|
Abstract:
|
The National Agricultural Statistics Service introduced the standard occupational classification system to categorize agricultural laborers reported on the Agricultural Labor Survey into finer-grained distinctions. The presentation of this classification system to respondents may be subject to response order effects given that close-ended response options often suffer from primacy effects, which may be exacerbated when surveys are administered via CATI. To assess the order effects present in the Agricultural Labor Survey, the distributions of agricultural laborers reported via the telephone and mail were compared. Analyses of various types of operations examine if these order effects present themselves across operations by size and farm type. The results highlight the order effects that emerge when respondents are presented with unordered lists of response options, and the need to randomize the presentation of occupational categories to reduce biasing the classification of agricultural workers.
|
