Abstract:

The article deals with the didactics of teaching statistics for non-statisticians, which is standard practice at university faculties of economics. The analysis is based on the authors' long-term personal experience and also relies on the results of surveys conducted among students over the past 10 years. It is based on hundreds of student responses, which allows for data analysis based on the respondents' attitudes. The experience comes from the University of Economics in Prague, where hundreds of students undergo statistics for non-statisticians every year.

The aim of the article is to reveal bottlenecks in teaching from the data set that can lead to weaker study results or even study failure. The basic statistics course is one of the more difficult subjects with worse study results. Another ambition of the authors is to also point out changes in the methodology of teaching statistics, which have been brought about by the rapid development of software in teaching over the period of approximately 10 years under review.

It turns out that there are about three to four bottlenecks in statistics teaching, which can gradually discourage students from intensive teaching during the semester, cause them to feel difficult and ultimately lead to poorer academic results, possibly even failing the exam. These bottlenecks are:
1) understanding and use of descriptive variability,
2) an excess of probability distributions (especially theoretical ones),
3) an excess of hypothesis tests and
4) inductive (inferential) understanding of regression and correlation.

The text – based on the data and experience obtained – mainly formulates recommendations in the conclusions for improving teaching didactics. It suggests specific measures that can help students and lecturers in the identified bottlenecks to create better didactic comfort during the semester.

Keywords:

Teaching statistics for non-statisticians, universities of economics, lecture didactics, teaching bottlenecks, data analysis from student surveys.