Questionnaire Design

Question 1

Lack of precision/reliability in measurements

Answer 1

random error – loss of statistical power; wider confidence intervals

Question 2

Lack of accuracy/validity in measurement

Answer 2

systematic error (bias) which will not be mitigated by increased sample size or repeat measurements; must change study design/procedures

Question 3

Principals of questionnaire design

Answer 3

Appropriate: able to provide answers to the research question/s

Intelligible: the respondent can understand (language, level of education etc.)

Unambiguous: the question means the same thing to the respondent and the inquirer

Unbiased: you are no more likely to elicit one kind of response than another (recall, social desirability etc.)

Capable of coping with all possible responses: omni-competent (try to anticipate possible answers; don’t know option)

Well-coded (for data capture & analysis): check for ambiguity & overlap

Ethical: informed consent, sensitivity (Autonomy, Beneficence etc.)

Piloted:
1. assess performance of tool
2. iron out design faults

Short and simple as possible

Logical sequence

Maximize response rate (majority of those eligible participate: representative sample)

Minimize missing data: avoid blanks, ‘other’, text is difficult

Good & well thought out data collection = high quality data = easier analysis & interpretation = useful conclusions

Justify your choices – tools, questions, scores

Question 4

Pros of interview administered questionnaire

Answer 4

Ensures good completion of questions, interviewer can re-explain certain terms, participant does not need to be literate, researcher can develop a good rapport with the participant (+++this aids in removing any fears the participant may have

Question 5

10 steps of questionnaire development

Answer 5

1. Research Question, Objectives & Study Design
2. What data will you need to collect? Variables
3. Map out key data elements
4. Select method of administration
5. Develop the questions
6. Anticipate the responses
7. Format the questionnaire
8. Coding
9. Maximize validity & reliability
10. Pilot
11. Ethics

Question 6

Cons of interview administered questionnaire

Answer 6

Cons: Time consuming, interpersonal factors can affect data collection (i.e. elderly researcher interviewing adolescent females about contraception), if interviewer does use a standardized method for administration- variation can arise and impact data quality and also more expensive

Question 7

Pros of self- administered questionnaire

Answer 7

: Good strategy for completion of sensitive information (i.e. adolescent health surveys), less time consuming {could use an audio-recorded questionnaire on a Tablet, could make format of questionnaire more visually appealing than paper copy}

Question 8

Cons of self- administered questionnaire

Answer 8

Participant must be literate, questionnaires must be clear and well laid out, lack of skip patterns/direction/explanation can impact completion of the questionnaire

Question 9

2 types of question structure

Answer 9

• open ended
  -close ended

Question 10

pros of open ended question

Answer 10

Can be used when there are an extensive list of possible options (good for exploratory work)

Question 11

cons of open ended question

Answer 11

Time-consuming to read and categorize, responders can get confused as to how to answer these

Question 12

pros of close ended questions

Answer 12

Less time spent coding and analyzing, shortens the interview

Question 13

cons of close ended questions

Answer 13

Responses could be limited depending on researcher’s knowledge
Include all reasonable responses & response options should be mutually exclusive

Question 14

Tips

Answer 14

• home language
  -simple, short, specific
• avoid technical and imprecise terms
• could use visuals or graphics
• include reference frames to avoid confusion
• avoid double barreled questions
• avoid leading/ emotional language
• use text boxes for explanations of key words

Question 15

Reliability/precision:

Answer 15

-The questionnaire delivers consistent results
-Standardize data collection procedures
-Train interviewers
-Supervise data collection
-Test-retest reliability: administer to same subjects after brief interval

Question 16

Validity/accuracy

Answer 16

-The questionnaire measures what it is intended to measure
-Free of bias (recall/social desirability/non-participation)
-Check results against an independent source
-Are you using a validated tool?