In the context of a medium-term study designed to integrate the simulation of different types and processes of learning—such as classical, operant, and some cognitive types—one must start with other more elementary ones that are facilitators of the more complex types and processes. Of special interest is habituation, owing to the filtering out of irrelevant stimuli, which means that the simulated agent does not have to respond to them. This paper presents two difference functions constructed to computationally simulate the characteristics that define habituation. The behavior of these functions is described, as are differences arising from stimulus intensity and interstimulus intervals. Results are compared with existing empirical data.

The one-year reliability of auditory evoked potential (AEP) amplitude/stimulus intensity slopes (obtained for single channels and with dipole source analysis) and the stability and criterion validity of alternative measures to the conventional AEPs augmenting/reducing (A/R) linear slope were explored in a sample of 21 psychology students. AEPs were obtained from two sequences of 288 Ss of different intensities (60, 80, 90, and 110 dB SPL), presented in four blocks of 72 tones of each intensity or in a pseudorandomized order. The Zuckerman's Sensation Seeking Scale, which has been associated to the A/R slope, was used as a criterion measure. The results supported the one-year reliability of N1P2 slopes obtained in single channels (especially for Fz and Cz and for the random presentation of stimuli) and for dipoles obtained with brain electric source analysis (especially for the tangential dipoles). Neither of the A/R measures was significantly related to the criterion. These results support the stability of AEPs A/R, but throw serious doubts on the use of this psychophysiological phenomenon as a correlate of personality dimensions such as sensation seeking.