Page 311 - Contributed Paper Session (CPS) - Volume 2

CPS1873 Ferenc M. et al.

Experiences with the theme coding method of
cognitive questionnaire testing at Hungarian
Central Statistical Office
Ferenc Mújdricza, Mónika Földvári
Hungarian Central Statistical Office

Abstract
Cognitive testing aims at better understanding of the question-response
process. It provides a basis for exploring and reducing sources of response
error, thereby it enables the design of survey questions that fit better to the
research goal as well as the ultimate enhancement of survey data quality. Our
presentation focuses on the experiences of the 2017-2018 cognitive test
project of Household Budget and Living Conditions Survey (HBLS) at
Hungarian Central Statistical Office (HCSO), with a special attention to the
recent methodological questions of designing, conducting, and analysing
cognitive interviews. These experiences contribute to the improvement of
questionnaire testing methodologies of surveys utilising internationally
harmonised, extensive questionnaires, with particular regard to the
possibilities of questionnaire testing at NSIs.

Keywords
cognitive testing methodology; Theme Coding method; complex, extensive
questionnaire testing at NSIs

1. Introduction
The cognitive testing of the HBLS questionnaire served a twofold purpose:
1
questionnaire testing and Hungarian pilot for the recently adapted, Grounded
Theory-based Theme Coding approach of cognitive testing methodology. The
purpose of the pilot was to gain valuable experiences that enable the
improvement of the HCSO’s questionnaire testing protocols. Cognitive
interviewing (CI) studies provide means for revealing how respondents
interpret and answer the questions in the light of their own lives and
experiences. Furthermore, interpretational pattern and response process
characteristics of particular social groups can be discovered. The usability of
the questions can be analysed, enhanced, and, if necessary, repaired based on
the knowledge of these factors.
The aim of the HBLS is to collect data on income and living conditions of
Hungarian households. Working with representative probability samples, the
HBLS collects data on approximately 8,000 households and 20,000 persons of

The final report (Földvári, Dorkó & Mújdricza, 2018, p. 1-508) contains the description of the
1
entire testing process and its results in full detail.
300 | I S I W S C 2 0 1 9