THE EFFECTS OF STRATEGY (“WILLING” VERSUS ABSORPTION) AND FEEDBACK (IMMEDIATE VERSUS DELAYED) ON PERFORMANCE AT A PK TASK BY CHRIS A. ROE & NICOLA J. HOLT 1 ABSTRACT: In recent work to evaluate the sender’s role in successful ganzfeld GESP experiments, we have used a random number generator (RNG) as a “virtual receiver” in a ganzfeld-like experiment. During the sending period, statements are “selected” from a pool of items to give an “RNG mentation” to be used by an independent judge. After early success in demonstrating the basic effect, later work has explored the effects of the lability of the target selection method (random number table, pseudorandom process, and live RNG) and of the participant (high, intermediate, or low, based on a composite measure) and found a predicted interaction effect between these factors. The present study was designed to confirm that finding and extend it by exploring a putative interaction effect between sending strategy (active/willing versus passive/absorbed) and feedback type (delayed versus immediate). Forty participants generated virtual readings consisting of 24 statements, 8 from each of the 3 selection methods. A significant interaction was found between target lability and sender lability, replicating our earlier effect. Although the interaction effect between sending strategy and feedback type was nonsignificant, a predicted significant effect of feedback for participants in the willing strategy condition was found. Many ESP experiments adopt a telepathy design in which psi is conceived as a dyadic interaction between one person (the sender) who is aware of some randomly selected target information and one person (the receiver) who is unaware of that information by normal communication channels. The sender’s task is to attempt to convey target information psychically while the receiver adopts a relaxed, passive state that might be suffciently labile to allow psi-mediated information to come to conscious awareness. In this paper we describe the fourth study in a series that uses a novel protocol to explore more directly the contribution that the sender might make in such a dyad. This protocol has utilised a random number generator (RNG) to act as a “virtual receiver” that might be analogous to a human receiver in providing a fuid, random system by which to select impressions that could be related to target information but might avoid some of the diffculties of working with complex and idiosyncratic human systems (see Roe, Holt, & Simmonds, 2003).  This research was funded by the Bial Foundation. We would like to gratefully acknowledge this support. An earlier version of this paper was presented at the Parapsychological Association 49th annual convention.