The Effects of Rotation Group Bias on Estimates from Panel Surveys

The Evolution of Rotation Group Bias: Will the Real Unemployment Rate Please Stand Up?

“…Bailar (1975) was among the first to document the rotation group bias in the CPS. Using data from 1968-1972, she found that the unemployment rate estimated from different rotation groups in the same time period were systematically different.…”

Measurement biases in panel data

“…One could expect that with an increase in the number of waves respondents are participating in, a between wave fatigue develops, leading to a decline in reporting accuracy. These between-wave biases are equivalent to what are commonly referred to as panel effects (Bailer 1975(Bailer , 1979Tanur 1981). Experienced panel-members may.keep their diaries differently than new members.…”

“…The use of multi-day travel diaries in a panel context may lead to additional problems (Neter & Wakesberg 1964;Scheuch 1972;Bailer 1975Bailer , 1979Tanur 1981). If the accuracy of reporting differs across waves, biased aggregate mobility trends will be found.…”

LLCS Vol 3, Issue 1, Full Issue

“…Most frequently, researchers compare responses from some later wave of a panel to a temporally corresponding cross-sectional sample to which the same or similar questions were asked (Corder and Horvitz 1989;Menard and Elliott 1993;Osgood, O'Malley, Bachman and Johnston 1989;Silberstein and Jacobs 1989;Wilson and Howell 2005). Rotating panel designs provide a better alternative because comparisons can be made between identical survey designs, sample designs, procedures and identical questions (Bailar 1975;Cohen and Burt 1985;Ghangurde 1982;McCormick, Butler and Singh 1992;Pennell and Lepkowski 1992;Silberstein and Jacobs 1989). A third analytic strategy involves examination of a single sample and predicting specific effects, from theory, about panel conditioning (Pevalin 2000;Sturgis, Allum and Brunton-Smith 2009).…”