Survey Design and the Art of Asking Questions
The productive tension
Structured measurementandrespondent reality
The synthesis
The measurement camp treats survey design as an engineering problem: optimise the instrument, standardise the scales, eliminate variance, and extract the truth. The respondent-reality camp argues that surveys are social encounters, not measurement instruments: the answers you get depend on how you ask, when you ask, who you ask, and what the respondent thinks you want to hear. Both perspectives are correct. A survey measures what you design it to measure, not what you think it measures. The evidence-based survey designer combines the rigour of psychometric measurement science with a deep understanding of how real human beings -- distracted, satisficing, context-dependent -- actually experience and respond to questionnaires. The instrument must be technically sound AND humanly navigable. The error is designing surveys for an idealised respondent who does not exist.
Learning objectives
- →Design effective survey questions by identifying and avoiding common question construction errors including leading, double-barrelled, loaded, and ambiguous questions
- →Evaluate different scale types (Likert, semantic differential, NPS, ranking, constant sum) and select the appropriate scale for different research objectives
- →Explain question order effects including primacy, recency, anchoring, and context effects and design questionnaire flow to minimise their impact
- →Critically assess the Net Promoter Score as a measurement instrument, articulating both its practical appeal and its methodological limitations
- →Design a complete questionnaire flow from screening through demographics, applying principles of pre-testing and piloting to improve data quality
Members only
This lecture is part of a paid plan
The first lecture of every module is free — no account needed. The rest unlocks with a subscription. One price, all 120 lectures, both languages.