Replicate or lie

Andren and colleagues (2008) recently published a paper reminding us of good scientific practice, taught and learnt during our early years of research, but frequently forgotten. This article has a similar objective. It is highly likely that the majority of microbial ecologists have taken a basic course in statistics. Unfortunately, the basic principles of statistical analysis and its significance and necessity are frequently ignored. My own exposure to this problem, and my suffering of the consequences, arise through reading, reviewing and editing articles on microbial diversity. The suffering is greatest with articles utilizing ‘expensive’ or cutting edge techniques, such as analysis of clone library sequences, microarray analysis and (increasingly) high-throughput sequencing. The problems, however, are more widespread and exist beyond these techniques and beyond studies of microbial diversity.