Abstract:
Affirmative responses to questions about conservation behavior in a multi-year phone survey are conceived of as a compound event, performing the behavior and then recalling it during a survey. By decomposing these two events an unbiased estimate of true, rather than reported, conservation behavior is possible. This technique is applied to the telephone survey data to form estimates of the probabilities of performing conservation behaviors in each of two years, as well as the conditional probabilities of performing the conservation actions in the second year.
Keywords:residential energy conservation; stated behavior survey (search for similar items in EconPapers) JEL-codes:D12 (search for similar items in EconPapers) Date: 2003-09-03, Revised 2003-12-28 Note: Type of Document - PDF; prepared on Kile Linux; to print on PostScript; pages: 29; figures: included View list of references