A Survey Service for University and College Health Care Administrators
Setting up a survey
Surveys are collections of questions which have selectable answers.
The answers can be either a single answer, e.g. Yes OR No, or multiple
choice. Surveys are typically broken into two basic types of questions,
demographic questions and general questions. Demographic questions, e.g.
sex or age, are used to qualify the population and can be only
be a single answer. The general questions ask anything, can be
either multiple choice or single answer and are analyzed related to the
The first step in establishing a survey is to define all the questions
and associated answers
which are to be asked, both demographic (typically done first) and the
Once the questions have been determined for the survey, configuration
must be done. This includes the following:
Once this is done, the survey is then placed online and is available for
- Defining the basic survey structure
- Defining the questions
- Defining the page flow of questions
Defining the basic survey structure
entails establishing a definition file which contains information
- Names the survey
- Defines the criteria of participation, e.g. how are participants
identified and/or screened.
- Defines the "look and feel" of the survey, i.e. colors, backgrounds,
- Define passwords for access control of the results.
- Defines the locations for the question definitions, page definitions
and the results.
Defining the questions
The questions are defined by establishing a file which contain
all the questions involved in the survey. This is an ASCII text file which
has the following format:
|Q:Number-|| The question number.
The question text. This continues until there is a Y:record
|Y:Type ||Either R for OR, a radio button
which permits only ONE answer,
or C for AND, or checkbox list which permits more than one answer.
or M for AND with limited Multiple answers, a checkbox list which requires no more than a set number of answers.
Using the letter code in lowercase provides a 2-column display of the available answers.
|M:Number ||Limits number of responses where Type M (Y:M) is defined
|C:Category ||Either D, Demographic or G, General
|S:Single ||Single Line Switch. If set to 1, put all questions on single
|O:Optional||Optional Question Switch. If set to 1,
then this question can be left blank. If this switch is NOT set, then the
question must be answered or the survey cannot be completed.
|A:Answer ||The list of selectable answers. The sequence
of answers is the sequence they are displayed and indicates the answer number.
|R:Relation ||[NOT YET IMPLIMENTED] A-Q-M
The answer A is prohibited if question Q answer M is selected.
Question Q must preceed this question.
|E:Number- || The End of the data for Question Number:
As an example:
would define Question 1, asking "Sex:" as a Demographic question which could
be answered as either Male or Female.
Defining the page flow of questions is done by creating an ASCII
page description file as follows:
Examples would be:
|P:num-(Rel)|| Where num is the Page Number and, optionally, a (Rel)ationship.
The relationship defines a conditional page which is displayed only if
the answer to a question is some answer. The format is Question-Answer,
so that a line P:3-7-2 indicates that this would be page 3 if question
7 has answer 2. If relationships are used, there must be a page
defined for each possible answer for the associated question, i.e. if
page 3 is conditional on question 7, then there must be a page
definition for answer 1 and 2 and all other answers for question 7. Every possible use case needs to be accounted for.
If a given response causes the survey to jump ahead several questions end of Page 5, for example, a Thank You page at Page 5 must be defined, as well as where the survey would normally end.
|T:text|| The Page title
|H:text|| The heading for the page
|Q:nums|| The question numbers of the questions
to be asked on this page. The question numbers are separated by blanks.
|L:1||An indicator that this is the last page. It has no
questions and triggers the addition of the accumulated answers into the
|E:num-|| End of this page.
Q:10 11 12 13 14 17 18
T:Why No Insurance Coverage
H:Why No Insurance Coverage
Q:15 16 12 13 14 17 18
This would generate page 3 and ask questions 10 11 12 13 14 17 and 18
if questions 9 had answer 1 or
ask questions 15 16 12 13 14 17 and 18
if question 9 had answer 2.
Putting the survey online involves establishing the initial
page and establishing the answer analysis page. These are two web pages
which tie together the previously defined definitions.
Word Secure Encrypted Messaging now available to as a part of our services to aid administrators with anonymous logistics securely.
In addition, the system can be configured to:
- Screen a user's participation based on passwords or student identifiers.
- Allow a user to "retake" the survey.
- A respondent's anonymity can be established. In this case, the user
is asked for his/her identification to validate that they are allowed
to access the survey. If they are valid participants, a flag is set to
indicate that they have taken the survey and cannot do so again. The
system then assigns them a "random" identifier which is used to register
the answers. An anonymous survey cannot allow "retakes" of the survey.
- The participants can optionally be allowed to email comments, suggestions
or questions about the survey.
- The question/answer dialog can be done on a secure server using the Secure