through the minimized truth table object. This inherited
information is used by the editors to compute frequencies, report
CaseIds, and view/edit the original truth table. These objects also
share in common the fact that they are minimized functions of some
type; i.e., they are not fully specified truth tables. These
descendants of the minimized truth table object are: (1) the
complement of the minimized truth table function, (2) the
assumptions (i.e., minimized don't cares) used by the minimized
truth table function, (3) hypotheses that can be tested using the
minimized truth table, or one of its descendants, (4) and
complements of the hypotheses. Because of inheritance, you can
always get back to the original truth table through one of these
objects. Simply point to the function with the cursor, and then
select the action that requires the truth table.

The final object in QCA's hierarchy is the intersection object.

This object is a kind of hybrid that inherits all of the information
about a truth table's independent and dependent variables through
one of the other objects but no information about the truth table's
data. Therefore, you cannot access a truth table through an
intersection object.

-5-
So, what do you really need to remember about QCA's object hierarchy?
The bottom line is that QCA's object hierarchy means that you can
run most procedures by positioning the cursor anywhere on an
appropriate function and then selecting an action to perform. QCA
handles most of the details for you. The environment where you
interact in this way with your objects is the QCA Log. It is the
heart of QCA 3.0.

-6-
 THE QCA LOG

QCA's Log is a collection of objects that constitute the "results"

of the current session. Functions appear in the Log along with a
header which identifies the function's data file, model, and purpose
(i.e., what it is). In addition, for all except truth table objects,
QCA lists the full Boolean equation for each function.

One important role of the Log is to act as a simple editor so that

you can see your results. The cursor control commands for moving
around the Log are:

Cursor Control Commands:

From the Keyboard

 Move Cursor Up One Line
 Move Cursor Down One Line
PgUp Move Cursor Up One Screen (about 18 Lines)
PgDn Move Cursor Down One Screen
Ctrl-Home Move Cursor to First Line of Screen
Ctrl-End Move Cursor to Last Line of Screen
Ctrl-PgUp Move Cursor to First Line of Log
Ctrl-PgDn Move Cursor to Last Line of Log
- (Minus Sign) Move Cursor to Header for Previous Function
+ (Plus Sign) Move Cursor to Header for Next Function

With a Mouse (Left Button)

Single Click Move Cursor to Line
Click on  Move Cursor Up One Line
Click on  Move Cursor Down One Line
Click above _ Move Cursor Up One Screen
Click below _ Move Cursor Down One Screen
Click below  Move Cursor to First Line of Log
Click above  Move Cursor to Last Line of Log

With a Mouse (Right Button)

Click on  Move Cursor to Header for Previous Function
Click on  Move Cursor to Header of Next Function

In addition, the Log is QCA's control center. Objects are created

and manipulated by executing commands, using a menu or hot key, from
within the Log. The commands available in the Log are presented
below by menu.

-7-
File Menu .... .... .... .... .... .... .... .....Alt-F

The File Menu presents commands for creating and loading data files,
reporting summary information about a file, saving and printing
results, browsing ASCII text files, and quitting the program. Many
of these commands are also used by QCA's various editors (discussed
below). The basic file commands are:

Open a Data File ... .... .... .... .... .....Alt-O

The Open a Data File command opens the "Select a File" window
so that you can select a data file for analysis. This command
also initiates a sequence of windows (Select a File, Build a
Truth Table, Minimize a Truth Table) that continue until a
truth table is minimized or you choose to escape from a window.

Set File Types .... .... .... .... .... .....Alt-Y

QCA can read data from six different types of files: QCA data
matrix files, Gauss system files, Lotus WKS or WK1 files, SPSS
Export files, Stata system files, and Systat system files.
Most of these files use default file extensions for easy
identification. By default, QCA uses the following file
extensions for classifying data files:

.QDM .QVN = QCA file

.DHT = Gauss system file
.WK1 .WKS = Lotus spreadsheet file
.XPT = SPSS Export file
.DTA = Stata system file
.SYS = Systat system file

The Set File Types command allows you to change these default
assignments so that QCA can identify your data files properly.
For example, if you save all of your Stata system files with
a .STA extension, then you would use this command to add .STA
to the line for Stata files.

New Data File . .... .... .... .... .... .....Alt-N

The New Data File allows you to build a QCA truth table from
scratch. QCA will ask you to name the dependent and independent
variables, and to indicate whether memory should be allocated
for CaseIDs. QCA will then open the Data Matrix Editor
(described below) so that you can enter your data.

-8-
Save Output ... .... .... .... .... .... .....Alt-O
The Save Output command allows you to save either the current
screen or the contents of the entire Log to an output file.
QCA asks you to enter a filename (default = QCA.OUT), select
the amount of text to be saved (All or Screen), and select how
the file should be written (replace the output file by
Overwriting it or add to the output file by Appending text to
the end of it).

Print Output .. .... .... .... .... .... .....Alt-P

The Print Output command allows you to send output directly to
a printer (LPT1) if one is connected. It is identical to the
Save command, except that it (obviously) does not need you to
indicate whether to Overwrite or Append results. Unless you
roll back or reload the paper, output will be appended to any
previous printing.

Browse File ... .... .... .... .... .... .....Alt-B

The Browse File command allows you to load and view any type
of ASCII text file (if there is sufficient memory). This
command can be handy if you would like to check the contents
of the current output file or if you would like to examine a
QCA data matrix. Use the following commands to browse through
your file:

From the Keyboard

 Move Screen of Text Up One Line
 Move Screen of Text Down One Line
-> Move Screen of Text Right One Column
<- Move Screen of Text Left One Column
PgUp Move Text Up One Screen
PgDn Move Text Down One Screen
Ctrl-PgUp Show Text at Top of File
Ctrl-PgDn Show Text at End of File
Home Show Text at Far Right of File
End Show Text at Far Left of File

With a Mouse (Left Button)

Click on  Move Screen of Text Up One Line
Click on  Move Screen of Text Down One Line
Click on -> Move Screen of Text Right One Column
Click on <- Move Screen of Text Left One Column
Click above _ Move Text Up One Screen
Click below _ Move Text Down One Screen

-9-
 Click right of _ Move Text Right One Screen
Click left of _ Move Text Left One Screen
Click below  Show Text at Top of File
Click above  Show Text at End of File
Click left of -> Show Text at Far Right of File
Click right of <- Show Text at Far Left of File

Which File? ... .... .... .... .... .... .....Alt-W

All of the objects appearing in the Log are linked to a data
file, either directly (truth tables) or through inheritance
(all other functions). The Which File? command tells you
exactly "which file" is linked to a function in the Log. To
get this information, position the cursor anywhere on an object
(i.e., next to any header line or any function term) and then
enter the command.

Exit QCA . .... .... .... .... .... .... .....Alt-X or Click
On
"Alt-X to Exit"
Basically, "that's all folks!" Before quitting, QCA checks to
be sure that you really want to exit the program and gives you
a chance to change your mind.

Edit Menu .... .... .... .... .... .... .... .....Alt-E

The Edit Menu presents commands for editing the text in the Log and
for starting QCA's various editors. To edit a function, point to
it with the cursor and then use the menu or a hot key to select the
desired action. The basic editing commands are:

View Function . .... .... .... .... .... .....Alt-U

The View Function command starts QCA's function editor and
loads the current function. The function can be a minimized
truth table, complement, set of assumptions, hypothesis, or an
intersection. The function editor shows each function term
along with frequencies for the number of truth table
configurations and raw data cases that it implies. In
addition, if the data file contains CaseIds, the function
editor will also report the CaseIds for the cases implied by
each term in the function. The specific commands for this
editor are discussed in detail below.

Edit Truth Table ... .... .... .... .... .....Alt-T

The Edit Truth Table command starts QCA's truth table editor

-10-
and loads the truth table associated with the current function.
The truth table editor shows each row of a truth table along
with the number of 0, 1, and - (don't care) outputs associated
with each row. The truth table can be changed by recoding the
output values associated with any row. This "new" truth table
can then be written to a QCA data matrix file. The specific
commands for this editor are discussed in detail below.

Edit Raw Data . .... .... .... .... .... .....Alt-R

The Edit Raw Data command starts QCA's data matrix editor and
loads data from the file associated with the current function.
In the data matrix editor, each line is a "case" of raw data
rather than a row of a truth table. Therefore, a configuration
appears as frequently as it actually occurs in the data. The
data matrix editor works like any type of simple text editor.
Any value of any variable can be changed, lines can be deleted
or inserted, and QCA will even calculate and insert Logical
configurations (configurations that exist logically, but are
not found in the data). Any changes that you make can be
preserved by writing the data to a QCA data matrix file. The
specific commands for this editor are discussed in detail
below.

Edit Hypothesis .... .... .... .... .... Alt-H

The Edit Hypothesis command starts QCA's hypothesis editor.
The hypothesis editor is another type of "text" editor that is
used exclusively for creating or editing hypotheses. A
hypothesis is always defined in relationship to a minimized
truth table function. The minimized truth table function
provides information about the specific variables needed to
create a hypothesis. QCA allows multiple hypotheses for a
minimized function. To keep things clear, QCA automatically
supplies each hypothesis with a number. To create a NEW
hypothesis, position the cursor anywhere on a minimized
function, and then invoke the editor. To edit an OLD
hypothesis, position the cursor anywhere on the hypothesis and
then invoke the editor. If you change a hypothesis that has
a computed complement as part of the Log, QCA will ask if you
would like to recompute or delete this complement since it may
no longer be valid. The specific commands for this editor are
discussed in detail below.

Label Function .... .... .... .... .... .....Alt-L

The Label Function command allows you to annotate your output

-11-
 by attaching a "note" to any type of function. QCA provides
a simple word-wrapping text editor for this purpose. Each note
can contain up to 64 lines of text. To create or edit a note,
simply position the cursor anywhere on a function and then use
the menu or hot key to start the editor.

Delete Function .... .... .... .... .... Alt-D

The Delete Function command allows you to delete the output
for a function from the Log. Simply position the cursor
anywhere on the function and then use the menu or hot key to
enter the command. QCA gives you a chance to change your mind
before it actually deletes the function.

Mark Intersection .. .... .... .... .... .....I

The Mark Intersection command allows you to mark a function
for use in the computation of an intersection. The
intersection symbol (  ) appears next to the header for the
function. If the function is already marked, this command
clears it. Please note that the command is NOT Alt-I, but
simply I.

Methods Menu .. .... .... .... .... .... .... .....Alt-M

The Methods Menu presents commands for creating and analyzing

minimized truth table functions. To apply a method to a function,
simply point to it with the cursor and then use the menu or a hot
key to select the desired action. The basic methods commands are:

Compute Assumptions .... .... .... .... .....Alt-A

Including don't care configurations in the minimization of a
truth table generally leads to a more minimal solution.
However, this increased simplicity comes with a theoretical
cost; you must be willing to assume that the don't care
configurations produce the same output(s) as the other
configurations in the truth table. The Compute Assumptions
command identifies the specific don't care configurations used
in the minimization of a truth table (and thus, assumed to be
equivalent in output to the other configurations in the truth
table) and summarizes these "assumptions" in the form of a
minimized function. This method only applies to minimized
truth table functions that actually use don't cares. To
compute assumptions, position the cursor anywhere on the
minimized function, and then use the menu or hot key to start
the method.

-12-
Compute Complement . .... .... .... .... .....Alt-C
The Compute Complement command computes the complement of a
minimized truth table function or hypothesis. The complement
is the minimized function of all configurations that are NOT
implied by the original function or hypothesis. To compute a
complement, simply position the cursor anywhere on the
minimized truth table function or hypothesis, and then use the
menu or hot key to start the method.

Compute Intersection .... .... .... .... Alt-I

The Compute Intersection command computes the intersection of
all functions previously marked with the Mark Intersection
Command (I). An intersection is defined as the set of
configurations implied by ALL functions in the intersection.
This set is then minimized using the QM algorithm just like
any other truth table (thus, you will see a Prime Implicant
Table and you may be asked to complete the minimization by
hand). Any function except another intersection can be used,
as long as all functions are compatible -- i.e., their truth
tables have the same variables in the same order. Before
computing the intersection, QCA checks to be sure that all
marked functions are compatible and returns an error message
if they are not. To compute an intersection, simply invoke
the command from the menu or with the hot key after you have
marked at least two compatible functions.

Minimize Truth Table .... .... .... .... Alt-Z

The Minimize Truth Table command allows you to minimize the
truth table associated with any object in the Log except an
intersection. It is basically a shortcut so that you can redo
a minimization (selecting different outputs if desired) without
having to select the data file and the variables. To minimize
a truth table, simply position the cursor anywhere on the
function whose truth table you want to minimize and then use
the menu or hot key to invoke this method.

Select Variables ... .... .... .... .... .....Alt-V

The Select Variables command allows you to select a different
set of variables and build a new truth table while inheriting
the data file from the current function. Like the previous
method, this command is a shortcut that allows you to begin
the process of minimizing a truth table without having to
select a data file to analyze. To select variables, simply

-13-
 position the cursor anywhere on the function whose data file
you want to use and then use the menu or hot key to invoke this
method.

All of QCA's main procedures are launched from within the Log. Some
of these procedures, like computing a complement, never require you
to leave the Log. However, most of these procedures are fairly
complicated, and present a number of other actions for you to use
to create or manipulate an object. In the following sections of
this Guide, some of the more involved procedures are discussed in
greater detail.

-14-
 OPENING A DATA FILE

The Open a Data File command opens the "Select a File" window. The
purpose of this window is to get the name of the data file that you
want to analyze using QCA. Data from this file will be used to
build a truth table that you can then minimize using QCA's Boolean
minimization algorithms.

QCA 3.0 can read six different types of data files. Since some of
these are system files for stat packages, it makes it easier to
combine a QCA analysis with a traditional statistical analysis. It
also means that the stat package can be used for data manipulation
(data entry, computing and recoding variables, selecting samples of
cases, etc.), thus freeing QCA to concentrate on what it does best
-- minimizing truth tables. QCA classifies data files by their file
extensions. Currently, the file types and defaults extensions used
by QCA are:

.QDM .QVN = QCA File

.DHT = Gauss System File
.WKS .WK1 = Lotus 123 System File
.DTA = Stata System File
.XPT = SpssX Export (NOT SYSTEM) File
.SYS = Systat System File

If you use a stat package that is not included in this list, check
to see if it will save files using the Lotus spreadsheet format.
However, please note that QCA makes the following assumptions about
the structure of Lotus or compatible spreadsheet files. First, and
most important, QCA assumes that the cells in the first row of the
spreadsheet contain variable names (Lotus Labels) for their
respective columns. Second, QCA assumes that the data begin in the
second row of the spreadsheet. Finally, QCA assumes that each
column contains cells of the same type of data (i.e., all entries
in a column are either Integers, Floating Point Reals, or Text
Labels). Cells can contain formula as long as the results match
other cell values in the column. Data types can vary across columns
(e.g., a file can contain columns of both Integers and Reals, or
Reals and Labels, etc.), but they must be consistent within columns.
If you have problems trying to load a Lotus compatible file, check
to be sure that the spreadsheet conforms to these assumptions.

With release 3.0, a default format for QCA data files has been
established. This format uses two files. One is a variable names

-15-
file and the other is a data matrix file. The QCA variable names
file is a standard ASCII file that has a .QVN extension. A variable
name corresponding to each column of the data file is listed on a
separate line. In addition, if a CaseId is present, it is ALWAYS
the first variable name and it ALWAYS uses the name CASEID. A file
with CaseIds would look like:

CASEID
VARA
VARB
VARC
VARD

The variable names must be unique, and 8 or fewer letters, numbers,

or symbols. It does not matter if they are upper or lower case
because QCA converts all variable names to upper case for purposes
of matching. As a result, the uniqueness of a name cannot be based
on case (e.g., VarA and VARA would not be treated as separate
variables). Also, it does not matter whether a name begins with a
letter, number, or symbol (although the convention is to begin all
names with a letter).

The QCA data matrix file is also a standard ASCII file with the
default extension of .QDM. Each row of the file corresponds to a
case of data. Each variable of data gets only one column in the
file (no 99, or 191), and QCA recognizes only three values: zero
(0), one (1) and don't care (- or 2). Any value other than these
is treated as missing (although a period [.] is the default missing
value indicator). QCA will remove any blanks or commas between
values, although it is best not to use these since they take up
space and limit the number of variables that can be stored in the
file. The one exception to the above rules is for files with
CaseIds.

CaseIDs are used to identify cases within a data matrix. They can
be 1 to 12 characters in length. Blank spaces are legal "characters"
within a CaseID. CaseIds are optional. However, if they are used,
they MUST appear first on a line and must be separated from the rest
of the data by a single comma. An example of a file with a CaseID
is:

Burma,10110011010101220
France,00110011001022101
USA,10101001001001000

-16-
 Germany,10100011002001011

Without CaseIds, this file would look like:

10110011010101220
00110011001022101
10101001001001000
10100011002001011

QCA can handle a maximum of 1 CaseId and 255 variables per case.
However, the number 255 assumes that NO spaces and commas appear
within the data portion of a line. In essence, QCA can read a
CaseId (of 1-12 characters), the comma separating the caseid from
the data, and then 255 columns beyond the comma.

Use the following fields and commands to select the file that you
want to analyze.

QCA will recognize CaseIds from data files other than QCA files.
All you need to do is give one of the variables in your file the
name CASEID. It can be any type of variable in any position; i.e.,
it does not need to be the first variable in the file nor does it
need to be a string varible. Whenever QCA opens a "foreign" data
file, it checks for a variable named CASEID. If found, QCA remembers
its location and uses it exclusively for reporting CaseIds.

Window Fields

Name A text field that is used to enter the name of the

data file that you want to open, to specify a new
drive or directory to search, or to change the file
mask (i.e., the DOS search criteria that QCA uses
for including files in the Files List).

Files A list field that presents the files in the current

directory that match the search criteria specified
in the Name Field. Files can be opened directly
from this list by highlighting them with the cursor
and then pressing <─┘ or by DOUBLE clicking on the
file name with the left mouse button.

Directories A list field that presents all system drives except

the current one, all subdirectories of the current

-17-
 directory, and the parent of the current directory.
You can change the current drive or directory by
highlighting an element of this list with the cursor
and then pressing <─┘ or by DOUBLE clicking on the
item with the left mouse button. QCA will
immediately update both the Files List and
Directories List. QCA displays the current drive
and directory at the bottom of the window.

Ok A button field that is used to quit this window and

begin the process of opening a data file. If
selected, QCA will try to open the data file that is
currently highlighted in the Files List. This button
can be selected by highlighting it with the cursor
and pressing <─┘, by holding down the Alt key and
then pressing O, by pressing the F10 function key,
or by clicking it once with the left mouse button.

Cancel A button field that is used to quit this window and

return to the QCA Log WITHOUT opening a data file.
This button can be selected by highlighting it with
the cursor and pressing <─┘, by holding down the Alt
key and then pressing C, by pressing the Esc key, or
by clicking it once with the left mouse button.

Help A button field that is used to access QCA's online

Help system. This button can be selected by
highlighting it with the cursor and pressing <─┘, by
holding down the Alt key and then pressing H, by
pressing the F1 function key, or by clicking it once
with the left mouse button.

Window Commands

Main Window

From the Keyboard

Tab Move the Cursor to the Next Field.
Shift-Tab Move the Cursor to the Previous Field.
F10 OR Alt-O Select the Ok Button. QCA will Open the Data File
Highlighted in the Files List.
Esc OR Alt-C Select the Cancel Button. QCA will Return to the
Log without Opening a Data File.
F1 OR Alt-H Access QCA's online Help system.

-18-
With a Mouse (Left Button)
Click on Field or Label to Move Cursor to Field.
Click on Ok Button to Open the Data File Highlighted in the
Files List.
Click on Cancel Button to Return to the Log without Opening
a Data File.
Click on Help Button to Access QCA's online Help system.

Name Field

From the Keyboard

<- Move the Cursor One Character to the Left
-> Move the Cursor One Character to the Right
Home Move the Cursor to the Beginning of the Field
End Move the Cursor to the End of the Field
Back Space Delete the Character to the Left of the Cursor
Del Delete the Character at the Current Cursor Position
Alt-D Delete All Characters in the Field
Ins Switch between Insert (Block Cursor) and Overwrite
(Line Cursor) Modes

File and Directory Lists

From the Keyboard

 Move the Cursor Up One Line
 Move the Cursor Down One Line
PgUp Move the Cursor Up One Screen
PgDn Move the Cursor Down One Screen
Home Move the Cursor to the Top of the List
End Move the Cursor to the Bottom of the List
<─┘ Select Item

With a Mouse (Left Button)

Single Click Highlight Item
Double Click Select Item
Click on  Move Cursor Up One Line
Click on  Move Cursor Down One Line
Click above _ Show Previous Screen of Items
Click below _ Show Next Screen of Items
Click below  Show First Screen of Items (Top of List)
Click above  Show Last Screen of Items (End of List)

-19-
Selecting Files and Directories

QCA provides three ways to open a data file. First, you can move
the cursor to the Name field and type a fully specified file name.
For example, if you want to open the file GAMSON.QDM from the current
directory, simply type:

GAMSON.QDM <─┘

QCA displays the current directory at the bottom of the Select A

File Window. If you want to open a file from a different directory,
type the appropriate file path along with the file name. For
example, to open the GAMSON.QDM file from the directory MYDATA on
drive C, simply type:

C:\MYDATA\GAMSOM.QDM <─┘

QCA also allows you to open a file by selecting a file name from
the list displayed in the Files List. Use Tab, Shift-Tab, or the
mouse to move the cursor to the Files List. Highlight the name of
the file that you want to open and press <─┘ or double click the
left mouse button on the name. Finally, selecting the Ok button at
any time will cause QCA to open the file that is currently
highlighted in the Files List.

QCA also allows you to change the drive and/or directory that will
be used for loading data files. First, you can move the cursor to
the Name Field and type the new drive and/or directory. IF YOU TYPE
A DIRECTORY NAME, BE SURE TO INCLUDE A \ AS THE LAST CHARACTER. If
you do not include a \, QCA will think that the directory is a file
name, and will try to open it. Optionally, you can also include a
File Mask for QCA to use when updating the Files List. If you do
not include a File Mask, QCA will add one for you. For example, to
change from the current directory to the root directory of drive D,
position the cursor at the Name Field and type:

D:\ <─┘

If you want to search for files with the extension .DAT in the
directory E:\MYDATA, position the cursor at the name field and type:

E:\MYDATA\*.DAT <─┘

Finally, to change to directory MYDATA on the current drive, type:

-20-
 \MYDATA\ <─┘

You can also change the drive and/or directory by selecting one from
the Directories List. In this list QCA displays all system drives,
except the current one, all subdirectories of the current directory,
and the parent of the current directory ([..]). To make a change,
use Tab, Shift-Tab, or the mouse to move the cursor to the
Directories List. Highlight the desired drive or directory and
press <─┘ or double click on the item with the left mouse button.
QCA will update the Files List (using the mask(s) shown in the Name
Field) along with the Directories List.

Selecting the Appropriate File Type

As discussed above, QCA can read data from several different types
of data files. Currently, these include: QCA files, Gauss system
files, Lotus 123 (or compatible) WK1 or WKS spreadsheet files, Stata
system files, SPSS Export files, and Systat system files. When
asked to open a data file, QCA first tries to identify the file type
by comparing the file extension of the current data file to the list
of default file extensions for each type of file. If this fails,
then QCA will open a window and ask you to provide the data file's
type. To select a file type from the list that QCA supplies: (1)
highlight the appropriate type with the cursor and press the space
bar, (2) hold down the Alt key and then press the highlighted letter
for the appropriate file type (e.g., Alt+G for Gauss system file),
or (3) click on the appropriate type once with the left mouse button.

If the file is not one of the types supported by QCA, or if you do

not know what type it is, then you should not try to load the file
(select Cancel). Otherwise, you might encounter some strange
results.

-21-
 SELECTING VARIABLES FOR A TRUTH TABLE

Once a data file has been opened, you can use variables from it to
define the truth table that you want to analyze. You provide this
information to QCA through the "Build a Truth Table" window. In
addition, you can also use this window to tell QCA how to handle
missing values and how to select cases for inclusion in the truth
table. QCA opens this window automatically after you have
successfully opened a data file and assumes that you plan to use
that file. You can also access this window at any time from the
Log using the Select Variables command. In that case, QCA assumes
that you want to use the data file associated with the current Log
function, and opens that file. Use the following fields and commands
to build your truth table.

Window Fields

Dependent Variable

A text field that is used to enter the name of a

single dependent variable for the current truth
table. Every truth table MUST have a dependent
variable.

Independent Variables

A horizontally scrolling text field that is used to

enter the names of the independent variables for the
current truth table. Every truth table MUST have at
least three independent variables. The maximum
number of independent variables for a QCA truth table
is twelve.

Missing to Don't Care

A horizontally scrolling text field that is used to

enter the names of variables that will have missing
values converted to Don't Cares (either 0 or 1) by
QCA. QCA can convert any combination of dependent
and/or independent variables. If no names are
specified, QCA will drop all data cases that have a
missing value for any of the truth table variables
(dependent or independent).

-22-
Select If A horizontally scrolling text field that is used to
enter a set of logical conditions, separated by
commas, that QCA will use to select cases for
analysis. Each condition is evaluated in the order
specified, and all conditions are connected by the
logical OR. QCA will include a case in the analysis
if it meets ANY ONE of the select if conditions.
QCA evaluates Select If conditions very simply; they
CANNOT be nested in parentheses.

Table A button field that is used to compute a table of

frequencies for the specified dependent and
independent variables. For each variable, QCA lists
the number of 0, 1, - (Don't Care), and Missing
values found in the data file. This button can be
selected by highlighting it with the cursor and
pressing <─┘, by holding down the Alt key and then
pressing T, or by clicking it once with the left
mouse button.

Ok A button field that is used to quit this window and

begin reading data into the truth table using the
variables and conditions that you have specified.
This button can be selected by highlighting it with
the cursor and pressing <─┘, by holding down the Alt
key and then pressing O, by pressing the F10 function
key, or by clicking it once with the left mouse
button.

Cancel A button field that is used to quit this window and

return to the QCA Log WITHOUT building a truth table.
This button can be selected by highlighting it with
the cursor and pressing <─┘, by holding down the Alt
key and then pressing C, by pressing the Esc key, or
by clicking it once with the left mouse button.

Help A button field that is used to access QCA's online

help system. This button can be selected by
highlighting it with the cursor and pressing <─┘, by
holding down the Alt key and then pressing H, by
pressing the F1 function key, or by clicking it once
with the left mouse button.

Variable Names (A Pop Up Window)

-23-
 A list field that presents the names of all variables
in the currently opened data file. Variable names
can be entered into the text fields directly from
this list (how this is done is described below).
This list is "popped" onto the screen by holding
down the Alt key and then pressing V or by clicking
the RIGHT mouse button once.

Window Commands

Main Window

From the Keyboard

Tab Move the Cursor to the Next Field.
Shift-Tab Move the Cursor to the Previous Field.
Alt-T Select the Table Button. QCA will Open the current
data file and compute a table of frequencies for the
dependent and independent variables.
F10 OR Alt-O Select the Ok Button. QCA will open the current
data file and Build a Truth Table using the specified
Variables and Conditions.
Esc OR Alt-C Select the Cancel Button. QCA will Return
to the Log without Building a Truth Table.
F1 OR Alt-H Access QCA's online Help system.
Alt-V Pop up List of Variable Names.

With a Mouse (Left Button)

Click Once on Field or Label to Move Cursor to Field.
Click Once on Ok Button to Open the Data File and begin
Building the Truth Table.
Click Once on Cancel Button to Return to the Log without
Building a Truth Table.
Click Once on Help Button to Access QCA's online Help system.
Click RIGHT Button Once to Pop up List of Variable Names

Variable, Missing, and Select If Fields

From the Keyboard

<- Move the Cursor One Character to the Left
-> Move the Cursor One Character to the Right
Home Move the Cursor to the Beginning of the
Field
End Move the Cursor to the End of the Field

-24-
Back Space Delete the Character to the Left of the
Cursor
Del Delete the Character at the Current Cursor
Position
Alt-D Delete All Characters in the Field
Ins Switch between Insert (Block Cursor) and
Overwrite (Line Cursor) Modes

Variable Names List

From the Keyboard

 Move the Cursor Up One Line
 Move the Cursor Down One Line
PgUp Move the Cursor Up One Screen
PgDn Move the Cursor Down One Screen
Ctrl-PgUp Move the Cursor to the Top of the List
Ctrl-PgDn Move the Cursor to the Bottom of the List
Space Bar Mark/Clear Variable Name
F10 OR Alt-O Select the Ok Button. QCA will add any marked
variable names to the current field.
Esc OR Alt-C Select the Cancel Button. QCA will Return
to the Main Window without adding any
variable names.

With a Mouse (Left Button)

Click on Name Mark/Clear Variable Name
Click on  Move Cursor Up One Line
Click on  Move Cursor Down One Line
Click above _ Show Previous Screen of Variable Names
Click below _ Show Next Screen of Variable Names
Click below  Show First Screen of Variable Names
(Top of List)
Click above  Show Last Screen of Variable Names
(End of List)
Click on Ok Select the Ok Button. QCA will add any marked
variable names to the current field.
Click on Select the Cancel Button. QCA will Return
Cancel to the Main Window without adding any
variable names.

Entering Variable Names

The "Dependent Variable" field and the "Select If" field require

-25-
you to enter individual variable names. QCA provides two ways to
do this. First, you can simply move the cursor to either of these
fields (using the keyboard or mouse) and type a name of eight or
fewer characters. Variable names are NOT case sensitive; i.e.,
both upper and lower case characters are acceptable. QCA actually
converts all variable names to upper case for matching purposes.
For example, if you want to use WAR as your dependent variable,
type:

war <─┘

QCA will also provide you with a list of all variables in the current
data file and allow you to select a name from this list. To display
this list, hold down the Alt key and press V or click the RIGHT
mouse button once. To mark a variable name, highlight it and press
the space bar or click on it with the left mouse button. Repeating
this operation will clear the selected variable. Select Ok to have
QCA insert the marked variable name into the current field.

The "Independent Variables" field and the "Missing to Don't Care"

field require/allow you to enter a list of variable names. QCA
provides three ways to enter lists of variable names. First, you
can simply type all of the names, separated by commas or blank
spaces. For example,

var1,var2,var3,var4 <─┘ OR

var1 var2 var3 var4 <─┘ OR

var1,var2 var3,var4 <─┘ ETC.

Second, you can use a - to specify a group or range of variable

names. For example, if you enter

var1-var4 <─┘

QCA will use var1, var4, and ALL variables found between the two.
It is ok to combine individual variable names with variable ranges.
For example, if you want to use the first, third through sixth, and
ninth variables from a file, you could type:

var1,var3-var6,var9 <─┘

Finally, you can also enter variable names from a list provided by

-26-
QCA. To display this list, hold down the Alt key and press V or
click the RIGHT mouse button once. To mark a variable name,
highlight it and press the space bar or click on it with the left
mouse button. Repeating this operation will clear the selected
variable. Mark as many variables as you like (within the appropriate
program limits). Select Ok to have QCA insert the marked variable
names into the current field.

Processing Missing Values

Normally, QCA tries to use all of the cases in a data file when
building a truth table. However, if a case contains a missing value
(which, technically, is any value other than 0, 1, or Don't Care)
for any of the truth table's variables (dependent or independent)
QCA drops it. This is a fairly standard approach, and you will
probably want to use it most of the time. However, under certain
circumstances, you may want to override this option and have QCA
include cases even if they contain missing values for some of the
truth table variables. This can be done by having QCA convert
missing values to Don't Cares, which are acceptable Boolean truth
table values. QCA allows you to perform this conversion process
selectively. QCA will only convert missing values for those
variables (either dependent or independent) named in the "Missing
to Don't Care" field. Truth table variables not included in this
list will be handled using the default rule (any Missing Value
results in the deletion of the case).

Choosing Cases Using Select If

As discussed above, QCA normally uses every case in a data file when
building a truth table (unless a case contains missing values). You
can make QCA be more selective in its use of data cases by specifying
Select If conditions. Select If conditions are simple comparisons
between a variable and a value having the following general form:

[Variable Name] Operator [Text or Numeric Value]

where Operator is one of the following:

= (Equal to - Text or Numeric Variables)

<> (Not Equal to - Text or Numeric Variables)
> (Greater than - Numeric Variables Only)
< (Lesser than - Numeric Variables Only)
>= (Greater than or Equal to - Numeric Variables

-27-
 Only)
<= (Lesser than or Equal to - Numeric Variables
Only)

Select If conditions can use ANY variables in the data file. They
are not restricted to the truth table's dependent or independent
variables. Select If conditions are entered in the "Select If"
field. Conditions must be separated using commas. QCA treats the
commas as logical ORs and will include a case in the analysis if it
returns a "true" value for ANY ONE of the Select If conditions. For
example, given the following Select If statement:

year > 1947, wars <> ., region = west

QCA will include cases if the value of year is greater than 1947
(regardless of the values of wars and region), OR if the value of
wars is not equal to missing (regardless of the values of year and
region), OR if the value of region is equal to west (regardless of
the values of year and wars). Note that QCA treats a . as indicating
the value "missing." Also note that text values are entered the
same way as numeric values; single or double quotes are not used.
However, be aware that matches of text values are case sensitive.
QCA evaluates the value west as "not equal to" the value West. If
you are not sure whether a variable is text or numeric, hold down
the Alt key and press V. In addition to reporting the names of all
variables in the data file, QCA also indicates the type of the
highlighted variable (as specified by the data file). QCA treats
the Boolean values 0, 1, and Don't Care (from QCA data files) as
text variables. They can be compared using only two operators: =
and <>.

-28-
 MINIMIZING A TRUTH TABLE

Truth tables can contain five different kinds of configurations

based upon the values of the dependent variable. These are: 0
Configurations (dependent variable = 0), 1 Configurations (dependent
variable = 1), Don't Care Configurations (dependent variable =
either 0 OR 1), Contradictory Configurations (dependent variable =
0 AND 1), and Logical Configurations (dependent variable = NOTHING;
the configurations exist logically, but are not found in the data).
Before a truth table can be minimized, you must tell QCA which of
these configurations to minimize and which, if any, to use as don't
cares. The solution to the minimization of a truth table implies
all of the configurations that are minimized. The solution may
imply don't care configurations if they have been used. QCA uses
the "Minimize a Truth Table" window to obtain instructions from you
about how to use the configurations from the current truth table.
In addition, if the truth table is large, QCA offers you a choice
of minimization algorithms -- one that emphasizes speed, although
it may not always produce the most minimal solution, or one that
always finds the most minimal solution, although it can be VERY SLOW
with large truth tables.
Use the following window fields and commands to select your
configurations for minimization.

Window Fields

Configurations that you want to Minimize

A check box field that is used to select the

configurations that you want QCA to minimize. The
solution will logically imply all of the
configurations that you indicate here. QCA lists
all types of configurations that exist in the current
truth table. To include a group of configurations
in the solution, simply highlight them and press the
space bar or click on them once with the left mouse
button. An X marks your selection. Pressing the
space bar or clicking on them again clears your
selection. In addition, you can mark/clear
selections by pressing 0, 1, - (Don't Cares), C
(Contradictions), or L (Logical). It is OK to
include more than a single type of output in your
solution. However, including all types of
configurations in a solution will generate an error

-29-
 since the solution is the NULL SET.

Configurations to be Used as Don't Cares

A check box field that is used to select the

configurations that you want QCA to use as Don't
Cares. As their name indicates, Don't Cares may or
may not be implied by the final solution. They will
be used to generate prime implicants only if they
can help "factor" an expression into a simpler form.
Including Don't Cares is optional; a truth table
can be minimized without them. As a general rule,
using Don't Cares produces simpler solutions.
However, there is a theoretical "cost" involved; you
must be willing to assume that the Don't Cares are
essentially equivalent in output to the other
configurations in the solution. Don't Cares are
chosen in the same manner as Solution
configurations. Simply highlight a group of
configurations and use the space bar to mark/clear
your selection or click on a group with the left
mouse button. In addition, you can mark/clear
selections by pressing 0, 1, - (Don't Cares), C
(Contradictions), or L (Logical). If you do not
want to include Don't Cares, leave this field blank.

Table A button field that is used to compute a table of

frequencies for the different types of
configurations present in the current truth table.
QCA lists the number of different combinations of
independent variables associated with each type of
output (0, 1, Don't Care, Contradiction, Logical).
These frequencies "measure" the amount of diversity
found in the data; they do not indicate the number
of times that any particular combination exists.
This button can be selected by highlighting it with
the cursor and pressing <─┘, by holding down the Alt
key and then pressing T, or by clicking on it once
with the left mouse button.

Ok A button field that is used to quit this window and

begin the process of minimizing the current truth
table using the Solution and Don't Care
configurations specified. This button can be

-30-
 selected by highlighting it with the cursor and
pressing <─┘, by holding down the Alt key and then
pressing O, by pressing the F10 function key, or by
clicking on it once with the left mouse button.

Cancel A button field that is used to quit this window and

return to the QCA Log WITHOUT minimizing the truth
table. This button can be selected by highlighting
it with the cursor and pressing <─┘, by holding down
the Alt key and then pressing C, by pressing the Esc
key, or by clicking on it once with the left mouse
button.

Help A button field that is used to access QCA's online

Help system. This button can be selected by
highlighting it with the cursor and pressing <─┘, by
holding down the Alt key and then pressing H, by
pressing the F1 function key, or by clicking on it
once with the left mouse button.

Window Commands

Main Window

From the Keyboard

Tab Move the Cursor to the Next Field.
Shift-Tab Move the Cursor to the Previous Field.
Alt-T Select the Table Button. QCA will list the number of
different combinations of independent variables
associated with each type of output.
F10 OR Alt-O Select the Ok Button. QCA will begin Minimizing the
current Truth Table using the Solution and Don't
Care configurations specified.
Esc OR Alt-C Select the Cancel Button. QCA will Return to the
Log without Minimizing the current Truth Table.
F1 OR Alt-H Access QCA's online Help system.

With a Mouse (Left Button)

Click on Field or Label to Move Cursor to Field.
Click on Ok Button to Minimize the current Truth Table.
Click on Cancel Button to Return to the Log without
Minimizing the current Truth Table.
Click on Help Button to Access QCA's online Help system.

-31-
Lists of Configurations

From the Keyboard

 Move the Cursor Up One Line
 Move the Cursor Down One Line
Space Bar Mark/Clear Choice

With a Mouse (Left Button)

Single Click Mark/Clear Choice

-32-
 THE PRIME IMPLICANT CHART

Generally, QCA uses the Quine-McCluskey algorithm for minimizing a

truth table. This method assures the most minimal solution,
although it can be time-consuming for large truth tables. The
Quine-McCluskey algorithm uses a two step process to minimize a
truth table. First, all possible prime implicants of the truth
table are generated. Second, a subset of prime implicants
sufficient to imply all configurations from the truth table is
selected. This subset of prime implicants becomes the solution to
the minimization problem.

When QCA reaches the second step in the minimization process, it

opens a window so that you can watch as it selects prime implicants
for inclusion in the final solution. Before describing the
particular features of this window, however, it is necessary to
discuss the basics mechanics of truth table minimization.

Generating Prime Implicants

As mentioned above, minimizing a truth table using the Quine-

McCluskey algorithm is a two step process. Initially, QCA
compares all configurations to one another, looking for
opportunities to eliminate irrelevant variables and combine
configurations into simpler forms. For example, assume that the
configurations AbCD and ABCD (uppercase letters indicate a value
of 1 and lowercase letters indicate a value of 0) both produce
the same outcome (or outcomes that are treated as the same for
purposes of minimization). According to the logic of
minimization employed by QCA, the variable B is irrelevant for
determining the outcome in these two cases -- the outcome occurs
whether B is 0 or 1. Therefore, B can be eliminated and the two
configurations can be combined into the simpler form ACD. ACD
indicates that when the variables A, C, and D have the value 1,
the outcome(s) of interest will occur regardless of the value
for variable B. ACD is called an implicant of the two
configurations AbCD and ABCD since it logically implies each.
QCA continually applies this logic of minimization to a truth
table until it reaches the point at which implicants cannot be
reduced to simpler forms. Those remaining are called prime
implicants. Together they imply the entire set of
configurations selected for minimization from the original truth
table.

-33-
The Prime Implicant Chart

Once prime implicants have been generated, QCA must decide which
ones to include in the final solution. QCA does this by first
constructing, and then, simplifying, a prime implicant chart --
the second step in truth table minimization. As mentioned
above, the purpose of this window is to show you, step by step,
how QCA reduces the prime implicant chart and arrives at the
solution for the current minimization problem.

A prime implicant chart is a table of rows, columns, and cells,

that shows the relationship between prime implicants and the
configurations from which they were derived. QCA displays prime
implicant charts by listing configurations across the top of the
screen (columns of the chart) and prime implicants down the left
hand side of the screen (rows of the chart). An asterisk in a
cell of the chart indicates implication; i.e., the prime
implicant in the row implies, or covers, the configuration in
the column. The basic goal of chart simplification is to select
the minimum number of prime implicants needed to cover all
configurations in the chart. This set constitutes the minimized
solution for the outcome(s) of interest. QCA attempts to
simplify a prime implicant chart, and arrive at a final
solution, by applying two basic rules. First, all
configurations must be implied by at least one prime implicant
in the solution. Therefore, QCA is constantly on the lookout
for columns that contain only one asterisk. These columns
indicate configurations that are implied by only one prime
implicant. Such prime implicants are part of the core of the
solution and must be included. If they are not included, then
the solution will not imply all of the relevant truth table
configurations. QCA marks core prime implicants with an "S" to
indicate that they are part of the solution. In addition, QCA
marks as "covered" (with an "X") and removes from the chart all
of the configurations implied by these core prime implicants.
Finally, QCA marks (with an "X") and removes from the chart any
prime implicants which cover a subset of the configurations
implied by a core prime implicant (since they are now redundant;
see the next paragraph).

Once all core prime implicants have been included in the

solution, QCA looks for opportunities to simplify the remainder
of the chart by applying the second basic rule of prime
implicant chart simplification -- remove all redundant prime

-34-
implicants. Redundant prime implicants are ones which cover a
subset of the configurations implied by another prime implicant.
For example, the prime implicant Ac implies configurations ABc
and Abc. The prime implicant A implies these two configurations
as well as ABC and AbC. If these configurations and prime
implicants were in the same prime implicant chart, then the
configurations implied by Ac would be a subset of those implied
by A. Keeping Ac would add no additional information -- the
effect of Ac is redundant in light of the presence of A.
Therefore, Ac could be eliminated from the table; the
configurations covered by Ac will now simply be covered by A.
QCA marks redundant prime implicants with the letter "X" once
they have been removed from the chart.

QCA continually applies these two rules until either all

configurations have been covered (and a solution reached) or
until no further changes can be made to the chart using the
rules (at which point the user must make the final decisions
regarding which prime implicants to include in the solution).
QCA displays the number of simplifying steps taken in the lower
left corner of the screen and the current status of chart
simplification in the lower right corner of the screen. < More
Steps > means that QCA is still trying to simplify the chart;
<Finish by Hand> means that QCA is finished making changes to
the chart, but a solution has not been reached; *** Solution
*** means that QCA has simplified the chart and found a minimal
solution for the truth table.

As mentioned above, the main purpose of the "Prime Implicant

Chart" window is to allow you to watch as QCA simplifies a prime
implicant chart. Therefore, after each application of the
simplifying rules, QCA pauses and displays the current status of
the prime implicant chart on the screen. At this point, QCA
allows you to page backward and forward, viewing the status of
the prime implicant chart after each step in the simplification
process, or to terminate the simplification process, retaining
all or some prime implicants as part of the final solution.

Window Commands

To View the Chart

From the Keyboard

 Move the Chart Up One Prime

-35-
 Move the Chart Down One Prime
<- Move the Chart Left One Configuration
-> Move the Chart Right One Configuration
PgUp Move the Chart Up One Screen
PgDn Move the Chart Down One Screen
Ctrl/<- Move the Chart Left One Screen
Ctrl/-> Move the Chart Right One Screen
Home Move to the Top of the Chart
End Move to the Bottom of the Chart

With a Mouse (Left Button)

Vertical Scroll Bar (Right Side of Screen)

Click on  Move the Chart Up One Prime
Click on  Move the Chart Down One Prime
Click above _ Move the Chart Up One Screen
Click below _ Move the Chart Down One Screen
Click below  Move to the Top of the Chart
Click above  Move to the Bottom of the Chart

Horizontal Scroll Bar (Bottom of Screen)

Click on <- Move the Chart Left One Configuration
Click on -> Move the Chart Right One Configuration
Click to the
left of _ Move the Chart Left One Screen
Click to the
right of _ Move the Chart Right One Screen
Click to the
right of <- Show First Screen of Configurations
Click to the
left of -> Show Last Screen of Configurations

Other Commands

All of these commands can be entered from the keyboard or with a

mouse. To enter a command from the keyboard, simply press the
appropriate key or key combination. To enter the command with a
mouse, click the left mouse button once anywhere on the screen
display for the command (e.g., to execute the F10 command click
anywhere on the phrase "F10 Continue").

- Show the status of the prime implicant chart

prior to the last set of simplifications.
Pressing this key enough times will eventually

-36-
 return you to the original prime implicant chart.
When you press this key, you are not actually
undoing the previous step in the simplification
process. You are merely seeing what the chart
looked like before the last application of the
core and redundant rows rules.

+ Show the status of the prime implicant chart

after the next set of simplifications. Pressing
this key enough times will eventually return you
to the current state of the prime implicant
chart. As was the case with the - key, pressing
+ does not actually perform the next
simplification (that has already occurred). It
merely shows you what happened next in the
process of simplifying the current prime
implicant chart.

F9 Finish simplifying the prime implicant chart

without further pauses. QCA will no longer stop
every time that it makes a change in the prime
implicant chart. Rather, QCA will not stop again
until a solution has been found or until no more
changes can be made to the chart.

F10 Continue to simplify the prime implicant chart

and pause after the next set of changes have been
made. Pressing F10 "re-starts" QCA after it has
paused to display the fruits of its latest
efforts.

F1 Access QCA's online Help system for this window.

Esc Cancel the current minimization effort and return

to the Log.

Alt-A Stop simplifying the prime implicant chart and

include all prime implicants in the solution.
All prime implicants, regardless of their current
status, will be included in the solution.

Alt-R Stop simplifying the prime implicant chart and

include any prime implicants that have not yet
been eliminated, along with all core prime

-37-
implicants, in the final solution. If a prime
implicant already has been removed because it is
redundant (i.e., marked with an "X"), then it
will not be included in the final solution. The
status of prime implicants is determined by the
last actual step of the simplification process,
not by the step that is currently displayed on
the screen.

-38-
 REDUCING THE PRIME IMPLICANT CHART BY HAND

As discussed in the previous section, QCA is not always able to

completely minimize a prime implicant chart. If this occurs,
then you must make the final decisions regarding which prime
implicants to include in the solution. The current window
allows you to mark those primes that you want to include in the
final solution using either the space bar or a mouse. Each time
that you mark a prime implicant, QCA updates the prime implicant
chart using the following symbols:

a  next to a prime indicating that YOU have marked it for

inclusion in the final solution;

O's across the top of the chart indicating configurations

that are now covered by one or more of the primes that you
have marked;

o's in cells (replacing *'s) indicating implications that

are now irrelevant because the affected configurations have
been covered.

QCA also keeps a running total of the number of prime implicants

that are currently included in the solution (displayed in the
lower left corner of the screen) and the number of
configurations that are still not covered by the current
solution (displayed in the lower right corner of the screen).
If the chart is large, only part of it can be displayed on the
screen. Therefore, you may need to use the cursor control keys
to scroll up and down or left and right in order to view
remaining primes and uncovered configurations. Once all of the
configurations have been covered, QCA displays a "*** Solution
***" message in the lower right of the screen. When this
message appears, press the F10 key (or click on it with a
mouse) to return to the Log with your solution. If you do not
like a particular solution, you can use the space bar or mouse
to "unmark" primes, restoring their implied configurations to
the pool of uncovered configurations.

Different criteria can be used to select primes for inclusion in

the final solution. From a strict minimization perspective, you
want to select the minimum number of primes (preferably, those
having the fewest variables) necessary to cover the remaining
configurations. This usually means marking primes that have a

-39-
large number of *'s in their rows (indicating that they imply a
large number of the remaining configurations). However, from a
theoretical standpoint, you may want to mark any primes that
contain certain variables, regardless of their relative ability
to cover configurations. However you approach the problem, the
ultimate goal is to keep marking primes until you cover all
remaining configurations. If you want to continue marking
primes beyond this point, that is fine. You will still have a
valid solution to the minimization of the truth table -- it just
won't be the most minimal solution.

Window Commands

To Highlight and Mark Prime Implicants

From the Keyboard

Space Bar Mark/Clear Prime
 Move the Cursor Up One Prime
 Move the Cursor Down One Prime
<- Move the Screen Left One Configuration
-> Move the Screen Right One Configuration
PgUp Move the Cursor Up One Screen
PgDn Move the Cursor Down One Screen
Ctrl/<- Move Configurations Left One Screen
Ctrl/-> Move Configurations Right One Screen
Home Move the Cursor to the Top of the List
End Move the Cursor to the Bottom of the List

With a Mouse (Left Button)

Click on
Prime Move Cursor to Prime and Mark/Clear

Vertical Scroll Bar (Right Side of Screen)

Click on  Move the Cursor Up One Prime
Click on  Move the Cursor Down One Prime
Click above _ Move the Cursor Up One Screen
Click below _ Move the Cursor Down One Screen
Click below  Move the Cursor to the Top of the List
Click above  Move the Cursor to the Bottom of the List

Horizontal Scroll Bar (Bottom of Screen)

Click on <- Move the Screen Left One Configuration

-40-
Click on -> Move the Screen Right One Configuration
Click to the
left of _ Move Configurations Left One Screen
Click to the
right of _ Move Configurations Right One Screen
Click to the
right of <- Show First Screen of Configurations
Click to the
left of -> Show Last Screen of Configurations

Other Commands

All of these commands can be entered from the keyboard or with a

mouse. To enter a command from the keyboard, simply press the
appropriate key or key combination. To enter the command with a
mouse, click the left mouse button once anywhere on the screen
display for the command (e.g., to execute the F10 command click
anywhere on the phrase "F10 Done").

F10 Accept the current minimization and return to the

Log.

F1 Access QCA's online Help system.

Esc Cancel the current minimization effort and return

to the Log.

Alt-A Stop simplifying the prime implicant chart and

include all prime implicants in the solution.
All prime implicants, regardless of their current
status, will be included in the solution.

Alt-R Stop simplifying the prime implicant chart and

include any prime implicants that have not yet
been eliminated, along with all core prime
implicants, in the final solution. If a prime
implicant already has been removed because it is
redundant (i.e., marked with an "X"), then it
will not be included in the final solution.

-41-
 THE FUNCTION EDITOR

The Function Editor lets you view a table of frequencies for a

Boolean function. Each row of the table displays a function
term along with the number and percentage of 0, 1, and - (don't
care) truth table configurations and raw data cases implied by
the term. The totals for the truth table and data file are
displayed at the bottom of the table (labeled "Total"). QCA
also identifies the truth table configurations and CaseIds (if
included in the data file) associated with each function term.
You can view this information with the View CaseIds command
(Alt-I) described below. Any type of Boolean function can be
loaded into this editor except a truth table or intersection.
Frequencies will be computed using the truth table and data file
used or "inherited" by the function.

The Function Editor also lets you measure the impact on total
frequencies of the addition or deletion of specific function
terms. QCA displays this information in a row labeled "Checked"
at the bottom of the table. Only terms marked with a  are
included in the computation of the frequencies reported in this
row. Use the space bar to mark or clear the term at the cursor,
or double click the left mouse button on a term. The Recompute
Table command (Alt-R) will compute frequencies using only the
marked terms and then update the information in the "Checked"
row. By adding and deleting terms, you can assess the relative
power of a group of terms to cover the truth table and/or raw
data. You can also get a pretty good sense of the extent to
which function terms overlap in the truth table configurations
that they imply.

QCA displays function terms using 0, 1, and - notation. To make

it easier to identify terms, QCA lists the names of the
variables in the function's model across the top of the screen
and the value of the term at the cursor using these variable
names at the bottom of the screen.

Cursor Movement

From the Keyboard

 Move Cursor Up One Line
 Move Cursor Down One Line
PgUp Move Table Up One Screen
PgDn Move Table Down One Screen

-42-
 Ctrl-Home Move Cursor to Top Line on Screen
Ctrl-End Move Cursor to Bottom Line on Screen
Ctrl-PgUp Show First Screen of Table
Ctrl-PgDn Show Last Screen of Table
Space Bar Mark/Clear Term at Current Line

With a Mouse (Left Button)

Click on  Move Cursor Up One Line
Click on  Move Cursor Down One Line
Click above _ Move Table Up One Screen
Click below _ Move Table Down One Screen
Click below  Show First Screen of Table
Click above  Show Last Screen of Table
Click on Line Move Cursor to Line
Double Click on Line Move Cursor to Line & Mark/Clear Term

File Commands.....................................Alt-F

Which File?..................................Alt-W
The Which File? command tells you exactly "which file" is
linked to the function in the editor.

Save Output..................................Alt-O
The Save Output command allows you to save either the
current screen or the entire table to an output file. QCA
asks you to enter a filename (default = QCA.OUT), select
the amount of text to be saved (All or Screen), and select
how the file should be written (replace the output file by
Overwriting it or add to the output file by Appending text
to the end of it).

Print Output.................................Alt-P
The Print Output command allows you to send output directly
to a printer (LPT1) if one is connected. It is identical
to the Save command, except that it (obviously) does not
need you to indicate whether to Overwrite or Append
results.

Exit Editor..................................Alt-X or Click

On "Alt-X to Exit"
The Exit Editor command allows you to exit the function
editor and return to the Log.

Edit Commands.....................................Alt-E

-43-
 View CaseIds.................................Alt-I or Click
Right Mouse Button
The View CaseIds command allows you to browse through the
CaseIds for the data cases implied by the current function
term. QCA presents the CaseIds by truth table
configuration, and also reports the number of function
terms that imply each configuration (providing a rough
measure of overlap among function terms).

Goto CaseId..................................Alt-G
The Goto CaseId command allows you to search for a
particular CaseId and, if found, moves the cursor to the
function term that implies it. QCA provides a window for
you to enter the CaseId and to specify the direction of the
search (forward or backward from the current cursor
position), the type of match (exact or partial), and case
sensitivity (match by case or not). If the match is
successful, you can repeat it by simply pressing Shift-
(repeat search in a backward direction) or Shift- (repeat
search in a forward direction). If a match is found, QCA
automatically calls the View CaseIds command so that you
can view the CaseId.

Methods Commands..................................Alt-M

Recompute Table..............................Alt-R
The Recompute Table command allows you to recompute
frequencies for a function using only those terms that you
have marked. The row labeled "Checked" at the bottom of
the table reports the number of truth table configurations
and raw data cases implied by all checked terms combined.
By adding and deleting terms, and then recomputing the
table, you can get a good sense of the degree of overlap
among terms in the function. You can Mark or Clear a term
by highlighting it with the cursor and pressing the Space
Bar, or by double clicking on it with the left mouse
button. QCA marks terms with a .

-44-
 THE TRUTH TABLE EDITOR

Qualitative Comparative Analysis focuses on the construction and

minimization of truth tables. Therefore, it is appropriate that
QCA provide an editor designed specifically for the viewing and
editing of truth tables. The Truth Table Editor lists every
unique configuration of independent variables found in the data,
along with the number and percentage of 0, 1, and don't care (-)
cases associated with the configuration. The value of the
dependent variable for a configuration is a function of the
distribution of 0, 1, and don't care cases. If a configuration
occurs in the data with both 0 and 1 values on the dependent
variable, then QCA treats the configuration as a contradiction
and assigns the dependent variable the value C. Otherwise, QCA
assigns the dependent variable for a configuration the value of
0 (only 0 and don't care cases exist), 1 (only 1 and don't care
cases exist), or - (only don't care cases exist).

QCA displays truth table configurations using 0, 1, -, and C

notation, with the dependent variable (output) as the far right
value in the configuration (above the cursor). To make it
easier to identify configurations, QCA lists the names of the
variables in the truth table across the top of the screen and
the value of the configuration at the cursor using these
variable names at the bottom of the screen.

Cursor Movement

From the Keyboard

 Move Cursor Up One Line
 Move Cursor Down One Line
PgUp Move Table Up One Screen
PgDn Move Table Down One Screen
Ctrl-Home Move Cursor to Top Line on Screen
Ctrl-End Move Cursor to Bottom Line on Screen
Ctrl-PgUp Show First Screen of Table
Ctrl-PgDn Show Last Screen of Table
0,1,-,C Move Cursor to Next Line with this
Output Value

With a Mouse (Left Button)

Click on  Move Cursor Up One Line
Click on  Move Cursor Down One Line
Click above _ Move Table Up One Screen

-45-
 Click below _ Move Table Down One Screen
Click below  Show First Screen of Table
Click above  Show Last Screen of Table
Click on Line Move Cursor to Line

File Commands.....................................Alt-F

New Data File................................Alt-N

The New Data File command allows you to save the current
truth table to a QCA data matrix file. QCA will ask you to
provide a name for the file, and will warn you if the file
already exists.

Which File?..................................Alt-W
The Which File? command tells you exactly "which file" is
linked to the truth table in the editor.

Save Output..................................Alt-O
The Save Output command allows you to save either the
current screen or the entire truth table summary to an
output file. QCA asks you to enter a filename (default =
QCA.OUT), select the amount of text to be saved (All or
Screen), and select how the file should be written (replace
the output file by Overwriting it or add to the output file
by Appending text to the end of it). The Save Output
command saves a summary of the truth table so that you can
view or print it later; it does NOT create a data matrix
file that can be analyzed using QCA. Use the New Data File
command to create a QCA data matrix file.

Print Output.................................Alt-P
The Print Output command allows you to send output directly
to a printer (LPT1) if one is connected. It is identical
to the Save command, except that it (obviously) does not
need you to indicate whether to Overwrite or Append
results.

Exit Editor..................................Alt-X or Click

On "Alt-X to Exit"
The Exit Editor command allows you to exit the truth table
editor and return to the Log.

Edit Commands.....................................Alt-E

-46-
 View CaseIds.................................Alt-I or Click
Right Mouse Button
The View CaseIds command allows you to browse through the
CaseIds for the data cases associated with the current
truth table configuration.

Goto CaseId..................................Alt-G
The Goto CaseId command allows you to search for a
particular CaseId and, if found, moves the cursor to the
truth table configuration that implies it. QCA provides a
window for you to enter the CaseId and to specify the
direction of the search (forward or backward from the
current cursor position), the type of match (exact or
partial), and case sensitivity (match by case or not). If
the match is successful, you can repeat it by simply
pressing Shift- (repeat search in a backward direction) or
Shift- (repeat search in a forward direction). If a match
is found, QCA automatically calls the View CaseIds command
so that you can view the CaseId.

Recode Output................................Alt-R
In the Truth Table Editor, QCA displays the number and
percentage of 0, 1, and don't care (-) cases associated
with each truth table configuration as well as the specific
output value assigned to each configuration based on these
cases. The Recode Output command allows you to change the
output value assigned to a configuration by assigning new
values to the 0, 1, and don't care cases. QCA opens a
window showing the current coding of 0, 1, and don't care
cases and gives you an opportunity to enter new output
values for these cases. QCA then updates the output value
assigned the truth table configuration and marks the recode
by placing "R+New Value" next to the output cases recoded.
This command is useful for recoding contradictory
configurations to 0 or 1 based upon their tendencies in the
data.

Methods Commands..................................Alt-M

Summarize Table..............................Alt-T
The Summarize Table command displays the total number of 0,
1, Don't Care, Contradictory, and Logical Configurations
for the truth table. QCA provides this information for the
current state of the truth table (taking into account any

-47-
recodes) under the column labeled "New" as well as for the
original state of the truth table (prior to any recodes)
under the column labeled "Old." This makes it easier to
see what impact, if any, recoding has had on the
composition of the truth table.

-48-
 THE DATA MATRIX EDITOR

You can use the Data Matrix Editor to create or edit raw data
matrices. Each row is a "case" of data, and each column
represents a variable. The dependent variable is always the
last column of the matrix. To enter a value for a variable,
simply press the 0, 1, or - key. If a data set contains
CaseIds, they will be listed to the left of the raw data.
CaseIds can be edited by pressing the space bar or clicking the
right mouse button.

QCA's data matrix editor recognizes only three values: 0, 1, and

-. Missing values are not permitted and "blanks" appear as
don't cares (-). QCA lists the data file being edited and the
names of the variables in the matrix across the top of the
screen. To make it easier to read the data, QCA shows the value
of the configuration at the cursor using variable names at the
bottom of the screen.

Cursor Movement

From the Keyboard

<─┘ Add Blank Line
Ctrl-<─┘ Add Blank Line and Place Cursor in
CaseId field.
-> Move Cursor Right One Space
<- Move Cursor Left One Space
Backspace Move Cursor Left One Space and Insert
-
Home Move Cursor to Beginning of Line
End Move Cursor to End of Line
 Move Cursor Up One Line
 Move Cursor Down One Line
PgUp Show Previous Screen of Data
PgDn Show Next Screen of Data
Ctrl-Home Move Cursor to Top Line on Screen
Ctrl-End Move Cursor to Bottom Line on Screen
Ctrl-PgUp Show First Screen of Data
Ctrl-PgDn Show Last Screen of Data
Space Bar Move Cursor to CaseId

With a Mouse (Left Button)

Click on  Move Cursor Up One Line
Click on  Move Cursor Down One Line

-49-
 Click above _ Show Previous Screen of Data
Click below _ Show Next Screen of Data
Click below  Show First Screen of Data
Click above  Show Last Screen of Data
Click on Line Move Cursor to Line
Click Right Button Move Cursor to CaseId

File Commands.....................................Alt-F

New Data File................................Alt-N

The New Data File command allows you to save the current
data matrix to a QCA file. QCA will ask you to provide a
name for the file, and will warn you if the file already
exists.

Save Output..................................Alt-O
The Save Output command allows you to save either the
current screen or the entire data matrix to an output file.
QCA asks you to enter a filename (default = QCA.OUT),
select the amount of text to be saved (All or Screen), and
select how the file should be written (replace the output
file by Overwriting it or add to the output file by
Appending text to the end of it). The Save Output command
saves a listing of the data so that you can view or print
it later; it does NOT create a data matrix file that can
be analyzed using QCA. Use the New Data File command to
create a QCA data matrix file.

Print Output.................................Alt-P
The Print Output command allows you to send output directly
to a printer (LPT1) if one is connected. It is identical
to the Save command, except that it (obviously) does not
need you to indicate whether to Overwrite or Append
results.

Exit Editor..................................Alt-X or Click

On "Alt-X to Exit"
The Exit Editor command allows you to exit the data matrix
editor and return to the Log.

Edit Commands................................Alt-E

Insert Line..................................Alt-I

-50-
 The Insert Line command inserts a blank line at the cursor.
Lines below the cursor are pushed down to provide room for
the new line.

Delete Line..................................Alt-D
The Delete Line command deletes the line at the cursor.
Lines below the cursor are pulled up to replace the deleted
line.

Add Logical Configurations...................Alt-L

The Add Logical Configurations command identifies any
configurations that exist logically, but are not currently
part of the data, and appends them to the end of the data
matrix. QCA assigns the dependent variable for these
configurations the value "L". Once created, you can edit
these configurations just like any other lines of data in
the editor.

Find Contradiction...........................Alt-C
The Data Matrix editor lists data case by case. Because
the data are not summarized in the form of a truth table,
it can be very difficult to identify contradictory
configurations. The Find Contradiction command does this
for you. When you execute this command, QCA looks for the
next configuration of independent variables in the data
that has both a 0 and 1 value on the dependent variable,
beginning with the line of data at the cursor. If a
contradiction is found, then QCA moves the cursor to the
next occurrence of this configuration.
Methods Commands..................................Alt-M

Browse Truth Table...........................Alt-B

The Browse Truth Table command creates a truth table from
the raw data in the editor and then lists each unique
combination of independent variables along with the number
of cases in which the dependent variable associated with
the configuration has a value of 0, 1, or don't care. This
command gives you a quick summary of the truth table that
could be minimized using the data in the editor. This
information is similar to that available in the Truth Table
Editor.

Summarize Table..............................Alt-T
The Summarize Table command creates a truth table from the

-51-
raw data in the editor and then displays the total number
of 0, 1, Don't Care, Contradictory, and Logical
Configurations in the truth table.

Summarize Row................................Alt-R
The Summarize Row command counts the number of times that
the configuration at the cursor occurs with an output value
of 0, 1, or don't care and then reports this information
next to the configuration. This information is similar to
that available in the Truth Table Editor.

-52-
 THE HYPOTHESIS EDITOR

The Hypothesis Editor allows you to create or edit Boolean

hypotheses which can then be "tested" by computing their
intersections with minimized truth tables. Boolean hypotheses
are nothing more than Boolean functions, and can be represented
as simplified data matrices where each row is a hypothesized
configuration and each column is a variable. To enter a value
for a variable, simply press the 0, 1, or - key. Unlike raw
data matrices, however, hypotheses do not contain a column for
the dependent variable.

QCA's hypothesis editor is actually a simplified version of the

data matrix editor. Like that editor, the hypothesis editor
recognizes only three values: 0, 1, and -. Missing values are
not permitted and "blanks" appear as don't cares (-).
Hypotheses inherit their variables from truth tables. QCA
displays the inherited model for a hypothesis across the top of
the screen. To make it easier to read hypotheses, QCA shows the
value of the configuration at the cursor using variable names at
the bottom of the screen.

Cursor Movement

From the Keyboard

<─┘ Add Blank Line
Ctrl-<─┘ Add Blank Line and Place Cursor in
CaseId field.
-> Move Cursor Right One Space
<- Move Cursor Left One Space
Backspace Move Cursor Left One Space and
Insert -
Home Move Cursor to Beginning of Line
End Move Cursor to End of Line
 Move Cursor Up One Line
 Move Cursor Down One Line
PgUp Show Previous Screen of Data
PgDn Show Next Screen of Data
Ctrl-Home Move Cursor to Top Line on Screen
Ctrl-End Move Cursor to Bottom Line on Screen
Ctrl-PgUp Show First Screen of Data
Ctrl-PgDn Show Last Screen of Data

With a Mouse (Left Button)

-53-
 Click on  Move Cursor Up One Line
Click on  Move Cursor Down One Line
Click above _ Show Previous Screen of Data
Click below _ Show Next Screen of Data
Click below  Show First Screen of Data
Click above  Show Last Screen of Data
Click on Line Move Cursor to Line

File Commands.....................................Alt-F

Which File?..................................Alt-W
The Which File? command tells you exactly "which file" is
inherited by the hypothesis in the editor. This file comes
from the truth table that is the model for the hypothesis.

Save Output..................................Alt-O
The Save Output command allows you to save either the
current screen or the entire hypothesis to an output file.
QCA asks you to enter a filename (default = QCA.OUT),
select the amount of text to be saved (All or Screen), and
select how the file should be written (replace the output
file by Overwriting it or add to the output file by
Appending text to the end of it). The Save Output command
saves a listing of the hypothesis so that you can view or
print it later; it does NOT create a separate data file
for the hypothesis.

Print Output.................................Alt-P
The Print Output command allows you to send output directly
to a printer (LPT1) if one is connected. It is identical
to the Save command, except that it (obviously) does not
need you to indicate whether to Overwrite or Append
results.

Exit Editor..................................Alt-X or Click

On "Alt-X to Exit"
The Exit Editor command allows you to exit the hypothesis
editor and return to the Log.

Edit Commands................................Alt-E

Insert Line..................................Alt-I
The Insert Line command inserts a blank line at the cursor.
Lines below the cursor are pushed down to provide room for

-54-
the new line.

Delete Line..................................Alt-D
The Delete Line command deletes the line at the cursor.
Lines below the cursor are pulled up to replace the deleted
line.

-55-