Statistical refereeing for the New Zealand Journal of Agricultural Research, the New Zealand Journal of Experimental Agriculture, and the New Zealand Journal of Crop and Horticultural Science

摘要

Summary information is presented on faults that have been evident in papers that have been refereed statistically since 1983. Most problems that are identified when papers are refereed, whether by a biological or by a statistical referee, turn on the quality and interpretation of the data. A common fault is inadequate explanation of the experimental design. Other faults were: inappropriate use of correlations; failure to make coarse checks on statistical assumptions; the confusion of different strata of variation; and use of individual significance tests where a trend or response surface analysis was required.