Efficient Evaluation of Qualitative Traits in Multisite Trials

Trials of vegetable varieties can contribute valuable information about quality traits. In particular, growers need to know whether they will get the desired quality on their farm. There are two challenges to meet that simple goal. The first is that the variety trials need to be done at multiple locations to show whether the quality trait holds up in an environment similar to that of an interested grower. The second challenge is the quality traits generally have to be rated on a qualitative scale. To succeed in such a multi-site variety trial, the quality scale needs to be used consistently over time, and the raters at each location have to be trained to use the scale the same way. When multiple traits are to be evaluated, consistency is better achieved if the scale for each trait works the same way. For the Eastern Broccoli Project, we have developed a system that is working well. For each trait of interest we use a five-point scale, with objective criteria for each value. Furthermore for each trait a score of 3 represents marginal commercial quality, 4 represents desired commercial quality, and 5 is an even stronger expression of the characteristic. This simple consistency has allowed raters to work quickly and confidently, and to provide comparable scores in different locations. The resulting data have substantial power to assess wide adaptability, and suitability for a particular growing region.