Precognitive Remote Perception: Replication of Remote Viewing

The following brief description of the Princeton Engineering Anomalies Research (PEAR) Remote Perception program has been prepared at the invitation of the Editor, in order to augment this special report section of the Journal with information about another substantial database of experiments relevant to those of SRI and SAIC. Given Utts' attention to the importance of replication (Section 3.4), and Hyman's challenge of interlaboratory consistency (Point #3 of his Introduction and Point #2 of his "Suggestions for Future Research"), we submit that the PEAR program has obtained the largest extant body of experimental data that meets their criteria for interlaboratory replication. In point of fact, both the PEAR remote perception program, and the prior studies of Dunne and Bisaha on which it was originally based, were undertaken as formal replications of the SRI experiments of Puthoff and Targ. Although the PEAR program has accumulated several hundred experimental trials, its primary goal has been to develop a sophisticated analytical judging methodology to replace the human judging process, and thereby to facilitate more precise quantitative assessment of results and their correlation with various experimental parameters. In our basic procedure, the "free response" of the percipient is encoded using a list of 30 binary descriptor questions, allowing algorithmic comparison with the target, similarly encoded by the agent at the scene. For randomly assigned targets, further comparison can be made with an encoding by the person who prepared the target pool. The analysis proceeds by constructing a square matrix of scores calculated by comparing each perception against all targets in the given dataset. The properly matched trials (on the main diagonal of the matrix) can be assigned statistical merit by comparison with the distribution of off-diagonal, mismatched scores, which has sufficiently Gaussian characteristics to allow robust parametric statistical tests. Beyond the primary experimental question of the degree of anomalous acquisition of information, several other issues have been explored, among them the correlation of analytical and human judge scores, the efficacy of different scoring algorithms and descriptor sets, ex post facto vs. participant encoded descriptions, agent chosen versus randomly assigned targets, single vs. multi-