IF YOU ARE USING A GRAPHICS BASED WORD PROCESSOR SOFTWARE PROGRAM, IT IS
RECOMMENDED THAT THIS MANUAL BE VIEWED AND/OR PRINTED USING A MONOSPACE
FONT TO ENSURE CORRECT ALIGNMENT OF TABLE VALUES AND FORMS.



SURVWARE EVALUATION PROGRAM
Version 1.3
January 1990

PROGRAM DOCUMENTATION
Copyright (C) 1990 CompStat Software, All Rights Reserved

No part of this manual or documentation may be reproduced, transmitted,
transcribed, stored in a retrieval system, or translated into any
language (natural or computer), in any form or by any means, except as
described in the following license without the prior written permission
of CompStat Software.

You are granted a limited license to use the software described in this
manual.  The software may be used or copied only in accordance with the
terms of that license, which is described below.

Information in this manual is subject to change without notice and does
not represent a commitment on the part of CompStat Software.

CompStat Software may make improvements and/or changes in this manual
and/or in the Survware Evaluation Program and/or in Survware at any
time.

Survware is a registered trademark, and Survware, Survware Evaluation
Program, CompStat, the CompStat logo, and the Survware logo, are
trademarks of CompStat Software.

IBM is a registered trademark of International Business Machines
Corporation.


THIS SOFTWARE IS NOT FOR SALE

The Survware Evaluation Program is subject to the following license
terms and conditions.


SINGLE COPY SOFTWARE LICENSE

THIS IS A SINGLE COPY SOFTWARE LICENSE granted by CompStat Software, an
Indianapolis company, with its mailing address at P.O. Box 16367,
Indianapolis, IN  46216.  The Survware Evaluation Program is licensed to
you as the end user.  It is not sold.

The Survware Evaluation Program is copyrighted material.  You may use it
on a trial basis provided you do not violate the copyright and if you
follow these simple rules.  Use of the Survware Evaluation Program on a
regular basis requires purchase as described later.

You may not make any changes or modifications to the Survware Evaluation
Program, and you may not decompile, disassemble, or otherwise reverse-
engineer it.  You may not rent or lease it to others.

You may make copies of the Survware Evaluation Program only under the
terms of the following section entitled "Limited License to Copy".

This software is Commercial Computer Software under Federal Government
Acquisition Regulations and agency supplements to them.  The Software is
provided to the Federal Government and its agencies only under the
Restricted Rights Provisions of the Federal Acquisition Regulations
applicable to commercial computer software developed at private expense
and not in the public domain.  The use, duplication, or disclosure by
the Government is subject to restrictions as set forth in subdivision
(c)(1)(ii) of the Rights in Technical Data and Computer Software clause
at 252.227-7013.


LIMITED LICENSE TO COPY

You are granted a limited license to copy the Survware Evaluation
Program only for the trial use of others subject to the software license
described above, and also the following:

*  The Survware Evaluation Program must be copied in unmodified form,
complete with the following files:

  README  -  The Survware Evaluation Program documentation

  SURVEY.EXE  -  The Survware Evaluation Program

*  The full and unmodified Survware Evaluation Program documentation
(README) must be included with the copy.

*  No fee, charge or other compensation may be accepted or requested by
anyone without express written permission from CompStat Software.  This
includes disk vendors.

*  The Survware Evaluation Program may not be distributed in conjunction
with any other product or service without a specific license to do so
from CompStat Software.

*  The Survware Evaluation Program may be distributed only in North
America.  DISTRIBUTION OF the Survware Evaluation Program OUTSIDE OF
NORTH AMERICA IS STRICTLY FORBIDDEN.

Operators of electronic bulletin board systems (Sysops) may post the
Survware Evaluation Program files for downloading by their users without
written permission only as long as the above conditions are met.  A fee
may be charged for access to the BBS as long as no specific fee is
charged for access to the Survware Evaluation Program files.

Distributors of public domain and user supported software, such as disk
vending services, may distribute copies of the Survware Evaluation
Program subject to the above conditions only after obtaining written
permission from CompStat Software.  Such permission usually is granted;
please write for details.


INTRODUCING THE SURVWARE EVALUATION PROGRAM

The Survware Evaluation Program is the "try before you buy" version of
Survware.  CompStat Software has created it to allow you to "sample"
Survware before committing your hard earned dollars.  This has benefits
for both the users and the authors as described below.

First of all, you must be aware that Survware and the Survware
Evaluation Program are not free software, nor are they in the public
domain.  Both programs are fully copyrighted works protected to the full
extent of U.S. copyright laws.

Why allow a "sample" program?  The Survware Evaluation Program is based
on a marketing technique called "shareware" or "user-supported
software," which is a means for the computing community to receive
quality software while directly supporting authors.  It is based on the
idea that the value and utility of software is often best assessed by
the user on his or her own system.  Only after using a program can one
really determine whether it serves personal applications, needs and
tastes.

The benefits to users are fairly obvious: you are allowed to thoroughly
evaluate the software before purchasing it.  How many times have you
purchased an expensive piece of software only to find that it doesn't do
what you need?

Software authors benefit by creating a distribution channel of satisfied
users who recommend and pass along their product.  This "pyramid effect"
can be a very effective marketing technique, and is vastly less
expensive than traditional marketing techniques.  This is not to say
that "sample" programs and more traditional methods like magazine
advertising are mutually exclusive.  They may be used together very
effectively to target different segments of the market.

What does this mean to you?  While you are granted a limited license to
evaluate the program on a trial basis, regular use of the program
requires purchase.  What is a trial basis?  Kicking the program around
over a weekend to see how it works is certainly using it on a trial
basis.  Installing and using the program in a production test
environment for two weeks is trial use.  Using the software twice a week
for regular work for three months is not and therefore requires
purchase.  Using the program regularly for any regular task, especially
if it has replaced other software, is not trial use.  These examples are
not intended as definitive rules, but more as general guidelines.  If
you have used the program longer than 15 days on a regular basis, you
should then register Survware.

You may copy the program for the trial use of others subject to the
rules outlined in the previous section "Limited License to Copy."  In
general, you may only copy the program in complete and unmodified form,
as we distribute it.

You may not charge any fee for copies of the program without written
permission from CompStat Software.  This applies especially to disk
vendors.  Some services mislead customers into thinking they have
purchased the software, rather than just the disk and duplication
service.  If you have purchased this product from a disk vendor and feel
they were misleading about what you have purchased, please write us and
we will investigate the matter.

You may not include the Survware Evaluation Program with any other
product, whether hardware or software, without specific license.  For
example, you may not include the Survware Evaluation Program on hard
drives of computers you sell, or include it with copies of software you
have developed.


PURCHASING SURVWARE

After using the Survware Evaluation Program for a while, you need to
decide if you want to continue using it on a regular basis.  If not,
please feel free to continue copying it for the trial use of others, or
to pass your copy along to a friend.

If you do decide to continue using it, it must be purchased.  You can
purchase the regular version of Survware directly from CompStat
Software; it can also be found at many retailers.

Please note that there are two versions of the software: Survware and
the Survware Evaluation Program.  The evaluation program version may be
copied and used for evaluation as described above.  The regular version
of Survware, however, may only be copied for backup purposes.  After
receiving your regular version of Survware you may continue to copy the
Survware Evaluation Program for friends; you may not copy Survware.

Why should you purchase Survware?  To pay for the software which you are
gainfully using, and to support its authors and encourage them to
produce more high quality software.  And because regular use of the
Survware Evaluation Program without purchase is a violation of the
license under which you use it.

When you purchase the regular version of Survware you will receive in
addition to a new program disk the following:

*  The latest version of the program.
*  Perpetual license to use the software you have purchased.
*  A printed manual.
*  90 days free telephone technical support.
*  Reduced costs for upgrades to future versions of Survware.

To order, send a check or money order for $50 to our mailing address at
P.O. Box 16367, Indianapolis, IN  46216-0367.  If you have any questions
call our business office at (317) 898-4186.  If you have purchased the
Survware Evaluation Program from CompStat Software, send $40 with proof
of purchase of the Survware Evaluation Program.  An order form is at the
end of this documentation.  An additional 10% discount is available to
educational organizations, their staff and faculty, and students (proof
required with order form).



SURVWARE MANUAL
Copyright (C) 1990, CompStat Software, Indianapolis, Indiana

This manual and the software described in it are copyrighted, with all
rights reserved.  Under the copyright laws, this manual and/or the
software may not be copied, in whole or in part, without the express
written consent of CompStat Software, except in the normal use of the
software or to make a backup copy of the software.  The same proprietary
and copyright notices must be affixed to any permitted copies as were
affixed to the original.  This exception does not allow copies to be
made for others, whether or not sold, but all the material purchased
(with all backup copies) may be sold, given, or loaned to another
person.  Under the law, copying includes translating into another
language or format.  You may use the software on any computer owned by
you, but extra copies cannot be made for this purpose.



CHAPTER 1  -  USING SURVWARE


About Survware

Survware is a simple to use, time-saving input and analysis software
tool designed for use with opinion surveys which have no more than 50
close-ended type questions.  As many as 10 of the questions can be used
to analyze the survey based on socio-economic type demographic
groupings.  All questions can have up to 10 response choices (except the
demographic questions which can have up to nine choices).  The number of
questionnaires the program can analyze is limited only by the amount of
disk space available from where the program was started.  To find out
how many questionnaires are needed for a survey, see the short tutorial
on opinion surveys in this manual.


Survware Requirements

-  A computer fully compatible with an IBM PC, XT, AT, or PS/2 computer
with a PC-DOS operating system, version 2.0 or higher (or equivalent).

-  At least 384KB of random access memory (RAM)

-  At least one 360KB-5" or one 720KB-3" floppy disk drive; or a hard
disk drive

-  An 80-column monochrome; or an 80-column color monitor with a CGA,
EGA, VGA, or equivalent display card


Installing Survware

Before operating Survware on a computer with only a floppy disk system,
you should make a working copy of the disk and use that copy whenever
you want to start the program.

If Survware will be installed on a hard disk, it is suggested that it be
placed in its own sub-directory.  All initialization responses, survey
data and analysis reports are saved to the sub-directory from where
Survware is started.  A batch file is included with the program to
assist with this.  Place the Survware distribution disk in drive A.  At
the A> prompt type:

       SETUP  C:

and press the ENTER key.

A sub-directory named SURVEY will be created on drive C, and the
software will be copied to it.  When drive A stops turning, the computer
will be logged onto C:\SURVEY.  If you wish to install the software on
a hard drive other than C, substitute the appropriate drive letter in
the above command.

To start Survware, you should be logged onto the current directory and/or
drive where the software is installed.  At the prompt type:

       SURVEY

and press the ENTER key.


Initializing Survware

The first time you start Survware you must answer the following
initialization questions:

1.  How many questions are in your survey questionnaire (at least two,
maximum of 50)?

2.  Which questions ask for the respondent's demographic groupings
(at least one, maximum of 10)?

3.  How many response choices are there for each question in the survey
questionnaire (at least two, maximum of 10 except for demographic grouping
questions which have a maximum of nine)?

4.  What is the text of the choices for each of the questions which ask 
for the respondent's demographic groupings?

During the initialization process, you will be allowed to review your
answers at various stages.  If you complete the initialization process
and there is something that you still are not satisfied with, then reset
the program by selecting "Reset the Software" from the menu (see
Resetting Survware below).

For the "text" of the choices for question #4 above, the program will
only accept up to five characters for each choice.  The "text" which you
input is used later in the reports' column headings.

After all initialization questions have been answered, the program will
save your responses to a file.  The program will then display the
Survware menu.

Each time you start Survware after it has been initialized, you will see
the same Survware menu which was displayed at the end of the
initialization process.  The Survware menu offers four choices:

  1. Input Survey Data
  2. Analyze Data
  3. Exit the Program
  4. Reset the Software


Questionnaire Data Input

By selecting "Input Survey Data" from the menu, you will be able to
enter individual question responses from one questionnaire into
Survware.  First, the program will display a screen with a cursor next
to the words "Question 1 = " and will wait for you to type a lowercase
letter and press the ENTER key.  After entering an acceptable answer
choice and pressing the ENTER key, each succeeding question will be
displayed, one at a time.  Only one letter can be entered for each
question.

Survware is designed to be used with only close-ended types of
questions.  If your questionnaire has any "True-False" questions, then
simply substitute the letters "a" or "b" respectively when inputting the
responses.  If any survey questions are open-ended, then group the
answers into similar categories and assign letters (up to "j" except for
demographic grouping questions up "i").

After selections have been entered for all the questions from the one
questionnaire, the entire questionnaire is displayed with the responses
next to each question number.  After review, if you find you entered an
answer differently than what the respondent indicated on the
questionnaire, then you can press the "N" key in response to the prompt,
"Is this correct?"  You will then be prompted to type the number of the
incorrect question.  The question will be redisplayed.  After entering the
correct response for this question, the questionnaire will be displayed.
The questionnaire will be redisplayed for review in this manner again and
again until such time you are satisfied with the selections and press the
"Y" key in response to the prompt, "Is this correct?"

After indicating the questionnaire is completely correct, all the
selections for that questionnaire will be saved to the survey database
file.  Then you will be asked if you want to enter another
questionnaire.  If you do, press the ENTER key.  If you do not want to
enter another questionnaire's responses, press the ESCAPE key and you
will be returned to the menu.

If you pressed the ENTER key by mistake and did not want to enter
another questionnaire, you can correct it by pressing only the ENTER
key for the response to any question.  After a reminder that none of the
entries which were made to the screen questionnaire, if any, will be saved
to the survey database file, you will be returned to the menu.

A running total of the number of completed questionnaires in the survey
database file is displayed at the bottom of the screen during all
questionnaire data input.


Creating Reports

By selecting "Analyze Data" from the menu, you will be able to create an
analysis report based on one of the demographic group questions
identified in the initialization process.  The program will display the
first demographic group's question number.  If this is the group you
want to analyze the data by, press the ENTER key.  If you want to
analyze the data by a different group, press the ESCAPE key and the
second demographic group's question number will be displayed.  You can
loop through all the demographic group questions in this manner until
the one you want is displayed.  If you loop through to the last
demographic group question and press the ESCAPE key, then you will be
returned to the menu.

After selecting a demographic group question, the program will remind
you that the report is written to disk as an ASCII file named GROUP1, or
GROUP2, etc., relative to which demographic group you selected.  Any
previous versions of the report based on the same demographic group
question will be replaced.  By pressing the ENTER key, the computer will
begin "thinking" about the questionnaire responses.  The first
question's results will be displayed in less than a minute (depending on
how many questionnaires the computer has to think about).  You can stop
the analysis at any time by pressing the ESCAPE key.

For each question analyzed, three cross-tab tables are created: one by
absolute frequency which reports the number of answers within each
demographic sub-group for each possible answer choice; and two by
relative frequency which report the percentage of answers within (by
column) each demographic sub-group and one with the percentage of
answers across (by row) the demographic sub-groups for each possible
answer choice.  Column and row totals are also provided in the absolute
and relative frequency (by column) tables.

Since Survware does not contain file text editing capabilities, plain
text reports (ASCII) are created so they can be edited by any IBM
compatible word processing software.  This feature allows you to use
your word processing software to enter the text of each survey question
above its respective cross-tab tables for final report presentation.  If
you will be using a graphics based word processing software program, a
monospace font, like Courier, should be used to maintain column
alignment of the numerical totals with their text headings.


Resetting Survware

By selecting "Reset the Software" from the menu, you will be able to
delete all report files, the survey database file and your responses to
the initialization process.  The program will offer you one chance to
change your mind before resetting the software.  The next time you start
Survware you will have to complete the initialization process.


Quitting Survware

The safest way to quit Survware is by selecting "Exit the Program" from
the menu.  This method will ensure all files are closed, and any
housekeeping which is needed will be taken care of automatically.



CHAPTER 2  -  TIPS AND TECHNIQUES


Survware Files

All files created by Survware are saved as plain text (ASCII) documents.
The program is contained in one application file which was created using
Microsoft QuickBASIC.

After initialization of Survware, at least three files are always
created.

-  A file which contains the initialization responses is named INSTALL.

-  A file which will hold the responses you enter from survey
questionnaires is named SURVEY.LST.

-  A file which will hold a demographic group report is named GROUP1.  If
your questionnaire has more than one demographic group question, then
there could be up to 10 such report files named GROUP1, GROUP2, etc.

If you were to open the INSTALL file with any word processor software,
you would see a list of numbers and some words.

If you had just completed the Survware initialization and you had
indicated during the process that your survey questionnaire looked like
this (a purely hypothetical situation):

- Four questions on the questionnaire;
- Two of which were demographic grouping questions;
- The demographic grouping questions were #3 and #4;
- All survey questions had four response choices each;
- The first demographic grouping question's choices were MAN, WOMAN,
BOY, GIRL; and
- The second demographic grouping question's choices were 12-17, 18-30,
31-40, OLDER;

your INSTALL file would look like this:

4,2,0 *
3       (first demographic group question number)
4       (second demographic group question number)
4       (number of response choices for question #1)
4       (number of response choices for question #2)
4       (number of response choices for question #3)
4       (number of response choices for question #4)
"MAN"     (response choice A for question #3)
"WOMAN"   (response choice B for question #3)
"BOY"     (response choice C for question #3)
"GIRL"    (response choice D for question #3)
"12-17"   (response choice A for question #4)
"18-30"   (response choice B for question #4)
"31-40"   (response choice C for question #4)
"OLDER"   (response choice D for question #4)

*  The first line of the file contains the number of questions in the
questionnaire, the number of demographic group questions and the number
of questionnaires which have been entered completely into the SURVEY.LST
file.

Unless you really know what you are doing, it is best to not tinker with
this file.  The consequences of making a mistake here would be jumbled
and incorrect analysis reports.  The only time any editing would be
merited would be if you had already entered a substantial number of
questionnaires and you discovered that a question needed to be allowed
more responses (up to 10, except for demographic group questions, up to
nine).  Even then, caution should be used and you should make a back-up
copy of INSTALL before doing any editing on it.

If you were to open the SURVEY.LST file with any word processor
software, you would see one long string of lowercase letters each inside
a set of quotation marks.  Using the above INSTALL file as an example,
if two respondents had completed their questionnaires and the first one
had answered every question by selecting the first choice (a), and the
second respondent had answered every question by selecting the second
choice (b), the SURVEY.LST file would look like this:

"a"   (first respondent's choice to question #1)
"a"   (first respondent's choice to question #2)
"a"   (first respondent's choice to question #3)
"a"   (first respondent's choice to question #4)
"b"   (second respondent's choice to question #1)
"b"   (second respondent's choice to question #2)
"b"   (second respondent's choice to question #3)
"b"   (second respondent's choice to question #4)

You can see that opinion survey data can easily be imported into
Survware.  You would have to ensure that it had been formatted or was
transferred to Survware in the same style as above with quotation marks
around each response, and that each response ended with a carriage
return and a line feed.  All responses would have to be capital letters
which fell within the range of possible answer responses as indicated in
the INSTALL file.

If you do attempt to import data into Survware, ensure that the number
of lines in the imported file is an exact multiple of the number of
questions in your questionnaire, e.g. if the questionnaire has 20
questions, the imported file would have to have 20, or 40, or 60, or 80,
etc. lines.  Otherwise, Survware will not be able to create correct
analysis reports.


Data Entry Tip

Entering survey data from questionnaires is rather boring and time
consuming when done alone.  Survware users have said that entering data
with the assistance of someone who reads the responses out loud helps
speed up the process.  The reader rhythmically calls out the responses
in groups of four or five letters using any form of a phonetic alphabet
to aid in letter recognition by the typist.  For example, Alpha could be
used for A, Bravo for B, Charlie for C, etc.  Survware users have
reported that a set of 50 questions can be entered in under 45 seconds
when reading response letters in groups of four.


Report Formatting Tip

Reports can be made more readable.  Many of the columns may contain
quite a few zeros.  Use the "Search and Replace" capability of most word
processor software to find all the "0" (zero) values in the absolute
frequency tables which have a blank space on each side, and then replace
each one with three blank spaces.  This technique precludes replacing
the zero at the end of a number like 10 or in the middle of a number
like 104 with a blank.  This same technique can be used on the relative
frequency tables to find all the "0.0%" values which have a blank space
on each side, and then replace each one with six blank spaces.  Again,
if you are using a graphics based word processor program, remember to
use a monospace font to maintain alignment of the column totals.



CHAPTER 3  -  OPINION SURVEY TUTORIAL


INTRODUCTION

Opinion surveys generally are a mystery to the uninitiated and a pain in
the neck for harried public relation specialists who have to conduct and
decipher them for the boss or their clients.  This tutorial may not be
the aspirin that relieves all survey headaches, but hopefully it will
help you with the basics of opinion surveying.

Why do we conduct surveys?  Mostly to find out what people think about
certain things that affect their lives.  Hence the name opinion surveys.

In the information field, you can use surveys to measure readership of
publications; audience size for broadcast programming; and the
effectiveness of public relations, educational and community action
programs.


MAIN ELEMENTS OF A SURVEY

Let's start with the two main elements of a survey: the problem and the
population.  The problem is the big question or questions you want
answered.  Population or universe is the group from which you select the
people to answer your questions.

For example purposes let's say that you want to find out how many people
read your community newspaper, the NEWS.  You are convinced that this
journalistic marvel is widely read.  But the publisher suspects that it
only provides expensive canary carpeting.  You are to conduct a sample
survey of community members to settle the issue.

Never having conducted one of these operations you seek a guide to
sample surveying which outlines these basic steps:

*  State the problem in 20 words or less.
*  Specify the survey population.
*  Choose a sampling method.
*  Decide how big a sample is needed.
*  Choose the sample subjects.
*  Prepare a draft of the questionnaire.
*  Train your interviewers.
*  Pretest the questions and interview procedure.
*  Gather the data.
*  Tabulate, audit, analyze and interpret the data.
*  Report the results.

Some of the above steps are self-explanatory.  Others require more space
than available to cover them.  The references listed at the end of the
manual cover those steps thoroughly.  Instead let's focus on those most
perplexing and critical areas of sampling method and sample size.

There are 4,000 people who live in the community serviced by the NEWS.
You obviously cannot ask everyone if they read the NEWS.  Sampling
allows you to save time and money while still getting reliable answers.


SAMPLING METHODS

There are several sampling methods available to you.  Simple random,
stratified random, cluster, systematic and quota sampling.  The first is
good, the second is best, the third and fourth are convenient, and the
fifth is usable only as a last resort.

Simple random sampling means that everyone in the population has an
equal chance of being included in the sample.  To select a simple random
sample of 350 people from a group of 4,000, you would use a table of
random numbers.  From this table you choose 350 different numbers
between 0001 and 4000.  You number the list of 4,000 people, and pull
out those with the selected 350 numbers for the sample.

Stratified random sampling is a refinement of simple random sampling.
Before a sample is chosen, the population is first divided into
categories or strata, e.g. age, sex, income, education.  Stratifying
ensures getting enough sample subjects from within smaller groups within
each category to make separate conclusions about each category.  If 10
percent (400 of the 4,000) of the community's population were college
graduates, you would need to include as part of your 350-person
community sample 35 college graduates (10 percent of 350).  You would
select 35 different numbers between 001 and 400 from the table of random
numbers.  Then you would go to a list of all the community's college
graduates, numbered serially, to select the 35 people.  You would do the
same thing within each smaller education category you established for
the sample within the community, e.g. high school graduate, some
college, etc.

Clustered sampling means that you pick your 350 community members in
groups.  For example, you might select 35 people by the random method
from each of 10 apartment buildings randomly selected from among the 40
which exist in the community.

Systematic sampling is sometimes called patterned sampling, because you
follow a definite pattern in choosing sample subjects from some
population list of names.  To systematically select your 350 community
members, you would take every eleventh name from a community roster or
similar listing.

Quota sampling is the fastest, cheapest and least reliable method.  To
obtain your sample, you would tell your 10 interviewers to each get
responses from 35 people.  The only restrictions might be that the
people live in the community and that at least 10 teen-agers be included
in the 35.  Who gets selected for the sample is then left up to the
interviewers.


SAMPLE SIZE

In determining sample size, you are concerned with two things: accuracy
(reliability) and confidence (risk).  Other factors such as method of
collecting data, interviewer bias and question structure can also
influence the reliability of your results.

When you take a random sample, you are using a carefully selected part
to show the characteristics of the whole population.  There will always
be some error, i.e. difference between the estimate provided by the
sample and the percent of the whole population who reads the NEWS.
However, random sampling allows you to specify ahead of time the amount
of this error and the confidence you have that that amount will not be
exceeded.

In sample surveys conducted by the country's most renowned pollsters,
the most commonly used error factor, or accuracy tolerance, is plus or
minus 5 percent with 95 percent confidence limits.  This means that if
you conducted the sample survey 100 times, in 95 of them the results
will not vary from the way the entire population would have responded by
more than plus or minus 5 percent.

For example, if 50 percent of the community sample said they read the
NEWS, that means the actual readership may be as many as 55 percent or
as few as 45 percent of the 4,000 community members.

The table below shows the size of samples you would need to achieve
different levels of accuracy with 95 percent confidence. It is easy to
see that when you increase the accuracy required, the size of the sample
quickly mushrooms.


                Sample Size for Reliability of Plus or Minus:
Population Size          1%      2%      3%      4%      5%
          1,000                                 375     278
          2,000                         696     462     322
          3,000                1334     787     500     341
          4,000                1500     842     522     350
          5,000                1622     879     536     357
         10,000        4899    1936     964     566     370
         20,000        6489    2144    1013     583     377
         50,000        8057    2291    1045     593     381
        100,000        8763    2345    1056     597     383
        500,000        9423    2390    1065     600     384
     and higher


The sample sizes indicated are for usable questionnaires, ones that can
be used to analyze the survey.

For a population of 4,000 and an accuracy tolerance of plus or minus 5
percent with 95 percent confidence, you need to randomly select and
interview 350 people.

Using the simple random sampling method, you selected, and got usable
questionnaires from, 350 people.  The survey's analysis showed that 80
percent read the NEWS.

In presenting the survey results on the NEWS to the publisher, you must
tell the publisher to what degree the results can be trusted.  You can
say you are confident that in 95 of 100 such similar surveys the true
NEWS readership is between 75 and 85 percent.


QUESTIONNAIRE CONSTRUCTION

Putting together a survey questionnaire is like working a jigsaw puzzle;
once you get started right, the pieces fall in place.  Let's see if we
can lay out the steps in getting one all together.

Questionnaires find many uses in the information field.  Mostly they are
used to gather facts or opinions from a survey group.  Sometimes they
are used to pretest the effectiveness of films, booklets, and other
educational materials.  Here we will focus on how questionnaires are
used for surveys.

There are basically two kinds of questionnaires; sometimes called survey
schedules.  For self-administered ones, the people fill out the
questionnaire themselves.  On personal interview ones, someone else
records the answers.  The main difference is that you must include
directions for the respondents with the self-administered kind.

The cardinal rule is to make it easy for everyone involved in the survey
while still achieving your objectives.  You must think ahead.  Who is
going to answer the questions?  How will the responses be recorded and
later coded?  What do you really want to find out?  What kinds of
statistical analysis are planned?  Then use this information to
construct your questionnaire and formulate the questions.

Most questionnaires take some general form and have some common
elements.  Usually they are printed on 8.5" by 11" paper using offset,
photocopy or some other form of legible reproduction.  Sometimes
supplementary materials such as 5" x 7" answer choice cards and machine
scorable answer forms are used in conjunction with the questionnaire.
Different colored paper is useful in keeping subject groups, before and
after responses, or variations in the questionnaire separate.  Light
colors and pastel shades are best.

It is a good idea to print about one-third more questionnaires than the
number you need completed.  These will be used for training
interviewers, providing sample copies, spoilage and lost copies.


QUESTIONNAIRE PARTS

There are three main parts of the questionnaire jigsaw puzzle which can
be broken down into smaller pieces.  Generally, three kinds of
information are sought: identifying information at the beginning, the
questions about the survey subject in the middle, and the demographic
and socio-economic characteristics of the respondent at the end.

The identifying information generally includes the name or title of the
survey such as, "Readership Survey for the NEWS."  It allows the
respondent and others to quickly identify what the survey is about.
Sometimes this is followed by a statement of the general purpose of the
survey and the confidential nature of the responses.

You may also find the name and address of the sponsoring agency in this
section or on a cover letter if there is one.  This is a big help if
someone later wants to refer to the researchers for the results.
Usually, there is the date the interview is completed, and a
questionnaire or sample number.  The date is necessary to determine if
there was any variation in the data due to changes in events during the
data collection period that might have influenced the responses.  The
questionnaire number is useful in sorting questionnaires, sequencing the
collected data and distinguishing data from different areas.

Developing the questions is the most important and difficult part of the
survey schedule.  Eliciting straight factual information is much easier
than getting a true measure of opinions.

You have several choices as to the kinds of questions for finding out
how much someone knows about a topic, or for census-type information.
Each has certain advantages and disadvantages.


QUESTION TYPES

Open-ended questions allow the respondents to answer the question in
their own words.  While allowing the respondents maximum opportunity to
express their viewpoint, you are faced with the problem of categorizing
the responses from many people.  This takes time, introduces greater
possibility for coder error, and sometimes people do not remember or
know how to express correctly certain facts.

Checklist or multiple choice questions present the respondents with a
selection of possible answers.  Their task is then to choose one or more
depending on the question.  The biggest advantage here it that the
respondents can be reminded of possible answers that may be otherwise
overlooked.  In most cases it is wise to have an "other" category not
given in the list.  This is particularly important when pretesting a
list of possible answer choices to determine if any have been left out.
A large percentage of "other" responses would indicate this.

Dichotomous questions are usually of the "yes or no" and "true or false"
types.  You should use a "not sure" or "don't know" choice also so that
the respondents are not forced into a decision if they are really
uncertain.  Many times this kind of question is used as a qualifying
question.  If the respondents answer "correctly," they are then asked to
answer the remaining questions about the subject.  For example, in a
readership survey you wouldn't want non-readers to express their
opinions of the publication's content.

Sometimes it is useful for respondents to rank items to indicate some
sort of preference or perceived importance for the items.  The
respondents are given a list of items and asked to assign a rank (1, 2,
etc.) to them.

Measuring a person's opinion can be accomplished by using different
kinds of questions.  The agree-disagree approach asks the respondents to
indicate whether they agree, disagree or have no opinion with regard to
the statement.  One should try to balance the number of statements with
an equal number of expected answers in the agree-disagree categories.

In addition, various verbal or numbered scales can be used.  These are
usually three-, five-, or seven-point scales which can be used to
measure the intensity of feeling about the subject.  For example, you
could ask respondents to indicate how important they feel various
sources of news are for them.  Each source would be ranked from very
important to not important, such as:

     As a source of news, I consider radio:
     A.  Very important
     B.  Somewhat important
     C.  So-so
     D.  Somewhat unimportant
     E.  Not important

For analysis purposes, the top and bottom two categories might be
combined.  Sometimes they are condensed into a three-point scale, such
as important, so-so, and not important for the question.

Usually you will find it necessary to use several different kinds of
questions to get the information you need.  However, there are some
basic things to remember about the wording of all questions.


QUESTION WORDING

Always make the questions as concise as possible--the shorter, the
better.  Simple words familiar to the respondents will help achieve
this.  Any technical words or specialized jargon should either be
substituted for, or defined.

The question should yield exactly the information desired, and in terms
that lend themselves to tabulation and statistical analysis.  A common
error often made is that of overlapping categories which makes
interpretation impossible.  A prime candidate is when the respondents
are asked to check their age category and are then confronted with
a set of choices like this:

     A.  18-20
     B.  20-22
     C.  22-24

Which group would you choose if you were 22 years old?  And how would
you know how many in the 20-22 group actually belonged in the 22-24
group?

Another common problem is that of double-barrelling or multiple-meaning
questions that are confusing to the respondents and defy any clear
interpretation of the responses.  "Which medium is best for news and
entertainment?" is typical.  Limit each question to one and only one
point.

Leading questions are ones that are worded in such a manner as to
suggest or lead to the answers.  "Would you say you are in favor of
continued federal subsidies for Medicare?" is leading.  "How do you feel
about continued federal subsidies for Medicare?" is not.

Catchwords, stereotypes or words with emotional connotations should be
avoided.  These words--such as, radical, Commie, pinko--elicit strong
feelings that influence how a person will respond.   "Should naturalized
Commie citizens be allowed to vote in the United States?" might draw a
predictably heavy negative response.

Be wary of question wording that might be interpreted as reflecting on
the prestige or honor of the respondent.  People will tend to answer in
terms of what they "ought" to think about themselves or a situation.
Questions about morality, religion, race relations and other socially
sensitive subjects are susceptible to this kind of problem.  "Do you
favor abortions?" would not be a particularly well phrased question to
ask Catholics or Pro-Life activists.  A similar problem exists when
questions call out responses toward socially accepted norms or values,
such as patriotism and motherhood.

When feasible it is sometimes good to use check questions; that is, two
questions that bring out the same fact, but are worded differently and
placed in different parts of the questionnaire.  These are used to check
the internal consistency of the responses.  "When were you born?" and
"How old are you?" serve this purpose.


QUESTION ORDER

How you order your questions is also very important.  The main idea is
to arrange them in a logical manner that will avoid confusion and
misunderstanding.  Keep the knowledge and opinion questions separate,
but try to keep questions dealing with similar subjects together.

The opening questions should be easily answered.  They should be
interesting and make the respondents want to cooperate.  In no way
should they indicate that the questionnaire or interviewer is leading up
to a sale of anything or has a special point to prove.

Personal questions about the respondents should be at the middle or end
of the questionnaire.  Sensitive questions should be couched with less
probing questions.  Information about age and income are sometimes
considered very personal by respondents.  For this reason, the
demographic and socio-economic section which includes this kind of
information is kept for the end of the questionnaire.

The demographic and socio-economic section basically asks the
respondents to describe themselves.  Responses to these questions are
useful when cross-tabulated with other questions in describing the
different kinds of people who either know something, or hold a
particular point of view.  Most surveys include at least questions of
this type on age, sex, education, marital status and income.

It is usually a good idea to provide space at the end for the
respondents to give their personal feelings about the survey subject.
This is another way of letting them know that their information and
observations are valuable to you.  And, of course, it is always good to
thank the respondents, either in writing or verbally, for their
cooperation.


SURVEYING METHODS

There are two basic ways that questionnaires can be used in gathering
facts or opinions from a survey group.  We are all familiar with the
George Gallup and Louis Harris public opinion polls.  These polls use
questionnaires in a personal interview.  Here the interviewer asks the
questions and records the answers.

The other survey method involves the use of a self-administered
questionnaire in an impersonal situation.  The respondents fill out the
forms themselves, either in a group setting, work place, or in the
privacy of their homes.

There are many variations of each basic method, such as telephone
interviews, direct mailings, questionnaires passed out at grocery
stores, post cards and tear sheets inserted in magazines, and so on.
We'll focus mainly on one-on-one interviews, group settings and direct
mailing situations.

Which basic approach--the personal interview or the self-administered
questionnaire--suits your data collection purposes best?  Let's look at
the strengths of each in doing opinion surveys.  Keep in mind that the
questionnaire must be tailored for each approach.

The main differences are that you must include directions for the
respondents with the self-administered questionnaire, and you must
select and train interviewers for the personal interview.


SELF-ADMINISTERED QUESTIONNAIRES

Whether used in the group setting or sent out by mail, the self-
administered questionnaire is the cheapest method of surveying the
population.  This approach makes it easy to gather data from a large
number of respondents over a wide area with a minimum number of workers
in a very short time.  Group sessions or direct mailings avoid the
problem of hiring and training interviewers, and the relatively slow
response rate of personally interviewing large numbers of people.

You also ensure a relatively high degree of uniformity of response by
using a standardized approach in the group sessions.  It is possible to
script out the administration of such group sessions so that any
variance from one session to the next is minimized.  This is not
possible with the personal interview.  Not only do the personalities and
appearances of the interviewers differ (and thus influence interview
results), but all interviewers are bound to vary their technique from
interview to interview.

The self-administered questionnaire also gives the respondents plenty of
time to think about their answers.  Group sessions do set some limits on
available time, but these are usually quite generous to allow adequate
consideration of each question.  The respondent is not faced with the
pressure of a spontaneous response as in the "eye-to-eye" situation of a
personal interview.

Perhaps the greatest advantage of the self-administered questionnaire is
the confidence the respondents have in the anonymity of their answers.
This is particularly important where sensitive issues are dealt with in
the questionnaire.  People are basically honest and want to cooperate in
surveys they feel are useful.  But they will not give straightforward
answers if they know their names will appear in the results.


PERSONAL INTERVIEWS

In certain situations the personal interview is needed to convince
people to cooperate at all in giving information on sensitive matters.
Let's look at some of the other advantages of the personal interview
approach.

The personal interview takes much of the burden off the respondent and
places it on the skills of the interviewer in eliciting answers.  Far
more skill and preparation are required in conducting personal
interviews than in using self-administered questionnaires.  However,
this is a basic strength of the personal interview approach.

The question and answer choices can be explained in a manner so that
even the least educated respondents can understand what is being asked
of them.  This is particularly important where you need information from
people of all socio-economic strata, and it is unlikely that they are
all capable of completing a written questionnaire.  This fact alone
helps the personal interview method achieve a much better return and
cross-section of any population than mail surveys.  The typical mail
survey nets less than a 50 percent return, and those returning the
questionnaire are generally bunched by certain socio-economic
characteristics.

Flexibility is another strong feature of personal interviews.  It is
possible to correct "on-the-spot" misinterpretations of the question by
repeating or rephrasing it.  Although questions should be worded in the
vernacular of the survey population, there are always some people who
will not understand the wording or aim of some questions.  The self-
administered questionnaire leaves no room for clarification.  Also, the
interviewer has a chance to make sure that complete data is obtained
from each respondent, and that no questions are omitted.


PRETESTING

Whether you plan to use the personal interview or self-administered
questionnaire approach, your results will be only as good as your
questionnaire.  The pretest is a trial run of the questionnaire to see
how well people understand it and whether it gives you the information
you need.  It lets you rephrase questions, substitute easy words for
hard ones, include additional answer choices, and even add or delete
certain questions.

The pretest also gives you practice in administering the study on a
small scale.  If you are using personal interviews, you will find out
how long it takes to complete a typical interview.  The interviewers may
need more training to speed up the process or some assistance in
correctly recording the answers.

Members of the survey population should be interviewed in the pretest.
Every attempt should be made to get a cross section of this population
so that problems unique to any group are identified.  This procedure
also prepares the interviewers for the realities facing them during the
full field study.

The situation is similar for self-administered questionnaires in a group
setting.  Procedures for getting the group assembled and the
questionnaires answered are practiced.  For mail surveys, a small
portion of the survey population is sent the questionnaires.  Rate of
return, completeness of answers, turn around time and characteristics of
respondents are some things that can be recorded from trial mailings.

However, the main value of the pretest of any questionnaire derives from
discussing the questions and procedures with the respondents.  You can
ask them what the questions meant, what words they didn't understand,
how clear were the instructions, were the answer choices appropriate or
complete enough, and so on.  The idea is to blend the pretesting of your
questionnaire with the pretesting of the way in which you plan to use it
in the survey.


MAIL SURVEYS

Since Survware was designed primarily for the do-it-yourself survey
taker, this tutorial will only cover the self-administered type of
survey method in detail.  Generally, the self-administered survey is
cheaper, requires a smaller staff and is done through the mail.

A successful mail survey is one that gets the needed information from
enough of your sample to make the results reliable.  Paul L. Erdos put
it this way: "No mail survey can be considered reliable unless it has a
minimum of 50 percent response, or unless it demonstrates with some form
of verification that the nonrespondents are similar to the respondents."
The Advertising Research Foundation recommends 80 percent.  Let's see
what techniques you can use to reach these goals.

Several of the references cited at the end of this tutorial will guide
you through the steps of a typical mail survey.  But let us zero in on
the key characteristics of the outer envelope, the cover letter,
questionnaire, incentives, pretest and follow-up mailings that affect
your response rate.  We will assume that you have carefully drawn a
sample from a population that represents exactly the kind of people you
want to respond.

Perhaps the biggest obstacle to overcome is getting the potential
respondents to open the envelope containing the questionnaire.  Advance
notice postcards, letters and telephone calls will alert them to the
forthcoming questionnaire.  A postcard or call will usually suffice, but
a letter is more effective when the sample consists of VIPs, or the
message is too long for a postcard, or the survey is confidential.  The
impact of advance notice is increased by referring to it on the
questionnaire envelope.  Print "ENCLOSED IS THE SURVEY WE WROTE (CALLED)
YOU ABOUT" on the envelope in colored ink in the lower left hand corner.

Even without advance notice, your survey envelope will have a better
chance of being opened if you follow these suggestions.  If at all
possible, address it to the individual using his or her full name.  Be
sure to include a title like Mr., Mrs., Ms., or Dr. where appropriate.
The name and address should be individually typed on all envelopes.
Even though this is a lot of work, it pays important dividends in
increasing returns.

The name of your organization and return address should be clearly
printed on the front.  Use first class mail and postage stamps, even if
your organization has franking privileges, metered stamps, or a business
permit.  Use the business reply or franked envelopes for returning the
questionnaire.  Since you will be paying postage, keep an eye on the
combined weight of your questionnaire and envelope.  A slight change in
the size of envelope or questionnaire can greatly increase mailing
costs.  The whole idea behind these suggestions is to help the
respondent distinguish your mail from junk mail or direct advertising.

Once the respondents have opened the envelope, you must arouse their
interest or convince them that it is worth their time to fill out the
questionnaire.  Here is where your cover letter comes into play.  The
cover letter must create a feeling of personal communication between
researcher and reader.

The text of the letter should be reproduced on official stationery that
clearly identifies who is doing the research, along with the
researcher's address and telephone number.  This information makes it
clear to respondents that it is a legitimate research effort, and that
they can get further information about the survey if they want it.

The main text of the letter should be written in a personal vein by
making liberal use of the pronoun "you" in brief, to-the-point
sentences.  While the precise wording and approach used must be tailored
to the study's objectives and the nature of the respondents, a typical
cover letter contains several items.

Start with a general statement of the study's purpose.  This provides a
frame of reference for what is to come later in the letter.  This
statement should be brief and interesting, but should not bias responses
to the questions.  Naturally, people want to know how and why they were
selected for the study.  Here you can indicate the main criteria for
selecting particular individuals from a given population.  This ties in
with two more elements of the cover letter, namely the importance of the
respondent's cooperation and the assurance of complete confidentiality.

People must feel it is worth their time to fill out the questionnaire.
This can be accomplished by pointing out the value of their replies both
from an individual worth and from sampling points of view.  Reassurance
of confidentiality is importance since most questionnaires are keyed for
follow-up mailing and other purposes.  Use of key numbers or other
identification does not permit you to ethically state that they can be
assured of anonymity, although you obviously have no plans to associate
names with responses.

Include some reference to how much time is involved.  However, when you
state, "It should take only a few minutes to answer the questions," be
sure it doesn't take longer even for the slowest respondent.  Such a
statement, coupled with a long, difficult appearing questionnaire, will
almost certainly result in a nonresponse.  Honestly is the best policy.
A promise to send a copy of the final report often will get people to
invest a greater amount of time, since they are personally interested in
the results of the survey.

Mention the enclosed, postage-free return envelope.  This reminds the
respondents of how easy and cost free returning the questionnaire is.
Often people feel obliged to complete the questionnaire, or even return
an uncompleted one if you provide return postage.

A simple "thank you" is always an appropriate way to conclude the cover
letter.  It shows appreciation for people taking time to read the
letter, and usually, completing the questionnaire.

With mail surveys, it is doubly important that all questions are
necessary, easily understood, take up a small amount of space, and can
be answered with a minimum of instructions.  Here are some pointers that
will help create the "short and easy" look.

As a rule, the rate of return increases as the size of the questionnaire
decreases.  A questionnaire of monarch size (7" by 10") will draw more
returns than one of letterhead size (8.5" by 11").  Likewise, the fewer
the pages, the higher the percentage of return.  Four pages, front and
back, is generally the upper limit to consider, but it is better not to
exceed two pages.  Always establish the questions you want to ask,
rather than determining how many pages you want to fill.

Use good paper stock, but avoid using colored paper unless you want to
separate respondent groups.  Then use only pastel shades which do not
hinder readability.  The printing must result in clear reproduction
which makes the questionnaire easy to read.  Also, use white space and
other layout techniques to help readability.

Questions can be made to appear easy to answer.  Use mostly multiple-
choice or close-ended types of questions.  Also, always provide a "no
answer" or "not applicable" choice.  Open-ended types of questions make
the respondent work, spend more time, and often decide not to complete
the questionnaire.  Most questions should not require any specific
instructions, thus saving valuable space.

In most cases, you will want to consider follow-up mailings to
nonrespondents.  Normally, surveys can be closed out after three or four
weeks.  Your first follow-up mailing should go out 10 to 14 days after
the initial mailing.  Additional mailings are usually not worth the time
or money.  Questionnaires mailed in each wave should have a key number
which matches a number and associated respondent on a master list or
card file.  A daily accounting of returns shows where you stand on total
returns, and who has not cooperated.  Although invisible ink and other
clever means are sometimes used to key questionnaires, it is generally
advisable to put a clearly visible number or mark in some not too
inconspicuous place on the questionnaire.  The reason most people do not
return the initial questionnaire is they simply mislaid it or forgot to
answer it.

The pretest is one way of ironing out many problems before doing your
full scale mailing.  By selecting a random sample of your survey
population for the pretest, you can check the rate of returns, test the
quality of your mailing list, review question wording and usefulness of
the information, and evaluate the effectiveness of different techniques
(e.g. incentives, advance letters and calls) on the rate or return.


REFERENCES

1.   Herbert Arkin and Raymond Colton, TABLES FOR
     STATISTICIANS, Barnes and Noble, Inc. New York, 1967.
2.   Charles H. Backstrom and Gerald Hursch, SURVEY
     RESEARCH, Northwestern University Press, 1963.
3.   Allen L. Edwards, TECHNIQUES FOR ATTITUDE SCALE
     CONSTRUCTION, Appleton-Croft, Inc. New York, 1957.
4.   Paul A. Erods, PROFESSIONAL MAIL SURVEYS, McGraw-
     Hill Book Co., New York, 1970.
5.   George Gallup, THE SOPHISTICATED POLL WATCHER'S
     GUIDE, Princeton Opinion Press, 1972.
6.   Dr. Eugene A. Kroupa, AUDIENCE SURVEYS, Military
     Media Review, Indianapolis, 1980.
7.   A. N. Oppenheim, QUESTIONNAIRE DESIGN AND
     ATTITUDE MEASUREMENT, Basic Books, Inc. New York, 1966.
8.   Mildred Parten, SURVEYS, POLLS, AND SAMPLES, Harper
     and Brothers, New York, 1950.
9.   Claire Seltiz et al., RESEARCH METHODS IN SOCIAL
     RELATIONS (revised edition), Holt Rinehart and Winston,
     New York, 1964.
10.  Slayton C. Shaw, COMPUTERIZED SURVEYS, Military
     Media Review, Indianapolis, 1988.
11.  Slayton C. Shaw, DATA TABLES, Que Publishing, Inc.,
     Indianapolis, 1986.
12.  Sidney Siegel, NON-PARAMETRIC STATISTICS FOR THE
     BEHAVIORAL SCIENCES, McGraw-Hill, New York, 1956.
13.  Morris J. Slonim, SAMPLING, Simon and Schuster, New
     York, 1967.


LIMITED WARRANTY.  CompStat Software does not warrant that the functions
contained in the software will meet your requirements or that the
operation of the software will be uninterrupted or error free.  The
warranty does not cover any media or documentation which has been
subjected to damage or abuse by you.  The software warranty does not
cover any copy of the software which has been altered or changed in any
way by you or others.  CompStat Software is not responsible for problems
caused by changes in the operating characteristics of the computer
hardware or operating system which are made after the delivery of the
software.

ALL IMPLIED WARRANTIES ON THE MEDIA AND MANUAL, INCLUDING IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, ARE
LIMITED IN DURATION TO NINETY (90) DAYS FROM THE DATE OF THE ORIGINAL
RETAIL PURCHASE OF THIS PRODUCT.

Even though CompStat Software has tested the software and reviewed the
documentation, COMPSTAT MAKES NO WARRANTY OR REPRESENTATION, EITHER
EXPRESS OR IMPLIED, WITH RESPECT TO SOFTWARE, ITS QUALITY, PERFORMANCE,
MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.  AS A RESULT, THIS
SOFTWARE IS SOLD "AS IS," AND YOU THE PURCHASER ARE ASSUMING THE ENTIRE
RISK AS TO ITS QUALITY AND PERFORMANCE.

In no event will CompStat Software BE LIABLE FOR DIRECT, INDIRECT,
SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES RESULTING FROM ANY DEFECT
IN THE SOFTWARE OR ITS DOCUMENTATION, EVEN IF ADVISED OF THE POSSIBILITY
OF SUCH DAMAGES.

The warranties AND REMEDIES set forth above are exclusive and in lieu of
all others, oral or written, express or implied.  No CompStat Software
dealer, agent, or employee is authorized to make any modification,
extension, or addition to this warranty.

Some states do not allow the exclusion or limitation of implied
warranties or liability for incidental or consequential damages, so the
above limitation or exclusion may not apply to you.  This warranty gives
you specific legal rights, and you may also have other rights which vary
from state to state.



ORDER FORM

CompStat Software                     CompuServe 72301,2564
P.O. Box 16367                        (317) 898-4186 (office)
Indianapolis, IN  46216-0367

*  Prices are subject to change without notice.
*  Price includes technical support for 90 days.
*  Price includes reduced prices on upgrades.

***********************************************************
* ALL EDUCATIONAL ORGANIZATIONS, THEIR STAFF AND FACULTY, *
*   AND STUDENTS RECEIVE A 10% DISCOUNT ON BELOW PRICES   *
***********************************************************
                                 QUANTITY   PRICE    PRICE
    ITEM                    QTY  DISCOUNT   EACH    EXTENDED
    =====================   ===  ========   =====   =========
    Survware
                  1 -  4    ___     0%    $ 50.00*  $_______
                  5 -  9    ___    20%      40.00    _______
                 10 - 49    ___    25%      37.50    _______
                 50 - 99    ___    30%      35.00    _______
                    100+    ___    40%      30.00    _______

        MINUS ADDITIONAL 10% EDUCATIONAL DISCOUNT   $_______
                                 (Proof required)

                            SUBTOTAL:               $_______

     (Indiana state residents only) 5% sales tax    $_______
     Standard shipping and handling FREE in U.S.
     Orders outside U.S. and Canada  -  add $5      $_______

                                            TOTAL   $_______

* If the Survware Evaluation Program was purchased from CompStat,
include proof of purchase for a $10 reduction on the first disk.


Name    _______________________________________________

Address _______________________________________________

        _______________________________________________

City    _______________________ State_____ Zip_________

Country ____________________ Phone (____)_____-________

         ----------------end-of-author's-documentation---------------

                         Software Library Information:

                    This disk copy provided as a service of

                           Public (software) Library

         We are not the authors of this program, nor are we associated
         with the author in any way other than as a distributor of the
         program in accordance with the author's terms of distribution.

         Please direct shareware payments and specific questions about
         this program to the author of the program, whose name appears
         elsewhere in  this documentation. If you have trouble getting
         in touch with the author,  we will do whatever we can to help
         you with your questions. All programs have been tested and do
         run.  To report problems,  please use the form that is in the
         file PROBLEM.DOC on many of our disks or in other written for-
         mat with screen printouts, if possible.  PsL cannot debug pro-
         programs over the telephone, though we can answer questions.

         Disks in the PsL are updated  monthly,  so if you did not get
         this disk directly from the PsL, you should be aware that the
         files in this set may no longer be the current versions. Also,
         if you got this disk from another vendor and are having prob-
         lems,  be aware that  some files may have become corrupted or
         lost by that vendor. Get a current, working disk from PsL.

         For a copy of the latest monthly software library newsletter
         and a list of the 2,000+ disks in the library, call or write

                           Public (software) Library
                               P.O.Box 35705 - F
                            Houston, TX 77235-5705

                                1-800-2424-PSL
                                 MC/Visa/AmEx

                          Outside of U.S. or in Texas
                          or for general information,
                              Call 1-713-524-6394

                          PsL also has an outstanding
                          catalog for the Macintosh.

