1) A slightly re-phrased indicator can also be used for assessing animal production – for example, average number of litres of milk per cow per day or average number of eggs per chicken per day (always specify the exact animal breed).
2) Self-reporting faces common errors which decrease the precision of the provided information, including lack of knowledge (guessing), intentional over- and under-reporting (due to certain expectations or concerns), non-standard harvest units (even among individual households), poor quality of responses due to the respondent being tired, and many others. To minimize these risks, use the following tips:
> always pilot the questionnaire and if you record some difficulties, address it accordingly
> explain to the farmer, in an easy-to-understand way, why it is so important that his/her responses are precise
> clarify any unnecessary concerns or unrealistic expectations
> collect the data as soon after the harvest as possible (minimizing recall errors)
> if possible, use portable scales to measure the weight (in kg) of the unit the farmer reports in (e.g. 8 bags of wheat)
> if possible, go and measure the actual size of the land the survey is concerned about (in areas where farmers are less likely to know the land size, this must be a mandatory part of the data collector’s job)
3) Consider replacing interview-based self-reporting by supporting farmers in recording yields and other required information in forms provided by your organization (note – farmers should have a clear reason and incentive to do so).
4) If the survey team’s expertise allows, replace reported yield with crop cut measurement methods – harvesting samples of the given crop and using them to calculate the average yield. IndiKit currently does not have detailed guidance on crop cuts – if you can help us to prepare it, contact us please. We would appreciate your help!
5) Select the most suitable units of measurement based on focus groups discussions with farmers. Make sure that your staff is well-trained in determining the size of the land and the weight of the units reported by respondents.
6) Always ensure that you are comparing what is comparable – comparing the yields of a low-input crop variety with a variety requiring high financial and time investment might be of limited value (unless you are conducting an analysis of the production efficiency).
7) When interpreting the results, keep in mind that the indicator does not consider the amount of invested inputs (money, time, water…) and the sustainability of the production. Where relevant, use it together with the Agricultural Productivity indicator.
8) The measurements always need to be conducted in the same season of the year – do not compare, for example, vegetable yields in the dry and rainy season.
9) It is important that you also assess whether the respondent incurred any significant production loses (e.g. due to pests), otherwise the practice you promoted might have had a positive impact on his/her yield but your endline data will show no improvement.
10) European Commission's DEVCO recommends to use a similar indicator: "Average crop yield".