RESUMO
Addressing the common problems that researchers encounter when designing and analysing animal experiments will improve the reliability of in vivo research. In this article, the Experimental Design Assistant (EDA) is introduced. The EDA is a web-based tool that guides the in vivo researcher through the experimental design and analysis process, providing automated feedback on the proposed design and generating a graphical summary that aids communication with colleagues, funders, regulatory authorities, and the wider scientific community. It will have an important role in addressing causes of irreproducibility.
Assuntos
Internet , Projetos de Pesquisa , Software , RetroalimentaçãoRESUMO
The need to improve reproducibility and reliability of animal experiments has led some journals to increase the stringency of the criteria that must be satisfied before manuscripts can be considered suitable for publication. In this article we give advice on experimental design, including minimum group sizes, calculating statistical power and avoiding pseudo-replication, which can improve reproducibility. We also give advice on normalisation, transformations, the gateway analysis of variance strategy and the use of p-values and confidence intervals. Applying all these statistical procedures correctly will strengthen the validity of the conclusions. We discuss how InVivoStat, a free-to-use statistical software package, which was designed for life scientists, especially animal researchers, can be used to help with these principles.
Assuntos
Experimentação Animal/estatística & dados numéricos , Projetos de Pesquisa/estatística & dados numéricos , Animais , Reprodutibilidade dos TestesRESUMO
InVivoStat is a free-to-use statistical software package for analysis of data generated from animal experiments. The package is designed specifically for researchers in the behavioural sciences, where exploiting the experimental design is crucial for reliable statistical analyses. This paper compares the analysis of three experiments conducted using InVivoStat with other widely used statistical packages: SPSS (V19), PRISM (V5), UniStat (V5.6) and Statistica (V9). We show that InVivoStat provides results that are similar to those from the other packages and, in some cases, are more advanced. This investigation provides evidence of further validation of InVivoStat and should strengthen users' confidence in this new software package.