COMPARISON OF THE MAXIMUM LIKELIHOOD ESTIMATOR WITH THE UNBIASED ESTIMATOR FOR THE FRACTION OF SENSITIVE QUESTIONS IN THE NON-RANDOMIZED CROSSWISE MODEL Cover Image

PORÓWNANIE ESTYMATORA NAJWIĘKSZEJ WIARYGODNOŚCI Z ESTYMATOREM NIEOBCIĄŻONYM ODSETKA PYTAŃ DRAŻLIWYCH W MODELU KRZYŻOWYM NIEZRANDOMIZOWANYCH ODPOWIEDZI
COMPARISON OF THE MAXIMUM LIKELIHOOD ESTIMATOR WITH THE UNBIASED ESTIMATOR FOR THE FRACTION OF SENSITIVE QUESTIONS IN THE NON-RANDOMIZED CROSSWISE MODEL

Author(s): Stanisław Jaworski
Subject(s): Social Theory
Published by: Szkoła Główna Gospodarstwa Wiejskiego w Warszawie
Keywords: nonrandomized crosswise response model; sensitive questions;

Summary/Abstract: One of the problems in surveys is estimating the fraction of responses to sensitive questions. It requires the use of special response models. The first model of this kind was the randomized response model [Warner 1965]. There are many other models nowadays. Among them is the non-randomized crosswise model [Yu i in. 2008]. The estimator of the fraction of the sensitive questions in this model can take values outside the range of the interval á0, 1ñ. There is shown in the paper how the truncation of the estimator to the interval is influencing its properties.

  • Issue Year: XXIII/2022
  • Issue No: 2
  • Page Range: 35-46
  • Page Count: 12
  • Language: Polish