Scottish household survey: methodology and fieldwork outcomes (2014-16)

Current Weighting Strategy

The Scottish Household Survey (SHS) previously used a fairly simplistic weighting strategy. It was intentionally simple, and reflected, in part, a desire to keep the processes of the survey straightforward so that data could be made available for analysis as quickly as possible. It also reflected the limited extent to which SHS data differs substantially from comparator data. Thus, no additional corrective weighting had ever been applied to the data beyond that required to account for sample design and differential response rates between local authorities.

This aspect of the survey has been subject to review, as part of a study comparing non-respondents to the SHS with Census data. Whilst comparison with the Census showed some bias in the SHS, this was not substantial, although some corrective weighting was recommended

Weighting Strategy Review

During 2008 and 2009, the SHS contractors were asked to make proposals for developing and implementing a revised weighting strategy for the survey. The key driver behind this is to account for any non-response bias in the estimates, and for the strategy include the production of grossing factors to allow the survey to provide numerical, population based, estimates for survey measures (rather than the current percentages).

A report has been produced which outlines the outcomes of the review in assessing the SHS weighting strategy and an evaluation of impact on key estimates.

• Scottish Household Survey - Weighting Strategy Review 2009

Further information on the weighting strategy used is available through our range of publications, in particular the Methodology and Fieldwork Outcomes report