The pervasiveness and implications of statistical misconceptions among academics with a special interest in business research methods

Abstract

Statistics play a very important role in business research, particularly in studies that choose to use quantitative or mixed methods. Alongside statistical analysis, aspects related to research design (such as sampling, reliability and validity issues) require a good grounding in statistical concepts reinforced by careful practice to avoid potential mistakes arising from statistical misconceptions. Although quite a considerable number of published studies have focused on students' faulty thinking regarding statistical concepts, little research explores the extent to which these are also held by academics who are their instructors. This empirical study addresses this by answering the following questions: First, are statistical misconceptions pervasive among academics with a special interest in business research methods? If so, second, is there an association between the pervasiveness of statistical misconceptions and the preferred research tradition (qualitative, quantitative, mixed methods)?. Data were collected via a web questionnaire from a purposive sample of academics with an expressed interest in business research methods. The questionnaire comprised 30 categorical statements (agree, disagree, don't know) focusing on statistical misconceptions (and conceptions) relating to descriptive statistics, design strategies, inferential statistics and regression, and five demographic questions. We targeted a critical case purposive sample of 679 potential respondents. Although 166 consented to take part, only 80 completed the questionnaire and their responses form the basis of the statistical analysis, a response rate of 11.8 %. The study provides empirical evidence of both an absence of knowledge and a high pervasiveness of faulty notions that have infected the thinking of academics relating to both research design and the use of statistics. This is particularly so for academics who prefer quantitative methods, those preferring qualitative methods being more likely to admit that they do not know. The study argues that such lack of knowledge and misconceptions reduce the true utility of statistics in research. Recommendations are offered regarding the teaching of statistics within business research methods.