Comparative studies need to rely both on sound natural history data and on excellent statistical analysis

Evolutionary biologists aim to determine why a particular trait evolved.Detailed studies of single species can identify fitness benefits of specific traits and why those traits evolved, but these results might be rather species specific.Thus, to reach general conclusions about which ecological factors favour the evolution of a specific trait, comparative studies in several taxa are needed.Currently, comparative studies arising from databases summarizing many field studies are flourishing and provide important insights for our understanding of social evolution [1][2][3].Large, diverse datasets require detailed specialist knowledge in a wide variety of systems or taxa that is very rare, making it difficult to assess the quality of such databases.Referees of papers submitted to assess these comparative studies should review the part of the database for which they do have specialized knowledge.In some cases, studies with large databases could benefit from one additional referee that focuses only on the database itself, just as the views of specialized referees are often sought on aspects of statistical analysis or of animal ethics.In sum, the quality of a study depends on the quality of the data, and this is also the case for comparative studies.A recently published comparative study in Royal Society Open Science about cooperative breeding in mammals [4] used a database previously used in a comparative study of the evolution of monogamy in the same taxa [5].It had already been shown that for the third largest mammalian order, the Eulipotyphla, the information in this database is, to a large extent, inadequate [6].In the supplement to their earlier publication on the evolution of social monogamy [5], Lukas & Clutton-Brock state that the database is the result of a systemic literature review on 'mammalian species . . .collecting information from the primary literature . . .