Use Of Statistical Methods and Complexity of Data Analysis in Recent Research Publications in Basic Medical Sciences

Abstract

Objective: This study was carried out to evaluate the use of various statistical methods; statistical software; complexity of statistical analysis; and appropriateness of use of statistical methods in recent research publications of basic medical sciences.

Methods: Original articles published in Journals of Basic medical Sciences, i.e Anatomy, Physiology and Pharmacology subscribed by the central library of our institute (SMIMER, Surat) were taken into consideration were reviewed for statistical applications in the manuscript.

Results: Total 143 original articles were reviewed; out of them 89.51% (95% CI 84.49 - 94.53) had used Statistical techniques. Most frequently used statistical method was ANOVA (42.97%) and it’s use is lower in Anatomy compared to other two subjects. One out of seven publications used either t-test (independent and paired) or contingency table (chi-square and fisher exact) or nonparametric test. Confidence intervals were estimated in approximately one fifth of all research papers. Use of statistical software was much more in Physiology and Pharmacology compare to Anatomy. Basic methods of statistics were used 63 times while “modern Analysis” techniques were used 88 times. Only 10% (15 out of 143) research papers had mention the study design; 7% (10 out of 143) research papers had mentioned appropriate formula for calculation of sample size; 1.4% (2 out 143) had defined sampling technique. In 7.03% (9 out of 128) papers statistical measures were presented inappropriately.

Conclusion: Statistics methods are used widely in basic biomedical research. However, conceptual understanding of methodology and bio-statistics was lacking.