Designing survey questions that clearly and precisely communicate the question's intent and elicit responses based on the intended interpretation is critical but often undervalued. We used cognitive interviewing to qualitatively assess respondents' interpretation of and responses to questions pertaining to maternal and child nutrition intervention coverage. We conducted interviews to cognitively test 25 survey questions with mothers (N = 21) with children less than 1 year in Madhya Pradesh, India. Each question was followed by probes to capture information on four cognitive stages—comprehension, retrieval, judgement, and response. Data were analysed for common and unique patterns across the survey questions. We identified four types of cognitive challenges: (1) retention of multiple concepts in long questions: difficulty in comprehending and retaining questions with three or more key concepts; (2) temporal confusion: difficulty in conceptualizing recall periods such as “in the last 6 months” as compared to life stages such as pregnancy; (3) interpretation of concepts: mismatch of information being asked, meaning of certain terms and intervention scope; and (4) understanding of technical terms: difficulty in understanding commonly used technical words such as “breastfeeding” and “antenatal care” and requiring use of simple alternative language. Findings from this study will be useful for stakeholders involved in survey design and implementation, especially those conducting large‐scale household surveys to measure coverage of essential nutrition interventions.

Objectives Designing survey questions that clearly and precisely communicate the question's intent and elicit responses based on the intended interpretation is critical but often undervalued. We used cognitive interviewing to qualitatively assess respondents’ interpretation and responses to questions pertaining to maternal and child nutrition intervention coverage. Methods We conducted interviews with mothers (N = 21) with children less than one year in Madhya Pradesh, India, to cognitively test 25 survey questions. Each question was followed by probes to capture information on four cognitive stages - comprehension, retrieval, judgement, and response. Interviews were recorded and notes were taken on verbal and non-verbal cues. Data were analyzed for common and unique patterns across the survey questions within the cognitive domains and grouped into challenges. Results We identified four types of cognitive challenges: 1) Poor retention of multiple concepts in long questions: difficulty in comprehending and retaining questions with three or more key concepts; 2) Temporal confusion: difficulty in conceptualizing recall periods such as “in the last 6 months” as compared to life stages such as pregnancy; 3) Misinterpretation of concepts: misinterpretation of the information being asked; meaning of certain terms such as “animal-source foods” was considered as referring to meat products only and not milk and eggs; scope of intervention using the phrase “talk with you” in referring to counseling was interpreted in different ways by respondents; and 4) Poor understanding of technical terms: difficulty in understanding even commonly-used technical words such as “breastfeeding” and “antenatal care” requiring the use of plain and simple alternative language. Conclusions Findings from this study will be useful for stakeholders involved in survey design and implementation, especially those conducting large-scale household surveys to improve coverage data of essential nutrition interventions, which is critical for policy actions Funding Sources Bill & Melinda Gates Foundation through the DataDENT initiative and the Improving Measurement and Program Design grant, and the CGIAR Research Program on Agriculture for Nutrition and Health (A4NH), led by the International Food Policy Research Institute.

No abstract