| Title |
Evaluating the impact of a non-probability sample-based estimator in a linear combination with an estimator from a probability sample |
| Authors |
Čiginas, Andrius ; Krapavickaitė, Danutė ; Nekrašaitė-Liegė, Vilma |
| DOI |
10.1177/0282423X251331346 |
| Full Text |
|
| Is Part of |
Journal of official statistics.. Thousand Oaks, CA : SAGE Publications. 2025, vol. 41, iss. 2, p. 649-674.. ISSN 0282-423X. eISSN 2001-7367 |
| Keywords [eng] |
propensity score ; variance estimation ; Poisson pseudo-sampling design ; super-population ; composite estimator |
| Abstract [eng] |
In this article, the estimators based on data from independent non-probability and probability samples are combined to estimate finite population parameters. Assuming that the values of the study variable are available in both samples, the integration of the non-probability and probability samples through a composite estimator of the population total is studied. The integration is done using a linear combination of the inverse probability weighted (IPW) estimator and a design-based estimator. By evaluating the variance of the former estimator, the randomness of the underlying non-probability sample is taken into account through the distribution of the estimated propensity scores. This approach is then compared with a variance estimator based on the asymptotic variance and with a bootstrap variance estimator. The proposed linear combination is not sensitive to the misspecification of the model for the propensity scores due to the incorporated estimator of the bias of the IPW estimator. The number of Lithuanian companies possessing websites is estimated in a simulation study. By combining the sample survey data and big voluntary sample data, the properties of the introduced estimators are demonstrated numerically. |
| Published |
Thousand Oaks, CA : SAGE Publications |
| Type |
Journal article |
| Language |
English |
| Publication date |
2025 |
| CC license |
|
