Designing Data Collection Forms

1
Designing Data Collection Forms

• Dr. John T. Drea
• Professor of Marketing
• Western Illinois University

2
Six Things a Questionnaire Does

• It translates the research objective(s) into
  specific questions.
• It standardizes how questions are asked.
• It can keep a subject motivated to complete the
  research.
• It is a permanent record of the research.
• It facilitates data analysis (clearer responses,
  faster)
• It allows for the calculation of reliability and
  validity.

3
Good Questionnaires Are No Accident!

• The commandments of question wording
• Each question must focus on one topic.
• Be brief.
• Be sure everyone interprets each question the
  same way.
• Question wording must be clear to subjects (not
  just to you).
• Be simple.
• Subjects should answer for themselves.
• Subjects must be capable of answering
• Dont ask subjects to guess.
• Dont lead the respondent.
• Avoid ambiguous wording.

4
Some problem words - avoid absolutes
Always, every, best, never, nobody, none
Absolutes will get you in trouble. For example,
do you always like your Mom?
5
Some problem words - words open to interpretation
American values, where, government, country, see,
today, you, dinner, service
Where did you see our ad? In my chair.
How often do you go to the grocery store? Never.
My wife does.
Have you seen our Ad? No. (but Ive heard it on
the radio)
6
Some problem words - confused meaning
Could/should/would, may/can
Theres quite a difference between
could/should/would lower prices get you to come
back?
Remember, can asks if its possible. May
asks for permission. For example Can we
contact you for more information? is wrong.
7
Some problem words - words needing a basis for
comparison
Would you like more service? Well, more than
what? Would you like just as much service as
you used to receive? Do you mean only as
much? Quite is often misused in the same way
the unique does. Quite little is an
oxymoron, and very unique doesnt make sense.
More, less, just, quite
8
Other problems to watch

• Response categories
• Overlapping categories
• Unbalanced categories (where balanced would make
  more sense)
• Balanced categories (where unbalanced would make
  more sense)

9
Everyone makes mistakes

• Heres my two worst mistakes in question design.

How long have you lived at your present
address? ___ lt1 year ___1-2 yrs. ___2.1-4
yrs ___4.1-6 yrs. ___ 6.1-8 yrs ___8.1-10
yrs. ___Over 10 years
How important are each of the following to you in
choosing how to travel to/from Chicago? Circle
the appropriate number. Important
Unimportant Cost 5 4 3 2 1 Comfort 5 4 3 2 1 Con
venience 5 4 3 2 1
10
Start at the beginningThe introduction!

• Who is doing the research? (unless a disguised
  approach is used.)
• What is the purpose of the survey? Tell them why
  you want to know their responses.
• Why was the subject selected?
• Get the survey to the right respondent!
• You need to persuade the subject to complete the
  survey.

11
Sequencing of questions

• The goal here is to get the right person to
  complete the survey in the right way!
• Double-check Is the question needed?
• Put screening questions at the beginning.
• Throw them a bone - place some easy questions
  near the start.
• Harder questions are usually placed in the third
  quarter (and later) of the survey.
• Put personal questions at the end (such as
  income, age.)
• You can organize by starting with the general and
  getting specific, or grouping questions into
  sections - depends on whats being measured.

12
Observational Studies

• Why do we observe consumers in the act of
  purchasing?
• Consider the roles of cognition and social
  desirability in survey research.
• Do you want to know why or what?
• Identification of situational influences.
• Mystery shopper programs

13
Observational Studies - Build-up or Break-down?

• Build-up do the observations first, then create
  the categories in which to put the observations.
• Avoids a priori beliefs, forcing categories
• Break-down design the observation form first
  with categories, then place the observations into
  these categories.
• Typically more consistent
• Focuses researcher on certain behaviors
• Complete

14
Case 11.3 Metro Research, Inc.

• Do you recall a specific advertisement or other
  promotional message of Seville Restaurant in the
  last five years?

___No ___Yes, where did you see the ad?

• Problems
• May be too difficult to start the questionnaire.
• Asks for specifics when only generalities can be
  provided.
• Where is a problem word.

15
Case 11.3 Metro Research, Inc.

• Please review the items below and then provide us
  with a seriatim ranging from 1-12 or 13 in terms
  of the importance to you in patronizing Sevilles
  Restaurant.

___Its a fun place ___I came to be with
friends ___I came to meet people ___I came
because I wanted a good place to eat ___Its a
good tavern ___I like the atmosphere ___I like
the entertainment ___I like to dance ___Seville
is a famous place ___Drinks are affordable ___To
play darts, pool, games ___To watch the big
screen TV ___Other? Tell us_______________________
___________________________
Problems Seriatim? (It means rank) Categories
overlap Too many items to rank (can you
distinguish between whats 8 and 9, for
example)
16
Case 11.3 Metro Research, Inc.

• How often do you come to Seville Restaurant?

___more than once per week ___twice per
month ___about once per week ___once a
month ___12 times or less per year ___once per
year ___less than once per year ___This is my
first time
Problems Categories shouldnt overlap (whats
the difference between once per month and 12
times or less per year?)
Would this be better? How regularly do you come
to Seville Restaurant? ___Very Often
___Regularly ___Seldom ___Almost Never
___This is my first time
17
Case 11.3 Metro Research, Inc.
All questions were scaled very satisfied/somewhat
satisfied/neither satisfied nor
dissatisfied/somewhat unstatisfied/very
unsatsisfied/dont know

• Friendliness of employees
• Knowledge of employees
• Promptness of the service
• Parking
• Cleanliness of the facilities
• Quality and variety of entertainment
• Amount of the entertainment fee
• Price of the beverages
• Quality and price of the food
• Variety of menu choices

• Problems
• Should we include dont know?
• Double-barreled question
• Service is vague
• Needs an overall measure of satisfaction