The Philosophical Foundations of Prescriptive Statements and Statistical Inference

Abstract

From the perspectives of the philosophy of science and statistical inference, we discuss the challenges of making prescriptive statements in quantitative research articles. We first consider the prescriptive nature of educational research and argue that prescriptive statements are a necessity in educational research. The logic of deduction, abduction, and induction in philosophy of science are briefly reviewed, and the logic of prescriptive statement is specifically considered. The inductive nature of statistical inference is examined for both classical frequentist statistics and Bayesian statistics. To conclude our discussion, we recommend conducting replications and building "research programs" in educational research. We also make recommendations on what to write in the discussion section and what evidence is required to make an appropriate prescriptive statement.Prescriptive statements frequently appear in the discussion/conclusion section of research articles and consist of recommendations about how things should be done in practice. While skimming through the issue 2, volume 102 of the Journal of Educational Psychology, we came across a very typical prescriptive statement in the first article: "The fact that the LSC had a significant impact on struggling sixth-grade readers' reported problem-solving strategy use suggests that the LSC can help struggling students in early adolescence develop at least an awareness of strategies for overcoming or at least compensating for their reading difficulties. This is particularly important, given research that suggests that problem-solving strategies are most associated with skilled reading" (Cantrell et al. 2010, p. 270); and in the third article, "Results of the current study point to technologically mediated (remote) delivery of literacy coaching as a promising alternative to the common coaching practice of in-person visits" (Powell et al. 2010, p. 310) and then Educ Psychol Rev (2011) 23:207-220

Making and justifying evidence claims: evidence synthesis of impact evaluations and systematic reviews in international development

“…In addition, a number of papers already described include aspects of relevance and usefulness in regard to making claims (Cartwright, 2021;Greene, 1987;Sun and Pan, 2011), and there is also overlap with the communication of claims (Lazarus et al, 2015;Shyagali et al, 2022).…”

When is It Acceptable to Make Prescriptive Statements in Educational Research Articles? Commentary on the Special Issue Papers

“…That divide not only cuts across different scholarly disciplines in education such as psychology, philosophy, and curriculum studies but also separates scholars within particular disciplines such as the teaching of reading, mathematics, and science. It is not merely a divide, I think, between those who gravitate toward quantitative and experimental methods on the one hand, versus those who prefer qualitative and descriptive methods on the other hand, as suggested in some of the papers in this special issue (e.g., Sun and Pan 2011;Marley and Levin 2011). Instead, it is a divide, I think, between those who believe in the scientific method in the classic sense and those who believe the world is too complex to be studied or analyzed in this way.…”

“…Several of the articles address the different philosophical beliefs and traditions that underlie different research approaches, as in the essays by Nolen and Talbert (2011), Sun andPan 2011, andMarley andLevin (2011). Several of the articles take detailed looks at particular quantitative approaches to research and the conditions, within these approaches, that best lend themselves to prescriptive statements, such as the discussion of Martin (2011) of structural equation modeling, the discussion of O'Connell and Gray (2011) of logistic regression models, and consideration of Brown and Wilson (2011) of measurement issues in cognitive assessment.…”