The Development of a Household Panel Survey Plan Prepared for Puget Sound Council of Governments by

From a statistical viewpoint, a panel survey has the definite advantage that it offers more accurate estimates of changes than would repeated cross—sectional surveys with the sample size (Cochran, 1977, p.345). For example, suppose one wishes to monitor changes in bus use that may follow an introduction of a new line, and a sample of 1,000 households is drawn for a survey prior to the introduction of the line. If estimating change is the primary concern, then it is best to survey the same 1,000 households again after the new line opens. This increased accuracy offered by the panel survey implies that its sample size is smaller than the sample size required for a cross—sectional survey to attain a given level of accuracy of the estimates of changes.