Academic journal article Electronic Journal of Business Research Methods

The Pervasiveness and Implications of Statistical Misconceptions among Academics with a Special Interest in Business Research Methods

Academic journal article Electronic Journal of Business Research Methods

The Pervasiveness and Implications of Statistical Misconceptions among Academics with a Special Interest in Business Research Methods

Article excerpt

1. Introduction

Statistical misconceptions are argued to hinder meaningful learning, impede research progress and interfere with decision making (Huck, 2009). For students, such misconceptions may be generated by poor understanding reinforced by statements uttered or written by one's mentors (Huck, 2009). The study seeks to determine whether academics with a special interest in business research methods hold mainstream statistical misconceptions, thereby extending a recent study that investigated the prevalence of research methods mis/conceptions with the same target group (Bezzina & Saunders, 2013). To date, limited research has examined the pervasiveness of statistical misconceptions among academics; the studies we reviewed focused on either identifying students' statistical misconceptions (e.g., Bezzina, 2004; Huck, Cross & Clark, 1986; Mevareck, 1983) or statistical flaws made by authors in published articles, reports and textbooks (e.g. Huck, 2009; Lance, 2011; von Hippel, 2005). Consequently, this research enables academics to determine whether faulty thinking has infected academics' notions about mainstream statistical concepts and considers the impact of these on their students. In addition, in the light of the findings that emerge, this paper provides some important suggestions for the teaching of business research methods, particularly on what the state of practice should be.

2. The nature of misconceptions and the role of academics

Misconceptions are "views or opinions that are incorrect due to faulty thinking or misunderstanding" (Bezzina & Saunders, 2013, p. 41)", representing deviations from widely accepted norms and conventions. In some cases, the practices themselves are not intrinsically faulty but rather, it is the reasoning why or rationalisation used to justify the practices that is questionable (Lance & Vandenberg, 2009).

Misconceptions arise from prior learning or from interacting with the social/physical world and interfere with learning concepts (Smith, diSessa & Roschelle, 1993). Some are grounded in human intuition that leads to faulty thinking, while others are generated by inconsistencies in textbooks and oral presentations in classrooms (Huck, 2009). Garfield (1995, p.32) highlights that misconceptions are often so strong and resilient that "they are slow to change even when students are confronted with evidence that their beliefs are incorrect". Similarly, Mevareck (1983) argues that when statistical misconceptions become deeply engrained in the underlying knowledge base of the individual, mere exposure to more advanced courses is not sufficient to overcome them. However, Brown and Clement (1989) note that successful instructional confrontation can replace faulty misconceptions with new expert knowledge in a short period of time while Smith, diSessa and Roschelle (1993) advise that the emphasis should be on knowledge refinement and reorganisation rather than replacement. Given that faulty thinking is such a pervasive phenomenon, it is important that academics as instructors are aware of their own misconceptions and the impact of these upon their students (Bezzina & Saunders, 2013).

3. Statistical data analysis and statistical misconceptions

Statistical data analysis is the process by which data are transformed with the aim of extracting useful information and facilitating conclusions. Each statistical technique has underlying conceptual and statistical assumptions that must be met if the results are to be valid (Gel, Miao & Gastwirth, 2005). Various structured- model building approaches and step-by-step guides are available to facilitate this process of data analysis. The scope behind them is to provide researchers with "a broader base of model development, estimation and interpretation" (Hair et al., 1998. p. 25) not a rigid set of procedures to follow. Structured approaches do not come without criticism. Conflicting viewpoints arise on various aspects such as the required sample size, the statistical model to analyse the data, and the quality of the input data. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.