2. How to write a health survey?
Mona Al-Dabbagh, MD, MHSc
Consultant Pediatric Infectious Diseases
King AbdulAziz Medical City
National Guard Health Affairs
3. Objectives
• To present a general outline to health surveys
• To demonstrate a systematic guide to writing
questionnaires and survey questions.
4. What is a survey?
• ‘A purposeful and systematic collection
of data from population by means of
asking questions’
5. Types of Health surveys
• Aim: Sampling
– Descriptive – Random population
– Analytical – Non-random population
• Topic: • Nature of information
– Health status – Subjective
– Health determinant – Factual
– Health services
• Data collection
• Study design: – Self-administered
– Cross-over – Interview
– Longitudinal – Mixed mode
6. Steps in designing a health survey
1. Formulating a survey objectives
2. Defining survey variables
3. Choosing the method for data collection
4. Formulating the questions
5. Formatting the questionnaire
6. Designing the method of sampling
7. Calculating the sample size
8. Pilot testing
8. Question type
• Concept measured
– Facts, opinions/attitude, feeling, knowledge, others.
• Measurement level
– Nominal, ordinal, interval, ratio
• Response format
– Open-ended, closed-ended, other
• Complexity
– Single questions, checklist, multi-item scale, other
• Sensitivity/ threat level
9. Factual questions
• How old are you?
___________ years
• What is your gender?
Male
Female
• How many children do you have?
1
2
3
4
More than 4 (specify)
10. Factual questions
• How many hours a day do you typically
study?
Less than 0.5 hours
1 hour to 1.5 hour
1.5 hour to 2 hours
2 hours to 2.5 hours
More than 2.5 hours
30% responded > 2.5 hours
Dillman et al, 2009
11. Factual questions
• How many hours a day do you typically
study?
Less than 2.5 hours
2.5 hour to 3 hours
3 hour to 3.5 hours
3.5 hours to 4 hours
4 hours to 4.5 hours
More than 4.5 hours
71% responded > 2.5 hours
Dillman et al, 2009
12. Opinions
1. Do you agree or disagree with the following
statement:
The research summer school course is very
informative.
Strongly agree
Agree
Neither agree nor disagree
Disagree
Strongly disagree
13. Opinions
1. In your opinion, how informative is the
research summer school course?
Not informative at all
Somewhat informative
Very informative
Most informative course ever
14. Opinions
1. Compared to other research courses that
you have attended, would you say that this
research summer school course is:
Very bad
Bad
Neither bad nor good
Good
Very good
15. Knowledge
1. Is the following statement true or false?
The incubation period for Influenza is 1 to 4
days.
True
False
2. What is the incubation period for influenza?
______________ days
16. Response format
• Open-ended
1. How old are you? ___________years
• Closed-ended
1. Are you a female or male?
Female
Male
• Visual analog
1. How much pain do you feel? Please mark on the line
No pain _____,____,____,____,____,____,____,_____ Very severe pain
• 1 2 3 4 5 6 7
17. Level measurement
• Nominal
1. Are you a female or male?
Female
Male
• Ordinal
1. How old are you?
<20
20-30
> 30
• Interval
1. What is your birth year? ________
• Ratio
1. What is your weight? ___________kgs
18. Single vs. Multiple response questions
• Single question
1. Have you ever watched the titanic movie?
Yes
No
• Checklist
1. Which sports do you play? Please check all that apply
Football
Swimming
Basketball
Tennis
Other (please specify)
19. Forced-choice checklist
1. Which sports do you play regularly, on monthly
basis?
Yes No
Football
Swimming
Basketball
Tennis
Other (please specify)_____________________
20. Matrix questions
• In the past month, how many times did you eat in
any of the following fast-food restaurants?
0 times 1-2 times 3-4 times > 4 times
1. McDonald
2. Burger King
3. Pizza Hut
4. Hardies
21. Threatening questions?
1. Did you ever join any protestation?
Yes
No
2. Have you ever seen a case of suspected child abuse
and did not report it to the child protection
services?
Yes
No
22. Random error vs. Bias
RANDOM ERROR (p):
• Deviation from truth in inference about underlying
population which happens because sampling was
part of study
SYSEMATIC ERROR (BIAS):
• Any systematic error in the design, conduct or
analysis of a study that results in a mistaken estimate
of an exposure’s effect on the risk of diseases.
23. Random error vs. Bias
Random error Bias
Ameliorated by repetition of the study Yes No
Ameliorated by large sample size Yes No
24. Random error vs. Bias
• In survey questions, bias (systematic error)
occurs when some factors cause errors in
responses to be more likely in one direction
than the other
• Random errors occur when there is no
systematic tendency for errors to be in either
direction
25. Health survey errors
• Coverage error
‘not everyone in the target population is
included in the sampling frame’
• Sampling error:
‘not everyone in the sampling frame included in
the samples (not a census)’
26. Cont. Health survey errors
• Non-response error
‘not everyone included in the sample actually
participated’
• Measurement error
‘not all responses are perfectly valid’
27. Issues to consider in questionnaires
• Questionnaire length
• Question order
o First question
o Logic order, grouping
o Demographic questions
o Attitude questions (general specific)
o Factual questions (context specific general ‘recall’)
• Instructions
o minimal, easy to remember
o Place specific instructions when needed, make distinct from
the questions
28. Issues to consider in questionnaires
• Questionnaire layout (self-administered)
o Cover page
o Number of pages
o Paper size
o Question grouping
o Matrices
o Double questions
• Visual elements (self-administered)
o Font
o Color/ shading
o contrast
o Spacing
o Response space
o Page breaks
o consistency
29. Ethical issues in surveys
• Ethics application
– Who is doing the study, funding, conflict of
interest, prior reviews
– Summary of the project
– Details of data collection procedures
– Risk and benefits to participants
– Informed consent procedure
– Confidentiality and data safety procedures
30. Cont. Ethical issues in surveys
• Informed consent/ cover letter
– Investigators/ institution
– Purpose of the study
– Tasks
– Duration
– Confidentiality
– Right to withdraw
– Risk and benefits
– Remuneration
– Contact
– Feedback
32. Do’s and don’ts
• Make it short
• Name your survey
• Start with a cover page for introduction
• Simple wording and easy grammar
• Assure common understanding
• Start with an interesting question
• Avoid leading questions
• Avoid double negatives
• Balance rating scale
• Avoid too long list of choices
33. Do’s and don’ts
• Avoid difficult concepts
• Avoid difficult recall questions
• Use closed ended questions
• Use logic order for questions
• Use square boxes with X placed in front, rather than
numerical labels for response questions
• Avoid double questions
• Avoid complex matrices.
• Use booklet format (not staples at the upper corner)
• Pre-test your survey
Data collection:1-Interview:Personal interviewTelephone interview2- self-administeredPaperComputer based3- mixed-mode survey
RANDOM ERROR (p):Deviation from truth in inference about underlying population which happens because sampling was part of studySYSEMATIC ERROR (BIAS):Any systematic error in the design, conduct or analysis of a study that results in a mistaken estimate of an exposure’s effect on the risk of diseases.