The construction of limiting distributions of response probabilities
G. S. Marliss, J. R. McGregor · Journal of Applied Probability · 1971
As stated elsewhere [9], the two experimenter-controlled events learning model of Bush and Mosteller [1] may be described by a Markov process X 0,X 1, ···satisfying the following conditions. (i) Xo has an arbitrary distribution on (0, 1). (ii) If Xn is given, then with probability π1 and with probability π2 (ii)