“…The design was based upon the procedure of Peters (1935), as later adapted by Bugelski and Scharlock (i952)> with the major pomts of (1) three separate sessions of paired-assoaate leammg, with 48 hours mtervenmg (S was usually tested on Monday, Wednesday, and Fnday, with the actual hour of testmg ahoays identical on each day), (2) three hsts of tngrams to be associated A, B, and C Joseph F Rychlak facihtates leammg, then the B list should hasten leammg at A-C, as compared to an A-C hst which has not had the benefit of common B-mediators Bugelski and Scharlock (1952) were mterested m A-C leammg to determme if mediation vs no-mediation exerted a differential experimental effect Their hypothesis, which received only weak support, was that mediation did facihtate leammg at A-C, with S actmg as his own control (1 e, S leamed a sublist at A-C with no common B-mediator, though everythmg else was comparable as to practice, AV, etc ) In the present study, although each S agam acted as his own control, the interest was m the possible uncontrolled (in the earher work) infiuence of RV on mediation There was no A-C condition without a common B-mediator in the present study Five PPI tngrams on each of the three hsts (A, B, and C) were identified from the PPI as hked (L), five as disliked (D), and five as ambivalent (A) An S therefore leamed a list of 15 paired associates on each of the three expenmental days The 15 tngrams m each hst were equated for AV so that S leamed paired associates at A-B, B-C, and A-C of eqmvalent meamngfulness as defined by Archer's norms (i960), but, S's paired associates for each of these expenmental days were broken dovwi mto three sublists of five pairs diffenng m RV. He always leamed five pairs (subhsts) havmg the RV charactenstics of h-h, D-D, and A-A, subhsts were not leamed serially, however (see below)…”