Common mistakes and misuse of statistics in agricultural experiments and guidelines on how to avoid them - A commentary

Abstract

The main objective of this commentary was to highlight some of the common mistakes and misuse of statistics in biological experimental designs, as well as possible ways of addressing these misconceptions. More often than not, researchers seek for guidance on their data analysis after data collection is completed, upon which the Biostatisticians detect poorly conceived experimental designs. Poorly conceived research often results in data analysis and/or results which do not correspond with experimental designs. The advent of ready to use statistical software seems to have both positive and negative benefits to agricultural research. Although it enables efficient data manipulation, processing and analysis, researchers with limited statistical experience are more likely to misuse statistical procedures, and this might lead to erroneous decisions. The validity, reliability and usability of biological research findings is dependent on adherence to statistical protocols and codes of conduct in the designing of experiments, analysis and interpretation of data as well as the conclusions made from the results. Therefore, researchers with limited statistical knowledge are encouraged to seek guidance from Biostatisticians right from the conception stage of their experiments. Researchers need to use relevant statistical analysis and interpretation protocols for their research results to be scientifically valid and usable.