Latest experiment by the Science Lab of Mozilla is all set to explore an interface between peer review related to scientific articles including codes and the code review of software engineer.
There is a widespread increase in overall research articles including codes that are utilized for manipulating data, in an era where research is getting highly dependent on a number of computational methods. A large number of people entrusted to produce such codes are highly trained and skilled computational scientists. Others engaged in the service are expert biologists who have mastered the process of coding effectively.
At recent times, a number of journals, including that of PLOS Computational Biology, are in the process of publishing innumerable articles on Software interfaces. However, the purpose of such publication aims towards fruitful peer reviewing of related software. It is proved that in majority articles that peer reviewing of codes appear to be at cursory best role. The notion even stands true for other publications and researches including the ones with smaller coding pieces where focus on research is not applied. However, it is not yet proved if such factor really causes any problem afterwards when any other party tries replicating or building anything on a previously published work.
On the other hand, several difficulties arise while trying to build anything on an existing code that is previously published, even though it is available as a whole. This eventually gives rise to several important questions – Is formal reviewing of any code is necessary? If so, what should be the method of approach?
Science Lab of Mozilla is going to perform a series of reviews related to software, on a trial basis. Experienced volunteers of Mozilla engineers would engage in reviewing code snippets from the papers published on papers of PLOS Computational Biology throughout the month. The dedicated Science Lab would approach different writers of the papers after completing an initial review. This would help them to take part in the process of dialogue at a later stage.
According to experts taking part in the review process, the move would offer the following insights –
- Nature of involvement by a non-specialist in reviewing a scientific software and even the duration of need for an expert in the domain
- Nature of effort required to compare reviews of a wide variety of software, including reviewing various papers themselves
- To what extent do such reviews help any scientist
The results of this trial experimentation will be published in the form of an anonymous summary by the Science Lab. This would eventually nullify the effects of reviews upon the status of such a publication.
According to one of the spokesperson of the experimenting team, they encourage all authors to utilize the reviews that come on their way through a feature or tool of post-publication commenting. However, since this experimentation is just carried on a trial mission, authors would be under no pressure or obligation to strictly follow the publications of the journal, even after they would receive the review of their documented article on Mozilla.