The reproducibility of scientific research has gained attention in recent years as the call for transparency has increased. A number of terms, such as reproducibility, replicability, repeatability and reliability, are often used with little understanding of what they actually mean. A recent article, written by Goodman et al, has tried to address this issue and provide definitions of these words.
The authors outline the confusion that arises due to a lack of clarity regarding the usage of such words and the fact that in the literature they have been used interchangeably. Goodman et al suggest “new terminology to distinguish between the various interpretations of reproducibility”. They propose breaking down the topic into methods reproducibility, results reproducibility and inferential reproducibility. The ability to repeat an experiment using the same materials and methods and obtain the same results comes under methods reproducibility. Results reproducibility is the generation of new supporting data using the same methods and inferential reproducibility encompasses the making of similar conclusions after repeating the work or reanalysing the data. The authors have tried to provide less ambiguous definitions in the hope that the concept of reproducibility will be clearer and its relationship to scientific truth better understood.
The field of cancer research has gone one-step further to examine the issue of reproducibility with the creation of the Reproducibility Project: Cancer Biology (RP:CB), a partnership between Science Exchange and Center for Open Science. The RP:CB is looking to replicate key results from prestigious papers published between 2010–2012. They plan to publish the study design, protocol, sample size and proposed analysis prior to replicating any research. The results will then be published separately. The RP:CB want the data to be independently analysed by individuals to remove bias, with them providing “their subjective opinion of whether the original result was successfully replicated”. In order to do this, they are currently looking to recruit people willing to review the results of the studies, as described in a recent blog from BioMed Central.