Inter-rater Reliability of Participatory Evaluation Data - Brachiaria Grasses in Nyandarua
17 Pages Posted: 21 Nov 2018
Date Written: October 25, 2018
This study investigates the place of various physical attributes of 8 different Brachiaria grasses and two common grass species used by farmer — Napier and Rhodes grass. A list of 5 attributes including the ability of the grass to cover the ground, hairiness, harvesting height, pest and disease resistance and forage amount were used to evaluate these grasses in order to choose the best performers. Using this list of attributes, farmers were asked to evaluate these grasses replicated in 4 plots at the KALRO DRI Sub Centre; Ol Joro Orok. The data generated from the evaluation was subjected to a consistency check by use of the intraclass correlation coefficient. Results imply that for some less visible traits such as resistance to biotic stresses (pests and diseases), the evaluators can fail the consistency test. On the other hand, ratings for traits such as plant cover and height, which are easily observable were consistent. The discussion points to the need for careful use of methods and the need to perform reliability checks when using such participatory evaluation approaches. Whereas the paper does not provide a clinical prognosis, it provides a reminder to procedures which we think are rarely reported of are largely ignored in such participatory evaluations.
Keywords: Brachiaria, Participatory Evaluation, Rating Data
JEL Classification: Q16, B40
Suggested Citation: Suggested Citation