New estimation techniques for ordinal sensitive variables

Methods to analyze multicategorical variables are extensively used in sociological, medical and educational research. Nonetheless, they have a very sparse presence finite population sampling when sensitive topics investigated data obtained by means of the randomized response technique (RRT), survey method based on principle that questions must not be asked directly respondents. The RRT is with aim reducing social desirability bias, which defined as respondent tendency release personal information according what socially acceptable. This nonstandard data-collection approach was originally developed deal dichotomous responses questions. Later, idea has been extended multicategory responses. In this paper we consider ordinal more than two categories. particular, first discuss theoretical framework for estimating frequency categories subjected misclassification due use particular RRT. Then, show how it possible improve efficiency inferential process employing auxiliary at estimation stage through calibration approach. Finally, assess performance proposed estimators Monte Carlo simulation study.