Extraversion and ESP performance: a meta-analysis and a new confirmation - extrasensory perception

Journal of Parapsychology, The, Sept, 1998 by Charles Honorton, Diane C. Ferrari, Daryl J. Bem

Editor's Note: This paper was presented originally at the 1990 convention of the Parapsychological Association. Ferrari and Bem gave permission to publish the paper in the Journal. Honorton is deceased.

The relationship between ESP performance and individual differences in psychological traits has been explored in many studies since the 1940s. Extraversion is one of the most frequently studied trait variables and three narrative reviews of the ESP/extraversion literature have concluded that ESP performance is positively related to extraversion (Eysenck, 1967; Palmer, 1977; Sargent, 1981). We present a meta-analysis of the extraversion/ESP literature. The purpose of the meta-analysis is to (1) update earlier narrative reviews of the ESP/extraversion relationship, (2) estimate the magnitude of the relationship, (3) assess potential threats to validity, and (4) identify procedural and other variables that moderate the relationship. We will then present a new confirmation of the ESP/extraversion relationship and compare its magnitude to that estimated from the meta-analysis.

THE META-ANALYSIS

Previous parapsychological meta-analyses have focused on evidence for psi functioning in such research domains as ganzfeld communication (Honorton, 1985), precognition (Honorton & Ferrari, 1989), and studies of the impact of conscious intention on random number generators (Radin & Nelson, 1989). In such cases the effect size index is based on the proportion of hits and the unit of analysis is the trial. In the present case, we are interested in the relationship between psi performance and a predictor variable, extraversion, rather than overall psi performance. The effect size index is the correlation coefficient between the two variables and the subject is the unit of analysis.

METHOD

Retrieval of Studies

We attempted to retrieve all English-language studies of the relationship between extraversion and performance in experimental ESP tasks. The source of studies includes the bibliographies of three narrative reviews (Eysenck, 1967; Palmer, 1977; Sargent, 1981) and inspection of the principal English-language outlets for publication of parapsychological research, including the Journal of Parapsychology, Journal of the American Society for Psychical Research, Journal of the Society for Psychical Research, and Research in Parapsychology. In addition, we conducted a computer search of Psychological Abstracts using the keywords "extraversion," "extroversion," "introversion," "intraversion," "sociability," and "outgoing."

Investigator Definition

For the purpose of the meta-analysis, we defined independent investigators as investigators who have not worked with other investigators in the data base. In cases of multiple authorship, studies are identified by the senior author of the earliest publication in the data base. For example, studies by Kanthamani (1966), Kanthamani and Rao (1972) and Krishna and Rao (1981) are all identified as Kanthamani studies. While Kanthamani and Krishna did not work together, they share Rao as a co-author and are therefore considered to represent a single investigator set. Similarly, studies by Humphrey (1945, 1951) and by Nicol and Humphrey (1953, 1955) are all identified as Humphrey studies. Laboratory affiliation was not used to identify independent investigators since several investigators worked in different laboratories and several generations of investigators worked in one laboratory.

Procedural Features

Besides bibliographical data identifying the investigator, publication source and date, we coded various procedural, sampling, and statistical features for each study. The procedural features include the type of ESP task (forced-choice or free-response), test setting (individual or group testing), ESP mode (telepathy, clairvoyance, precognition, mixed), and amount of feedback in the ESP task. Procedural features relevant to assessing research quality are described in the following section. Sampling and statistical features coded include the number of subjects, the subject population used, the instrument used for measuring extraversion, and the inferential statistics reported for testing the relationship between ESP performance and extraversion.

Criteria for Assessing Research Quality

Methodological variables were coded in terms of procedural descriptions (or their absence) in the research reports. This approach was used in an earlier meta-analysis of the ESP ganzfeld domain (Honorton, 1985), resulting in study quality ratings that were generally in agreement ([r.sub.26] = .77, p = [10.sup.-6]) with independent "flaw" ratings by an outside critic (Hyman, 1985). Two sets of criteria were used. One set assessed threats to the validity of the ESP measure. The other set assessed threats to the validity of the relationship between the ESP and extraversion measures.

The ESP quality analysis includes four criteria. One point was given (or withheld) for each of the following:

Control against sensory leakage. Credit was given to GESP (telepathy) studies if the report specified that each of the following criteria were met: sender and receiver were located in separate rooms, the test situation prohibited auditory or other cues from sender to receiver, the sender and receiver were monitored by experimenters, and in free-response studies involving subject judging, duplicate target sets were employed. Credit was given to clairvoyance studies if the report specified that cues from the targets were prohibited by means of physical distance, screens, or opaque packaging. Precognition studies were considered to be immune to sensory leakage problems.