The Use of Pre-Existing Survey Questions: Implications for Data Quality

When designing survey questionnaires, the potential inclusion of existing questions is a possibility that is not often contemplated by researchers and students, despite it being perfectly feasible This is partly due to the pressures of being ‘original’ in the academic and research worlds, but also because of a general lack of awareness of the availability of ready-made questions used in UK social surveys (See for instance: Lamb 2004). This paper draws attention to a facility that seeks to remedy this problem. The ESRC Question Bank (Qb) [URL:http://qb.soc.surrey.ac.uk] is an online resource that provides access to a repository of social survey questionnaires, as well as information and commentary on social measurement in key topic areas. It holds questionnaires (from 1991 onwards) from 57 large scale UK national probability sample surveys, as well as further information about each survey. One advantage of using these pre-existing questions is that they will have been extensively tested at the time of first use. Information on the exact reliability of each question cannot always be easily accessed; where such information is available, it can be found using the further resources pointed to by the Question Bank. In a few cases, for example the conceptualisation and measurement of ‘social class’ and socio-economic position, quite a lot of methodological work has been done. On other topics, there is much less, at least publicly available. The topic commentaries in the Qb allow concepts to be explored in terms of how they can be measured. Taking these into account could help to shed light on the extent to which particular questions can accurately measure one’s concept of interest. This extent, unfortunately, may not always be large – compromises may have to be made, or questions adapted to fit one’s requirements. This paper discusses the implications this could have for the quality of survey data obtained from these questions, particularly in terms of validity. One major caveat of using preexisting questions is the potential result of low data quality if measures are unreliable. However, if ‘recycled’ questions are accurate measures of the concept of interest (and many will have been pre-tested to ensure this), the degree of validity is ikely to be high, resulting ultimately in obtaining data of higher quality.