Application of The Sequential Hot-deck Imputation Method for Identification of Indonesian Standard Classification of Business Fields (KBLI)

Abstract

The Covid-19 pandemic requires the adjustment of new habits in daily life, including in a series of data collection processes. One of the new adjustments is to use alternative types of data collection other than face-to-face, such as the telephone and the web. Information collected through telephone interviews is less accurate than the same information collected through face-to-face interviews, such as the level of non-response, consistency between entries, and outliers in the data or often identified as missing values. Missing value will be very influential on data quality when it appears on important variables. One of these variables is the Standard Classification of Business Fields (KBLI). Imputation is one method that can be used to deal with this problem. One method that is quite popular is Sequential Hot-deck Imputation. Therefore, this study aims to facilitate the identification of 5-digit KBLI by utilizing the Sequential Hot-deck Imputation method. The results of this study indicate that the use of the Sequential Hot-deck Imputation method in the KBLI identification process gives very high accuracy results. In addition, the use of this method is very efficient in the identification process, because the time required is very short, even in large datasets.