Statistical software for non-statisticians and non-computer programming students in education and social science disciplines: An evaluation of four contemporary tools

Abstract

Most doctoral students in education and social science disciplines struggle to become proficient in statistical analysis for multiple reasons (e.g., instruction, textbooks, motivation, attitudes, etc.). A factor that seems to have been overlooked is the design of software used for statistical analysis. Most software has not been designed for non-statisticians or students who are not familiar with computer programming. Hence, this paper explores human-computer interaction (HCI) factors in contemporary software tools to identify possible issues that contribute to or hinder successful statistical problem-solving. HCI factors include the range of statistical operations available; technical properties (user interface design, data visualization, entry, and manipulation); and usage properties (speed, ease of used, and efficiency). SPSS, RStudio, R Commander & jamovi software systems were selected for detailed evaluation. Analysis suggests that HCI factors are likely to interfere significantly with the completion of statistical tasks for doctoral students in education and social sciences,