In the paper we consider the most common inaccuracies and mistakes in the use of statistical methods that are applied in educological investigations. Some inaccuracies are related to the definition of population or sample. In several papers that we have considered, it is not clear also how a sample is related to a population of interest. The samples that are used for surveys are sometimes not representative and therefore the statistical inferences which are based on the data of such samples are in doubt. The results of educological investigations are also sometimes doubtful because of the assumptions of some statistical methods that are often ignored by the researchers. The other inaccuracies that we have analyzed are related to formulation of statistical inferences and unit/item nonresponse.
The recommendations how to avoid the mentioned inaccuracies and mistakes are also proposed, as well as the example of real statistical survey in the real educology research. That is, we consider in the selected population the question “What is the proportion of students that intend to become a teacher”, the answer to which is in recent time absolutely controversially assessed from one side by the employees of the Ministry of Education and Science that administer financing of universities, and from the other side by educology specialists who perform research of the mentioned type.
The survey results show that slightly more than 64 percent of the respondents intend to become a teacher or already are working as a teacher. This is a very joyful tendency.
Almost 90 percent of the respondents are pleased that they are studying at the faculty of Mathematics and Informatics, Vilnius Pedagogical University. This is the fact that shows a high quality of studies, as well as the trust in the advantages of the chosen specialty that may be useful adapting to the changeable today’s economic and social conditions.