Web Survey Bibliography

Title Propensity score and calibration as bias reducing techniques in surveys based on Internet panels: application to the outcome of the Swedish Parliament
Author Johansson, C., Lorenc, B.
Source The American Association for Public Opinion Research (AAPOR) 58th Annual Conference, 2003
Year 2003
Access date 06.05.2004
Abstract When systematic differences in key background variables with respect to a population are present, unweighted results of surveys based on Internet panels are often misleading. An inventive application of propensity score weighting, otherwise an established technique for reducing bias in observational studies (Rosenbaum, Rubin, 1984), to the field of Internet surveys was suggested by George Terhanian of Harris Interactive (e.g. Terhanian et al. 2001). The method is though still a matter of some controversy, possibly owing somewhat to the fact that its formal presentation has not been given yet.

After a preparatory study that showed viability of the propensity score technique in a simulated population with known properties (Lorenc, in preparation), we report here on its application to reducing bias in responses on voting behaviour of an Internet panel with respect to the population in the recent Parliament election in Sweden. For the present purpose we assume that the reporting is correct and that the difference in the unweighted means of the response variables between the panel and the population as a whole is due to systematic differences in auxiliary variables between the groups.

In addition, we compare propensity score weighting with calibration (Deville, Särndal, 1992), a method that uses known univariate or cross-tabulated population totals of auxiliary variables. Thus, a contrast is established with the propensity score technique, which uses data on the individual level and can exploit more richly eventual covariation in the auxiliary information.

Deville, Särndal (1992). "Calibration estimators in survey sampling". JASA, 87:376-82.

Lorenc (in preparation). "A study of effectiveness of propensity score weighting in a simulated population". Statistics Department, Stockholm University.

Rosenbaum, Rubin (1984). "Reducing bias in observational studies using sub-classification on the propensity score". JASA, 79:516-24.

Terhanian, Marcus, Bremer, Smith (2001): "Reducing error associated with non-probability sampling through propensity scores". JSM, Atlanta, Georgia, USA.
Year of publication2003
Bibliographic typeConferences, workshops, tutorials, presentations
Print

Web Survey Bibliography (6374)

Page:
Page: