The law of initial values: some further considerations.

Several methodological issues pertaining to the "law of initial values" (LIV) are explored. It is indicated that change scores, not final levels, are the appropriate measure for testing if the LIV applies, and that normalization of data is unnecessary. Heart-rate responses of infants to pacification were found to adhere to the law despite the inverse nature of the stimulus, suggesting the importance of considering behavioral state independently of sequence of measurement. Tests of the applicability of the LIV to group data revealed that spurious confirmation may result from operation of the law within the individuals comprising the group, due to differences in behavioral state. Techniques for mitigating this effect are considered. The role of time parameters is examined. The responses of adult subjects to sustained psychological stress indicates that the LIV may apply to prolonged as wall as transient physiological responses.