Publication type
Survey Futures Working Paper Series
Series Number
24
Series
Survey Futures Working Paper Series
Authors
Publication date
July 29, 2026
Summary:
Social surveys frequently collect information on industry and occupation, typically using open ended questions about job title and duties, with responses coded by expert coders. The shift towards self-completion surveys has made this increasingly challenging, as respondents may provide insufficient detail without interviewers present to probe for clarification. Closed-list questions, in which respondents select from pre-defined categories, could reduce fieldwork and coding costs while removing ambiguities inherent to open-ended responses. This paper assesses whether closed-list questions produce classifications comparable to those derived from manually coded open-text responses. Using data from the NatCen panel, a UK probability-based panel, we find overall agreement rates between self-selected and office coded categories are 61.7% for industries and 55.8% for occupations at the highest classification level (1-digit), with considerable variation across categories, and much lower agreement rates (45.8%) at the 2-digit level. Although these results appear to caution against the use of closed-list methods, the reduced question response times and lower costs point to clear practical advantages, particularly if highly detailed classifications are not required. However, for this approach to succeed, the categories presented to respondents would need refinement to align with how individuals conceptualise their jobs. Further research is needed to establish how to achieve this.
Subjects
Paper download#589118