2 – Assessing Available Information

Access to information is no longer a limiting factor in scientific progress. All the data generated in a study should be made available for scrutiny. Internet technology has made space limitations imposed by print journals obsolete. Data sets that are prohibitively large to be included in print form, such as those arising from DNA microarray studies, are being made available via Internet web sites. Similarly, comprehensive data sets should be made available for all studies, including clinical trials. All raw data and calculations should be made available, including those for individual experiments that “failed to work” for whatever reason. The practice of reporting data only from the subset of experiments that yielded data consistent with the hypothesis should be abandoned. The results of all studies, even ones in which the null hypothesis was supported (negative studies), should be reported. Such negative results, assuming the study was conducted and interpreted correctly, provide valid and potentially important information. This should be a requirement for all studies conducted with public funds, even incorrectly conducted studies. This information would be useful in making funding decisions. Regardless of how scientific information is communicated, its critical evaluation always will remain central to the scientific enterprise. This skill must be learned and practiced by its participants.