كليدواژه :
تحليل همبستگي , Common Statistical Mistakes , رياضيات , حجم نمونه , آناليز وارياش
چكيده لاتين :
Today, there are a few researches that do not benefit from statistical techniques and paradigm. In fact, the use of statistics for exploration, analysis, inference, and conclusion in any field of study is inevitable. Statisticians are encouraged to develop and introduce new creative methods in order to meet high demand for powerful and appropriate techniques. Despite of remarkable advances in statistics, more and more common misuses are still evident.
In Iran, this issue is dominated by inadequate (both in hardware and software) statistics. In other words, the widespread gap between the available resources and application requirements on one hand, and lack of a suitable statistical consultancy (appropriate and on time, at the beginning of a research) from highly qualified (expert) statisticians on the other hand, have lead to obvious mistakes.
In some cases, there are some critical misinterpretations and mistakes so that one canʹt use the findings, for example, as a case study. Non-clear definition of statistical population, vague sampling methods and techniques, unknown unit of analysis, biased estimations and low precision are the examples of some pitfalls in the phase of data generation. Poor description of research and analysis methods, mismatch between the related statistical approach and the research objectives, inconsideration of assumption, and ignoring premises are the examples of very common statistical pitfalls in the phase of statistical analysis. This article is intended to recommend a number of cautions to be taken when using statistics. We will discuss on some real (actual) cases to uncover and clarify mistakes and statistical pitfalls.
