Detection Careles From Responden Within Examination Outlier Data Identifying Respondent's Carelessness Within The Outlier Data

摘要

This research examined data of the selected respondents of 300 students majoring in management by processing instrument tests andidentifying primary data from the possibility of respondents’ carelessness in filling in the questionnaire. This study applied data analysis technique ofMehanolibis test in order to diminish bias results of the questionnaire. The use of questionnaires for this quantitative research is to answer and describethe extent to which the respondent answers with responsibility or only perfunctory. The researcher uses Rasch model in an attempt to acquire a moreaccurate result compares to any other models. Rasch model assists researcher to obtain a maximum result even to the level of person correlation sinceit generates a more replicable one. Instrument testing and validation are inevitable and essential elements before stepping into inferential statistics thatseeks to acquire an answer to the proposed research question.