Summary

The statistic survey perspectives can be viewed as a design perspective and quality perspective (Groves et al., 2004). The design perspective leads from concepts through "constructs" and measurements to questions to become a process and one of its stages is the response process. The quality perspective makes numerous references to "error". The sampling error, nonresponse error are just two examples. Measurement errors refer to the gap between what is called the ideal value and the obtained response, i.e. at the response process stage. Survey methodologists attribute deviations from perfect measurements to cognitive problems in the response process. Hence, these problems lie at the heart of the response process model. Originally, the model was developed to reflect aspects of households and individual surveys. Further development of cognitive research extended the model to fit the response process in business surveys. A merger of the two produced a Hybrid Response Process Model for Business Surveys, a complex and general model encompassing the entire response process in business surveys. Since it still did not fully and clearly address numerous aspects the model has recently been developed into the Multidimensional Integral Business Survey Response Process Model.

The response process models can serve as a framework for the evaluation of business surveys (Giesen, 2007). The linkage between model steps and observations of real respondent behaviour when dealing with survey requests, provides the structure which can help to analyse this complex activity. This is a way to spot problems and try to fix them for the future. Furthermore, considering the data collection instrument and the response burden connected with answering items it contains, response process steps make it possible to establish at which stage the burden is especially heavy and what can be done to ease it. This can improve the questionnaire and even influence its design. The division of the response process into separate stages was the foundation of cognitive methods for pretesting survey questions. Cognitive interviewing, understood as an extension of the standard interviewing process of eliciting answers to questions, studies processes distinguished in the response process model (Willis, 2004). The foundation of the response process for establishment surveys, which is more complex and contains more steps, adequately allows to split survey evaluation into the response process steps. When the data collection process and the response burden are assessed using different methods (Giesen, 2007) and the findings are linked with the response process stages it is possible to establish the nature of the problems, whether cognitive or logistic, and consequently, adopt the results to improve the data quality or ease the response burden.

To read the entire document, please access the pdf file (link under "Related Documents" on the right-hand-side of this page).

Your feedback is appreciated. Please send your remarks, suggestions for improvement, etc. to memobust@cbs.nl.