You are here
The coefficient of variation (CV) is used to measure variability of a metric. CV is popular because it can be used to compare distributions with different units of measurement.
How to operationalize the metric
Method of data collection and data needed to compute the method:
In this case we will use CV to assess the variability of net income or gross margin. If a number of technologies are tested for profitability in different locations or multiple interventions on plots for a single technology, then net incomes or gross margins need to be calculated using the approach discussed earlier. Once these data on the net income or gross margin are obtained, the coefficient of variation can be calculated.
Unit of analysis:
The coefficient of variation is unitless but takes on values between zero and
, where N is the number of non-negative values of the metric. The coefficient of variation can be computed then as;
Limitations regarding estimating and interpreting:
Coefficient of variation is only meaningful when there are non-negative profits in the sample. Negative values might be possible due to crop failure or high costs of inputs. Caution must be taken in calculating a CV where negative values are observed because the measure is not valid for cross-item measurement unless other statistical measures are taken.
Variability of profitability is best measured directly through farmers’ actual production, input costs, and output prices. However, it is also feasible to estimate variability in profits using production variability and price volatility values. Assumptions would need to be made about when farmers sell if their output prices vary seasonally (such as for maize in southern Africa). Thresholds and critical values for variability may relate to food security, poverty lines or similar objectives. Resilience of profitability to shocks (e.g., production, price, political) should be considered as a separate indicator.