Sir

John Kirkland's review of Toward a Global PhD? by M. Nerad and M. Heggelund (Nature 454, 408; 2008) emphasizes the increasing popularity of doctoral degrees among students around the world. As a graduate students' representative, I have noticed a general and worrisome side effect of this trend in the three universities on two continents where I have been based.

It is striking how many early graduate students around the world complain about being unprepared for science itself, despite having passed their undergraduate degrees with ease. For graduate ecologists, for instance, it is crucial that the experimental design and statistics should be carefully thought through before the project is started. However, despite most universities' requiring a basic course in statistics, many graduates struggle to create properly designed experiments.

Ways around this problem could include practical experience on a project-based MSc course as a prerequisite for enrolling for a PhD, stemming the trend to shorten PhD courses and enhancing the quality of supervision.

Ideally, students should be motivated to learn for themselves and should have access to good library facilities so that they can acquire all the information they need before beginning data collection. But some will struggle to make the leap from an initial grasp of theory to the complexities of experimental design and analysis, and will suffer from this for the remainder of their PhD.

This large gap between undergraduate and graduate courses inevitably leads to money being wasted on badly designed projects that will never be published, and the demoralization of graduates who will never go on to become scientists.