Özet:

This study, in the presence of lost data, aims to study the relationships between the material and the test parameters cut using different methods of lost data in already selectively tested tests and identify the methods of lost data that are suitable for use in such tests. In this study, designed as a research in the basic type of research, the relative scan model, the analysis was carried out on the SBS (level determination test) 2011 Mathematical Test A Library data for 527517 respondents. From the methods based on deletion in data analyses, the methods of ‘range deletion’, the methods of simple assignment, the methods of ‘0 assignment’, the methods of ‘series average assignment’, the methods of ‘series average assignment’, the methods of ‘series average assignment’, the methods of ‘series average assignment’, the methods of ‘series average assignment’, the methods of ‘series interpolation’ and the methods of ‘series trend point assignment’, the methods of ‘regression assignment’, the methods of ‘series-maximation algorithm’ and the methods of ‘data multiplication’, the methods of ‘Markov chains’ and the methods of ‘Monte Carlo’ and the methods of ‘Markov chains’ and the methods of ‘Markov chains’. The findings obtained indicate that in statistical cuts for already selectively tested tests, if the lost data is not neglected, it is necessary to use a suitable method of loss data. Removal-based methods and the 0 assignment method are not the appropriate methods for such data. The simple methods of assignment are likely to produce side-cuts. The most probability and multi-data assignment methods are considered to be appropriate for use in such data as lost data methods.

Anahtar Kelimeler: