Statistical methods used in Arabic journals of library and information science

Abstract

Statistical methods are crucial for research studies in LIS to reach, analyze, and critique results. The current study aims to investigate the use of statistical methods in the Arab literature in LIS, provide a comprehensive review of commonly used statistical methods, and reveal the variables affecting the use of statistical methods. Research papers published in the eight Arab journals from January 2014 to December 2018 were investigated according to three categories of variables related to characteristics of article, author, and research methodologies. Statistical methods were classified into four categories: descriptive methods, parametric inferential methods, nonparametric inferential methods, and predictive methods. The total usage of statistical methods reached 1037 times. Descriptive statistical methods were the most commonly used, followed distantly by predictive statistical methods. Frequency distributions and percentages were used in most quantitative studies. While t-test and ANOVA were the most parametric methods used, Chi-squared was the most non-parametric method used. Most of the studied variables influence the Arab researchers' choice of the category of statistical methods used in their research in LIS. Besides, the author's characteristics are the most influential variables in the use of statistical methods, followed by the research methodology, while the effect of article characteristics is less.