Is there a ''file drawer problem'' in biological research?