Factors Influencing Research Productivity at Njala University: A Count Regression Approach

International Journal of Research and Scientific Innovation (IJRSI) | Volume VII, Issue VII, July 2020 | ISSN 2321–2705

Factors Influencing Research Productivity at Njala University: A Count Regression Approach

Regina Baby Sesay^*1, Sheku Seppeh², Mohamed Kpangay^3
1,2,3Department of Mathematics and Statistics, School of Technology, Njala University, Njala, Sierra Leone

Abstract— Research promotes professional excellence as it helps academics to be more innovative thereby enhancing outstanding student education. Like most universities, Njala University (NU) academic staff are required to teach, research, and carry out community outreach. Publishing in a peer-reviewed journal is evidence of the effort made by academic staff to fulfill the obligation of one of the job descriptions of Njala University. The University like almost all other academic institutions promotes it, academic staff, purely based on the strength of their research work. The research strength of each academic staff is measured by the number of original research papers published in peer-reviewed journals. As such, the publication of original academic papers in recognized peer-reviewed journals has become the dream of each NU academic staff. However, despite the huge desire for publication, some unavoidable factors are infringing on the research activities of most academic staff. This research paper, therefore, used a statistical modeling technique for count data, to identify the main factors influencing the research productivity of NU academic staff. A stratified random sampling method was employed to select 113 respondents proportionately from each school. Data were collected from the selected respondents using structured questionnaires. The Poisson regression model was used as the baseline model. Due to the evidence of over-dispersion and excess zeros in the response variable, three additional count regression models were used in the analysis. Based on statistical tests, the zero-inflated hurdle model significantly outperformed the Poisson and Negative binomial regression models. However, the difference in performance between the zero-inflated poison and the zero-inflated hurdle model was not statistically significant. Initially, several factors were considered as possible determinants of research productivity of NU academics. However, the empirical analysis showed that academic qualification; teaching experience and hours spent on research are the main (significant) factors influencing the research productivity of Njala University academic staff. Increase in the number of hours spent on research can increase the number of research publications. Academics with more teaching experience tend to publish more than those with little or no teaching experience. The higher the academic degree attained by the academic, the higher the possibility of publishing more research papers.

Keywords—Count Regression, Njala University, Research Strength, Academics, Non nested models

І. INTRODUCTION

Research is a step by step search for knowledge. It involves fact discovery, and revision of accepted theories in the light of newly discovered facts, and the applicability of such discoveries to contribute to the development of the environment in which we live. According to [5], research is an objective and systematic analysis and recording of controlled observations that may lead to generalizations, principles or theories, resulting in predictions and possible control of events. This clearly points out the importance of research in the universities as it leads to innovations and invention of new knowledge. Many concerns have been raised with regards to factors influencing the research performance of universities, Colleges and academics. Amongst these are the work of [31] and [32].

Read More