ENVM7003 Chap.4 Designing a Questionnaire
Designing a Questionnaire
Two decisions sit upstream of every item: who is in the sample and how the instrument reaches them. The workshop puts both before drafting because they govern length and question type.
A twelve-minute instrument is reasonable online and impossible at a bus stop, and an item about a local trial fails with anyone outside the trial area.
The build is then structured by four question families, which are kinds of claim about a respondent rather than subject categories. Attribute items establish characteristics. Behaviour items ask what a person does. Attitude items ask for a judgement.
Values and beliefs items ask what a person holds important. Each family has its own error risk: overlapping categories, unrecallable counts, judgements the respondent has no basis to make, and value lists contaminated by activities. Within closed items the course distinguishes scale, category and ranking forms, and asks for a mixture with open items too.
Piloting closes the loop, and the weekly task is explicit that you must report one issue the pilot revealed, the revised item and the reason for the change.
What this chapter covers
- 01
4.1 Sample population and mode of distribution decided first
- 02
4.2 The four question families and the error each one invites
- 03
4.3 Open and closed items, and the three closed forms
- 04
4.4 Preliminary information that makes answers comparable
- 05
4.5 Piloting: what to watch for and how to report the repair
- 06
4.6 The pre-submission checklist the worksheet supplies
An eight-item instrument on household water use
- +2Fix the frame. Question: how do households in a metered outer-suburban area describe and justify water-use choices during restrictions? Population: adult householders responsible for utility decisions. Mode: online self-complete.
- +2Two attribute items. Household size in exhaustive bands, and dwelling type with an 'other' option whose text you keep separately.
- +3Two behaviour items. Watering frequency bounded to the past seven days, and a tick-all list of comparable actions taken in the past year.
- +3Two attitude items. Acceptability of a summer outdoor-watering limit, and willingness to pay a fixed levy, both on mirrored five-point ladders.
- +2Two values items. A capped 'select up to two' list of what matters about household water, followed by an open item asking the reason behind the choice.
Key terms
- Mode of distribution
- How the instrument reaches respondents: online, in person, by telephone or on paper. It constrains length, layout, whether anything can be explained, and response rate.
- Behaviour question
- An item asking what a respondent does, such as transport use or consumption patterns. Its characteristic failure is asking for a count over a period nobody recalls.
- Attitude question
- An item asking for a judgement about something, such as a view on a proposed development. It fails when the respondent has no basis on which to judge.
- Values question
- An item asking what a respondent holds to be important, such as the importance of environmental protection. It fails when activities are mixed into the option list.
- Ranking item
- A closed form asking respondents to order options against each other. It forces trade-offs and says nothing about absolute importance, so a first rank may still be trivial.
- Pilot
- A trial run of the instrument with a small number of people before fielding, used to find problems with clarity, flow and difficulty while they can still be fixed.
- Frame of reference
- Wording that gives the respondent a shared basis for answering, such as a map defining a place or frequency bands translated into everyday terms.
Designing a Questionnaire FAQ
How long should a questionnaire be?
Long enough to answer the question and short enough for the mode. The weekly task asks for eight items, which is a deliberate constraint rather than a recommended length. In a real project the binding constraint is the mode: an intercept survey has to be short and closed, while a self-complete online instrument can carry a map, a definition and an open item.
Do I have to include open-ended questions?
The workshop asks for a mixture of open and closed items, and there is a methodological reason beyond compliance. Scales tell you how strongly and never why, so an instrument with no open item produces a discussion section of numbers the group cannot interpret. One well-aimed open item usually carries the whole discussion.
What does piloting actually involve?
You swap instruments with someone, complete each other's, and give feedback on five things: whether the questions and instructions were easy to follow, whether any were unclear, how the flow felt, whether any were skipped and why, and what worked or could improve. Then you revise and report the change.
What should I look for when I pilot?
The hesitations rather than the answers. Where the respondent re-reads, asks a question, or qualifies with 'it depends', you have found a missing condition or an ambiguous term. Note it on the spot, because respondents rationalise those moments away if you ask afterwards.
Can I ask for demographic details?
Only where a planned comparison needs them. The worksheet checklist warns against unnecessary, intrusive or identifying information, and every demographic item is both a privacy cost and a drop-off risk. If you cannot name the comparison the item supports, the item does not belong.
Assessment move
Draft the instrument in one sitting and revise it in another, with the pilot in between. The revision is where the marks are: forty per cent of the weekly mark is applied research reasoning, and a piloted repair with a named cause demonstrates that better than a polished first draft ever will. Build the instrument your group project actually needs rather than a practice one, so the weekly task doubles as a project deliverable.
Organisational Behaviour · Identifying and Assisting Students at Risk · Digital Marketing and Social Media
