QUANTITATIVE-ANALYSIS OF RECOVERY CURVES

It is assumed that recovery curves may be produced either by a dynamic process caused by a feedback mechanism or by a linear compensation process which is independent of previous compensation. Based on this assumption, 3 types of recovery functions are derived. Two behavior functions correspond to each of these types of recovery curves, one starting at a low level and the other at a high level. For each of the resulting 6 types of behavior curves, examples of real data have been reported in literature. Elementary parameter estimates are provided and substantial interpretations of the parameters are given. Behavior measures are derived for the data from an illustrative example, and 1 of the 6 behavior curves is fitted to this data. By additionally taking into account a physical repair function which is assumed to cause a certain delay in recovery, it is shown that an even better fit can be achieved. © 1990.