For the reason that this critical essay has such a title it requires a subjective choice of topic, so I have decided to choose the following: I would like to touch upon one of the most important parts of any psychological research and psychology in general — mathematics, namely mathematical statistics.
As McConnell said: «Statistics are primarily a way of thinking, and you only need to have a little common sense and know the basic mathematics to apply it[AL1] ». There is some truth in this sentence, but the reality is much more complex. Mathematical methods are the basis of psychological research. Сertainly this topic is not very interesting, but the fact that it is useful – without a shadow of doubt: there would be no modern psychology without mathematics. Let's briefly describe why a psychologist needs mathematical statistics. Adequate application of statistics allows the following (Ермолаев-Томин, 2017): to prove the correctness and validity of psychologically methodical techniques and methods; to justify experimental plans; to generalize data obtained in experiments; to find various kinds of dependencies between experimental data; to identify the presence of significant differences between groups of subjects, to identify the relationship between the studied variables; to make statistical predictions; to avoid many logical and substantive errors. As we can see, statistics offer unlimited possibilities for studying. Of course, there are many psychological theories without any mathematics (e.g. activity theory, psychoanalysis), but this is a high level of abstraction, a certain level of methodology; without using mathematics, we cannot go to the level of specific methods (Юдин, 1977).
What is observed in psychology at the present time? According to the most conservative estimates, more than 1/3 of published scientific works contain mathematical errors. And Stanton Glantz (2002) notes that statistical errors occur in about half of the articles. This is a big problem, and the problem not just for psychology, which leads to biased research results. How can this be fixed? Surely, the first thing needed is to conduct a strict review of the statistical part for all journals. Some journals do this, but great majority don't pay much attention to statistics. The second, in my opinion, is to educate people who have decided to connect their lives with psychology (in fact not only with it), to be careful and precise with mathematical statistics. And the second, I think, is much more important.
The next problem appearing from these arguments is that many postulates, laws that are existing in psychology for a long time, are not confirmed or disproved by mathematical methods. And this is also a big problem. Consider soviet psychology. Such geniuses as Vygotsky, Leontiev, and Luria use statistics too little. For example, if you skim through Luria's works (1963), you will find that there are almost no statistics. Therefore, a lot of positions in psychology need to be re-examined using mathematics.
There are also a huge number of psychodiagnostic[AL2] methods that are used in practice very often, but whether the question on they their work effectivenessit remains in questionunanswered. Often the explanation of a particular result comes only from the thoughts of the a psychologist. As an instance, projective test such as a non-existent animal: this method has many stamps (for example, that teeth are a sign of aggression and so on), but after evaluating the validity of this method using mathematical statistics it was found that this method is absolutely invalid (Щербатых, Ермоленко, 2016)
However, with the development of science and technology, there is another problem – to choose program to use for data processing. Psychologists most often use SPSS, Statistica, and JASP. But currently there are no scripts for calculating certain correlation coefficients, criteria for differences, and etc. Furthermore, there are some errors, so you can't conduct global research or meta-analysis with them. As a solution to this problem, scientists need to find good programmers, but it is noteworthy that there is a program like RStudio, where you can write scripts yourself (Verzani, 2011), so a psychologist who is going to conduct a complex study should learn programming and then calmly use the statistical methods that he or she needs to calculate.
Thus, mathematical statistics are a necessary base for psychological research, the base without which the subsequent analysis of complex psychological problems is impossible.
References:
1. Glantz, S. A. (2002). Primer of biostatistics.
2. Verzani, J. (2011). Getting started with RStudio. "O'Reilly Media, Inc.".
3. Ермолаев-Томин, О. Ю. (2017). Математические методы в психологии: учебник для академического бакалавриата. М.: Издательство Юрайт.
4. Лурия, А. Р. (1963). Мозг человека и психические процессы: Нейро-психологические исследования N 1. Изд-во АПН РСФСР.
5. Щербатых, Ю. В., & Ермоленко, П. И. (2016). Оценка валидности проективного теста «Рисунок несуществующего животного». Вестник по педагогике и психологии Южной Сибири, (4).
6. Юдин, Э. Г. (1997). Методология науки. Системность. Деятельность. М.: Эдиториал УРСС, 445.
[AL1]Page and year
[AL2]assessment