Investigating the impact of training influence on employee retention in small and medium enterprises: a regression-type classification and ranking believe simplex analysis on sparse data

This study investigates the impact of available training alternatives TAs on employee retention in small and medium enterprises SMEs. A noticeable problem with this research issue is that individual SMEs may utilize different combination of TAs. The considered survey questionnaire allowed respondent SME owners/managers the option to gauge the level of satisfaction of a TA or to indicate that they did not use it. It follows, therefore, that the survey-based data set is sparse, in the sense that the 'did not use' option infers that a form of missing value is present Likert-scale-based satisfaction value present if a TA was used. To facilitate an effective analysis of the considered sparse data set, because the missing values have meaning, the nascent regression-type classification and ranking believe simplex RCaRBS technique is employed. As a development of the CaRBS technique, this technique is able to undertake multivariate regression-type analysis on sparse data, without the need to manage the missing values in any way. Results are presented from the RCaRBS analyses relating to SME owner/managers' satisfactions with TAs and their impact on two employee retention facets, namely greater employee loyalty and, conversely, losing an employee to a competitor. Emphasis here is on the graphical elucidation of findings in regard to model fit and TA contribution. The pertinence of the study is the inclusiveness of the data considered a novel approach to analysing sparse data, and the comparisons between these associated issues of TA satisfaction and employee retention.