Analysis of surveyor variability from data obtained from surveys of dwellings represented on video film

The surveying of houses is not an exact science. A survey is largely based on what are essentially subjective judgements by surveyors. These judgements are based upon the available evidence of the condition of a house, the judgement of one surveyor may well differ from another's, even if the judgement is based on the same evidence. This phenomenon is termed surveyor variability. This paper seeks to explore the issue of variability in the field of house condition surveys, from data obtained from a training exercise using video film of dwellings. The exercise was undertaken in 1996. The results of the analysis of surveyor performance show a wide range of repair costs for both of the dwellings in the video. The findings raise issues for both individual house surveys and large sample size surveys in terms of the accuracy of repair cost estimation for dwellings. The paper concludes that no one technique is likely to significantly improve the consistency of surveyors, and that a reduction in the potential for subjective decisions is the quickest, but not necessarily the best, strategy to reduce variability.