A novel method for quantifying passive-avoidance behavior based on the exponential distribution of step-through latencies
Résumé
In the extensive literature dealing with the one-trial passive-avoidance task the data are usually represented by the median latency to respond. We propose here a novel representation and analysis of passive-avoidance data which is based on the observation that the complement of the cumulative distribution of step-through latencies (i.e., the fraction of animals remaining in the safe compartment) decays exponentially with time from the onset of the trial. A remarkably close fit of this complementary distribution is seen when the best-fitting straight line is drawn through the data points plotted on semilog coordinates. The slope of this line k, which we call "the step-through rate constant," (or alternatively, the T1/2 which is equal to 0.69/k) provides an accurate description of the population behavior as a whole in most cases. In view of the exponential distribution of passive-avoidance data this treatment appears to be more appropriate than the widely-used measures of central tendency, the median and mean. It is applicable to research on the effects of drugs on passive-avoidance memory, and probably appropriate to other behavioral paradigms and species.