A current widespread procedure relies on the use of a reactivity meter for differential control rod measurements in nuclear reactors. From an observational point of view, the reactivity meter will measure the correct core reactivity if the input to the meter is proportional to the neutron amplitude function. A single detector reading is generally proportional to the amplitude function only if the point kinetics approximation is valid, that is, if the spatial flux distribution is invariant in time. Results of numerical simulations show that small spatial deviations can lead to large errors in reactivity determinations in current thermal-power reactor designs. In going from the use of single point to single long detector schemes, errors in reactivity determination are decreased but not eliminated and are very dependent on control rod location detector position. An integral approach has been developed and evaluated for a wide variety of core conditions for the reconstruction of the amplitude function from a set of detector readings. The scheme developed uses only observed detector readings and results in a combination of improved, consistent accuracy and ease of data handling.