As any other statistical methodology, AGRIS requires adequate testing to ensure that it is statistically sound, easy to implement and able to address the ever increasing data demand.
Accordingly, the GO in partnership with the Ghana Statistical Service (GSS) and the Ministry of Food and Agriculture (MoFA) has been undertaking a number of activities since December 2016.
The first part of this activity covered cognitive tests, field experiments and testing the generic AGRIS questionnaires.
The objectives of the cognitive tests were to probe problematic questions/sections and refine the questionnaires before the actual field tests. This has helped to ensure that respondents understand questions, concepts or tasks in a consistent way and as intended. Moreover, it helped to understand if respondents feel comfortable and are willing to provide pertinent and reliable answers. During the cognitive test, five AGRIS questionnaires were tested through in-depth cognitive interviews on a small group of respondents (about 20 individuals per questionnaire).
In order to refine the measurement approach used for collecting time worked on the farm, a field experiment was conducted on the labour module of the AGRIS questionnaire. In this experiment, labour measurement was conducted in three different ways and results were compared. The other field experiment undertaken, involved testing the respondent’s declaration against objective measurements on production and post-harvest losses. The experiment concludes that even if countries adopt objective measurements taking into consideration the time and costs required for objective measurements, farmers’ declarations can be implemented for AGRIS.
The generic questionnaire field-testing that was administered in July 2017 on 128 selected agricultural holdings (including household and non-household sectors) was part of the critical work undertaken by the GO in collaboration with GSS and MoFA. This test aims to validate all the AGRIS generic questionnaires and strongly contributed towards bringing the AGRIS questionnaires and the methodological notes into their final shape. Moreover, this test helped to fine-tune the CAPI questionnaires and made it possible to understand the workload and the time required to administer each of the AGRIS questionnaires.
Once the generic questionnaires have been finalized, taking into consideration all the inputs from the above mentioned field tests and experiments, the GO in collaboration with GSS and MoFA has launched the second phase of its exercise, which involves customizing the AGRIS methodology to the Ghanaian context. This will serve to test how countries will undertake the customization process and make AGRIS part of their regular survey. More specifically, it will assist the Ghanaian government on how to use and adapt the AGRIS generic methodology to their context and, through this exercise, build their capacity scaling it up to the national level. The customized AGRIS Ghana pilot survey will be using the integrated survey questionnaires and it is designed to produce reliable estimates for the four selected districts. While the actual pilot survey is scheduled for January 2018 and the survey results are expected in April 2018, all the preparatory activities have been underway since November 2017.