Page 40 - Contributed Paper Session (CPS) - Volume 1

CPS694 Ordak Michal
What education path should be chosen to increase the quality of
statistical analysis performed by the future medical community?
It is necessary to implement an appropriate method allowing students to
obtain practical knowledge necessary to conduct statistical analysis. Effective
education requires taking account of several factors. I have drawn the
following conclusions after conducting several thousand statistical
consultations. Thanks to this type of education, the knowledge of medical staff
regarding medical statistics has grown rapidly.
Let us suppose that a teacher deals with a specific statistical test, e.g. a
Student's t-test for independent samples. He or she explains when to apply this
test by referring to students' interests. Then, he or she demonstrates the
practical aspect of this statistical test based on a specific publication in a medical
journal. An example here can be a work published in The Nature Medicine in
which the authors compared two independent groups: one was continued on a
doxycycline diet and the other one was fed a normal-chow diet (Beckerman et
al., 2017). The next step is to perform this statistical test using a specific scheme.
It should include a step-by-step description of the methods of conducting the
test (and meeting the assumptions) and interpreting the results obtained. If
classes are carried out in this way, the knowledge of medical students will
increase by almost 100%. There is no stress associated with the rapid
implementation of individual steps necessary to conduct a test.
Another recommendation that would reduce problems associated with the
performance of statistical analyses by students for thesis, publications, etc., is
to conduct classes in medical statistics in later years of study. Due to a few-
year break in contact with statistics, a significantly larger number of students
do not even know how to get down to it. Conducting classes in medical
statistics in the last year of study would reduce the percentage of people
having problems with analysing statistical results. Students would have an
additional motivation that would encourage them to participate even more in
the classes. Thanks to this, the quality of conducted research necessary, for
example, to write a diploma thesis would significantly increase. Freshly
acquired knowledge could be used in a practical way in the further course of
study.
My last advice concerns the method of conducting a medical statistics
exam. The best way would be to use proper schemes during the exam
containing a step-by-step description of how individual statistical tests are
carried out. The exam should involve reading several research problems,
formulating hypotheses and selecting the appropriate statistical test. Then,
based on the scheme, the selected statistical test would be carried out and the
data would be interpreted. This would save the main problem of remembering
the great number of procedures necessary for conducting statistical analysis.

29 | I S I W S C 2 0 1 9