Open AccessStatistical analysis of cytogenetic data
Author(s) -
Л. А. Атраментова
Publication year - 2021
Publication title -
faktori eksperimentalʹnoï evolûcìï organìzmìv
Language(s) - English
Resource type - Journals
eISSN - 2415-3826
pISSN - 2219-3782
DOI - 10.7124/feeo.v28.1391
Subject(s) - data set , statistics , statistical analysis , set (abstract data type) , computer science , confidence interval , statistical hypothesis testing , sample size determination , sample (material) , interval (graph theory) , statistical model , data mining , econometrics , algorithm , mathematics , chemistry , chromatography , combinatorics , programming language
Using the data obtained in a cytogenetic study as an example, we consider the typical errors that are made when performing statistical analysis. Widespread but flawed statistical analysis inevitably produces biased results and increases the likelihood of incorrect scientific conclusions. Errors occur due to not taking into account the study design and the structure of the analyzed data. The article shows how the numerical imbalance of the data set leads to a bias in the result. Using a dataset as an example, it explains how to balance the complex. It shows the advantage of presenting sample indicators with confidence intervals instead of statistical errors. Attention is drawn to the need to take into account the size of the analyzed shares when choosing a statistical method. It shows how the same data set can be analyzed in different ways depending on the purpose of the study. The algorithm of correct statistical analysis and the form of the tabular presentation of the results are described.
Keywords: data structure, numerically unbalanced complex, confidence interval.