JACKKNIFE APPROACHES FOR IMPROVING DATA QUALITY

Olayiwola O. M1, Akintunde A.A.1, Yusuff K. M.1, Ajibade F.B.2and Adekpe D.O.1

Department of Statistics, College of Physical Sciences, Federal University of Agriculture, Abeokuta, Nigeria

Department of General Studies, Mathematics units, Petroleum Training Institute, Effurun, Delta State

Nigeria.

Abstract

Knowledge of the magnitude of error at each sampling step is necessary to know how

to improve data quality. This study derived three jackknife approaches for reduction

of sampling error. Jackknife leaving one cluster value out (JK1), Jackknife leaving

one stratum out (JK2), and Jackknife within each stratum (JK3). The Statistical

properties of these Jackknife approaches were examined and compared. JK1 is the

most efficient approach for sampling error reduction followed by JK3 and JK2.

Keywords: Jackknife, Data quality, sampling error, Statistical properties