Considerable research suggests that people who manifest, or are prone toward, certain forms of psychopathology report higher levels of magical or other types of paranormal beliefs. Meehl (cited in Eckblad & Chapman, 1983) noted that people who were prone toward manifesting schizophrenia were high in magical ideation, which he defined as a "belief, quasi-belief, or semi-serious entertainment of the possibility that events which, according to causal concepts of this culture, cannot have a causal relation to each other, might somehow nevertheless do so" (quote from Eckblad & Chapman, 1983, p. 215). The fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (1994) indicates that schizotypal personality disorder is characterized, in part, by "odd beliefs or magical thinking that influences behavior and is inconsistent with subcultural norms (e.g., superstitiousness, belief in clairvoyance, telepathy, or 'sixth sense'; in children and adolescents, bizarre fantasies or preoccupations)" (p. 645). The International Statistical Classification of Diseases and Related Health Problems (World Health Organization, 1992) contains a similar description.

Eckblad and Chapman (1983) developed a self-report scale to tap magical ideation. Items assessed belief in thought transmission, psychokinesis, precognition, astrology, and other dimensions. A group of students attending a large university in the United States completed the final scale (N = 1,512). Of these, high (n = 28) and low (n = 27) magical ideation participants were identified for further study. Based on clinical interviews, Eckblad and Chapman found that those scoring high in magical ideation reported more psychotic and psychotic-like symptoms, schizotypal experiences, and mood symptoms (depressed and manic) than those who scored low on magical ideation. George and Neufeld (1987), Chapman, Chapman, Kwapil, Eckblad, and Zinser (1994), and Kwapil et al. (1994) confirmed and extended Eckblad and Chapman's findings.

In Australia, Thalbourne and his colleagues have found a number of intriguing relations among levels of magical ideation, paranormal belief, and symptoms of psychopathology (e.g. Thalbourne, Bartemucci, Delin, Fox, & Nofi, 1997; Thalbourne & Delin, 1994; Thalbourne & French, 1995). For example, Thalbourne and Delin found that level of magical ideation and paranormal belief were positively correlated with extent of manic experience, depressive experience, manic-depressive experience, and hypomania. Other research has supported the view that paranormal and other types of magical beliefs are positively correlated with certain types of psychopathology - especially mood disorders, psychosis, and schizotypy - although the relationships are not always found or they may suggest a more complex view than described above (e.g., Irwin & Green, 1998-1999; Peltzer, 2002; Thalbourne, 1998). Furthermore, a relation between neuroticism and paranormal belief has not been found consistently (e.g., Zusne & Jones, 1989).

In summary, available research suggests there is a positive relation between extent of people's belief in magical causes (including belief in putative paranormal phenomena) and the level of certain symptoms associated with psychopathology (i.e., manic, hypomanic, depressive, and psychotic-like symptoms). People who experience higher levels of psychopathology are more likely to believe in the reality of causal connections between events that are not, according to generally accepted concepts, causally connected.

Study 1

In Study 1, we examined the relation between belief in magical causes and symptoms of psychopathology among a group of college students, using their responses on an "illusion of control" task as a measure of magical belief. This approach is different from those used to date, in that we exposed participants to an actual experimental task rather than asking them to report on their general beliefs.

In some situations, there is a contingency between responses and outcomes; outcomes are to some extent controllable. In contrast, in other situations, the probability of outcomes is the same irrespective of a person's responses; outcomes are uncontrollable. Langer (1975) defined an illusion of control as "an expectancy of a personal success probability inappropriately higher than the objective probability would warrant" (p. 313). She hypothesized that when factors normally associated with skill situations are present in uncontrollable situations, people will often perceive contingency between actions and outcomes when none objectively exists. Skill-related factors are those commonly associated with an increase in the probability of achieving outcomes. These factors include – but are not limited to – choice, stimulus or response familiarity, active involvement in the task, and competition (Presson & Benassi, 1996). The presence of skill-related factors makes uncontrollable outcomes appear to be controllable and, consequently, increases the likelihood that a person will approach the situation with a skill orientation.

Langer and Roth (1975) hypothesized that, because people want to see themselves as the cause of their successes (cf. deCharms, 1968), they attend to cues from the environment that support this notion. One such cue is a consistent pattern of task outcomes. According to Langer and Roth, people assume that if outcomes are controllable, one will be successful or fail rather consistently depending on one’s ability to perform the task. In contrast, one's performance on a chance task is characterized by a random series of outcomes. According to Langer and Roth, "the motivation to see events as controllable is so strong that the introduction of just one cue, a fairly consistent sequence of wins (with runs of no more than four in a row), is enough to induce an illusion of control" (p. 955). They found that when a string of successes on a coin-flip-prediction task occurred early in the outcome sequence, participants reported a higher number of successes than did those who received a string of an equal number of successes later in the task or no consistent string of successes.

We exposed participants to a psychokinesis task used in previous illusion-of- control research (Benassi, Sweeney, & Drevno, 1979). We sought to determine the relation between the extent of college students' level of depressive symptoms and their judgments on this psychokinesis task. We used this task for two rea-sons. First, a belief in personal psychokinetic ability is the type of dimension tapped in Eckblad and Chapman's (1983) magical ideation scale and involves a belief in causal connections (i.e., "mind over matter"). Second, previous researchers have implicated illusory control as an explanatory construct for belief in personal paranormal ability. Ayeroff and Abelson (1976), for example, demonstrated that belief in success on a mental telepathy task was influenced by the introduction of the types of skill-related factors described by Langer (1975). They found that participants who were permitted to choose the symbols they were to transmit in a mental telepathy task and/or who had some practice trials on the task reported higher levels of success than did participants who were not provided with these experiences (see also Benassi et al., 1979).

We exposed students to a task that has been shown to induce an illusion of control orientation (Benassi et al. 1979) and we also asked them to complete a magical ideation scale (Eckblad & Chapman, 1983). We analyzed the data using a principal components analysis and determined that judgments and ratings were well represented by a single factor. We formed two groups of students based on whether they were relatively low or high in depressive symptomatology and compared these two groups on a composite illusion-of-control factor score. Based on previous research that used general scales to measure magical ideation (and paranormal belief), we predicted that participants who reported greater depressive symptomatology would show greater belief in personal success on the psychokinesis task/higher levels of magical ideation.

Method

Participants Eighty-five female undergraduate psychology students enrolled in introductory psychology courses at a public university in the northeastern United States participated in the study as part of a course requirement. A large majority of the students were in their first year of college and most were 18 or 19 years of age.

Materials Participants completed a 46-item measure of depressive symptoms (Plutchik, Platman, Tilles, & Fieve, 1970). Each statement was a declarative sentence to which a participant answered "yes" or "no" (e.g., "I feel empty," "I feel sad"). The measure taps the present state of depressive symptoms, not a general tendency toward having such symptoms. Participants' total scores on the measure were calculated consistent with the protocol of Plutchik et al., with higher scores indicative of higher depressive symptomatology. The coefficient alpha for the sample was .87.

Participants completed the Magical Ideation Scale (Eckblad & Chapman, 1983). The scale consists of 30 statements each of which is responded to as "true" or "false" (e.g., "I think I could learn to read others’ minds if I wanted to," "I have felt that I might cause something to happen just by thinking about it."). Participants' scores on the measure were calculated consistent with Eckblad and Chapman's protocol, with higher scores indicative of higher magical ideation. The coefficient alpha of .81 for the present sample compares well with the value reported by Eckblad and Chapman for female college students (.85).

Apparatus We used a device called a die funnel in such a way that trial outcomes were hidden from a participant's view (see Benassi et al., 1979). Aparticipant sat at one end of the apparatus and tossed a die into the funnel. The experimenter retrieved the die at the other end, recorded the outcome of the toss, and returned the die to the participant.

Procedure A male experimenter tested each participant individually. Sessions were conducted in a quiet laboratory setting and lasted approximately 45 minutes. After signing an informed consent form, the participant completed the depression scale, the magical ideation scale, and several additional measures and tasks not relevant to the present study.

The experimenter next described to the participant that her task was to attempt to influence the outcome of a die toss. The experimenter showed her two dice, each of which had three sides painted red and three sides painted green. He asked her to select one of the two dice and, for each of the 20 trials, to select a color and for five seconds focus all her mental energy on making the target color land up at the end of the die funnel. The experimenter explained that he would keep track of the time and after five seconds he would tell her to throw the die into the funnel. The experimenter recorded the result of each trial, informed the participant of the outcome, and returned the die to her. The feedback was rigged so that the participant received 10 correct and 10 incorrect outcomes with seven correct responses occurring in the first half of the trials. After the 20 trials, the participant wrote the number of trials on which she believed the target color landed up and how much control she believed she had over the outcome of the trials. Participants reported control estimates on a scale ranging from 0 (no control) to 10 (complete control). Upon completion of the session, the experimenter debriefed the participant orally and in writing. Special efforts were made to make clear that the outcome feedback participants received was rigged and not an indication of how they actually performed on the task. The experimenter also went further to emphasize the purpose of rigging the feedback. Participants acknowledged their understanding of what the experimenter had done and none expressed annoyance or concern about the outcome feedback.

Results and Discussion

We analyzed judgments on the two dependent measures from the psychokinesis task and scores on the magical ideation scale using principal components factor analysis with varimax rotation. The solution produced a single factor, with an eigenvalue of 1.62 that accounted for 53.8% of the variance. The factor loadings for the die toss performance, control judgments, and magical ideation scores were .82, .73, and .64, respectively.

We transformed participants' depressive symptom scores to z-scores and then formed two groups. People whose score fell below zero were categorized as low in depressive symptomatology and those participants whose score was zero or above were categorized as high in depressive symptomatology. Illusion-of-control factor scores (M = 0; SD = 1) differed reliably as a function of the level of depressive symptoms, F(1, 83) = 5.02, p < .03, r = .24. Participants in the group that showed the higher level of depressive symptoms reported a higher degree of illusion of control (see Table 1). These results are consistent with Thalbourne and colleagues' (e.g., Thalbourne et al., 1997) view that a greater tendency toward psychopathology is associated with greater belief in paranormal phenomena. They are also consistent with the findings of Eckblad and Chapman (1983) and Chapman et al. (1994) that belief in magical causation - a type of illusory belief - is positively associated with depression, mania, hypomania, schizotypal, and psychoticlike symptoms. We found this relation among a nonclinical sample of college students.

Study 2

We conducted a second study with several purposes in mind. First, we wanted to replicate our finding from Study 1. Second, we wanted to create a measure of paranormal belief that was based solely on task-related judgments, not a composite that included scores on the magical ideation scale. Third, we sought to resolve what appears to be an inconsistency between the results that we found in Study 1 and the so-called "sadder but wiser" effect.

Alloy and Abramson (1979) investigated the link between illusion of control and depressive symptomatology. For example, Alloy and Abramson (1979, Experiment 2) compared the judgments of college students on a response-outcome contingency task. On each of a series of trials, a student either pressed or did not press a response key during a specified trial interval. The probability of an outcome occurring at the end of the interval was the same whether or not the student responded and was programmed to occur on 75% of the trials. Using a 0 (no control) to 100 (complete control) judgment-of-control scale, Alloy and Abramson found that students who scored in the depressed range on a measure of the severity of depressive symptoms reported significantly lower judgments of control than did nondepressed students on a task in which the programmed contingency between responses and outcomes was zero. That is, the depressed students did not demonstrate an illusion of control relative to nondepressed students, leading Alloy and Abramson to suggest that depressed individuals may be "sadder but wiser." Alloy, Abramson, and colleagues have reported that, whereas nondepressed participants' judgments on contingency tasks are sometimes accurate (e.g., when outcomes are response dependent) and sometimes inaccurate (e.g., when outcomes are response independent and occur at a high frequency), mildly depressed participants' judgments are consistently accurate relative to those of nondepressed participants (e.g., Alloy & Abramson, 1979; Alloy & Clements, 1992; Martin, Abramson, & Alloy, 1984). These findings have been replicated in a number of experiments (e.g., Benassi & Mahler, 1985), but not in others (e.g., Bryson, Doan, & Pasquali, 1984; but see Benassi & Belli, 1989).

The sadder-but-wiser effect reported by Alloy, Abramson, and others appears to be inconsistent with our finding in Study 1 that people scoring higher in depressive symptomatology showed greater illusory control than did those who scored lower on the depressive symptoms measure. In their studies, level of depressive symptoms was negatively associated with illusory control. Perhaps the key to reconciling these different findings lies in examining the type of task used by Alloy and Abramson and others - contingency detection - and the experimental task used in Study 1 to induce illusory paranormal beliefs. Study 2 specifically addressed this possibility.

In Study 2, we used tasks designed to tap precognition and psychokinesis beliefs. We also included a standard contingency judgment task that has been used to test predictions of the "sadder but wiser" hypothesis and that did not resemble a psi-type task. We predicted that, for the contingency judgment task, higher levels of depressive symptoms would be associated with lower illusory control. In contrast, we predicted that, for the precognition and psychokinesis tasks, higher levels of depressive symptoms would be related to higher illusory control. The former result would replicate findings reported by Alloy and Abramson (1979) and others (e.g., Benassi & Mahler, 1985). The latter result would replicate our finding from Study 1.

Method

Participants The sample consisted of 105 female undergraduate psychology students enrolled in introductory psychology classes who participated as part of a course requirement. As in Study 1, most participants were first-year students enrolled in a northeastern United States public university.

Contingency Task We used a contingency judgment task modeled after the one used by Alloy and Abramson (1979). The experimenter showed the participant a telegraph key and the box with blue and yellow lights and said that her task was to learn the relation between telegraph key responses and blue light onset. The experimenter explained that the yellow light signaled the start of each new trial and would stay on for two seconds during which she had the option of either making a key press response or not making a response. He told the participant that she should press on some trials and not on others in order to learn the relation between her responses and onset of the blue light.

The experimenter next informed the participant that she would have 24 trials in which to learn the relation between the responding on the key and the onset of the blue light, after which she would be asked to indicate her judgment of control on a scale that ranged from 0 (no control) to 100 (complete score). Finally, the experimenter informed the participant that she might discover there was no relation between her responses and blue light onset, or that there might be some relation, either moderate or perfect. When the participant indicated she understood the instructions, the experimenter left the room and activated the apparatus. Objective contingency may be defined by ∆P, the probability of an outcome (O) given a response (R) minus the probability of an outcome without a response (∆P= |P[O|R] - P[O|not-R]|). The apparatus was programmed so that the probability of the blue light onset on the trials on which the participant pressed the telegraph key was 75% as well as 75% on the no-press trials. That is, there was no programmed relationship between pressing and not pressing the telegraph key and blue light onset. After the participant had completed the 24th trial, the experimenter returned and asked her to complete the judgment-of-contingency scale.

Procedure A male experimenter tested participants individually in a quiet laboratory room. Each session lasted approximately 45 minutes. Upon entering the laboratory, the participant completed an informed consent form and the depressive symptoms scale (Plutchik et al., 1970). The experimenter then described the key and light apparatus and gave her instructions for completing the first task.

The apparatus, instructions, and procedure for the second task were the same as for the die toss task used in Study 1. After the 20th trial, the participant wrote the number of trials on which she believed the target color landed up. The third task was a coin flip task. The experimenter showed the participant a quarter and told her he would flip the coin 20 times. He instructed her to focus her mental energy on trying to foresee the outcome prior to the coin flip. The experimenter explained that he would sit behind her while flipping the coin. The actual reason he positioned himself behind the participant was to hide the fact she was given a prearranged pattern of feedback on the task. The pattern of feedback of 10 correct and 10 incorrect predictions, with 7 correct responses occurring in the first 10 trials, was based on the pattern used by Langer and Roth (1975). After the 20th trial, the experimenter asked the participant to write down how many correct predictions she had made. Following the final task, the experimenter debriefed the participant orally and in writing.

Results and Discussion

We analyzed the three dependent measures using principal components analyses with varimax rotation. Table 2 shows the factor loadings when the measures were forced to load onto a single factor. A single factor model is not appropriate. Table 2 also shows the factor loadings for the three illusion-of-control measures based on a two-factor model. This model shows that the psychokinesis/precognition tasks (eigenvalue = 1.26, 41.8% of the variance) and the contingency task (eigenvalue = 1.00, 33.3% of the variance) tapped separate dimensions. We created factor scores (M = 0; SD = 1) to represent participants' standing on each factor.

Using the same procedure we had adopted in Study 1, we transformed participants' depressive symptom scores to z-scores and then formed two groups to create a measure that differentiated participants who were low and high in depressive symptoms.

We analyzed the psychokinesis/precognition illusion-of-control factor scores and contingency factor scores (see Table 1) using a multiple analysis of variance (MANOVA), with level of depressive symptoms as the independent variable. For the psychokinesis/precognition measure, illusory judgments were higher for the high depressive symptoms group. For the contingency judgment measure, illusory judgments were lower for the high-depressive-symptoms group. The MANOVA confirmed the predicted type of factor score × level-of-depressive- symptomatology-interaction effect, F(1, 103) = 8.09, p < .004, r = .27. The effects for level of depressive symptoms (F < 1) and type-of-factor score (F < 1) were nonsignificant.

The pattern of means for the psychokinesis/precognition measure is consistent with the results reported in Study 1 for the psychokinesis/magical ideation factor. The present results also extend the work of Thalbourne and colleagues who demonstrated that higher levels of depressive (also manic and manic-depressive) symptoms were associated with greater belief in paranormal phenomena and magical ideation (e.g., Thalbourne et al., 1997; Thalbourne & French, 1995; Thalbourne, Keogh, & Crawley, 1999). These researchers assessed paranormal belief or magical ideation with general measures. Study 2 demonstrated that illusory control judgments on experimental psychokinesis/precognition tasks were positively associated with depressive symptoms.

Our results also support, with qualification, Alloy and Abramson's (1979) sadder-but-wiser hypothesis. Participants in the group with the highest level of depressive symptoms showed the lowest level of illusory control on the contingency task. The tendency for those high in depressive symptoms to make illusory judgments occurs in areas related to magical causation and putative paranormal experiences but not on the contingency judgment task used in Study 2 – a task unrelated to magical causation. Even in the area of contingency judgment, however, several researchers have found that higher depressive symptoms may be associated with higher illusory judgments under certain conditions (e.g., Benassi & Mahler, 1985; Martin et al., 1984).

General Discussion

The present research did not, and was not intended to, provide a test of whether serious psychopathology was associated with illusory control judgments. However, we did intend to uncover, using students from a normal college population, a link between depressive symptoms and paranormal illusory control. We found a positive relation between symptoms and illusory control in both Study 1 and Study 2. We also found a negative relation between depressive symptoms and illusory control judgments in Study 2 for a standard contingency judgment task (Alloy & Abramson, 1979). The results for the contingency judgment task (and the work of Alloy, Abramson, and colleagues) demonstrate that those high in depressive symptomatology do not simply show an across-the-board tendency to misperceive causal relations. In Study 2, the same participants' judgments were either relatively high or low on the two illusory-control-factor scores depending on whether they had reported a relatively high or low level of depressive symptoms. Thus, the nature of the task used to assess illusory control had an important effect on the results obtained in the study (cf. Presson & Benassi, 1996).

The between-group differences in judgments that we obtained on the paranormal tasks were not large in absolute terms. We found relatively small, but reliable, effects. However, the pattern of results we found was consistent with what others have found in their tests of the association between magical thinking and psychopathology. Some of these other studies also tested college students (e.g., Eckblad & Chapman, 1983; Thalbourne et al., 1999), while others tested adults diagnosed with a manic-depressive disorder or schizophrenia (e.g., Thalbourne & Delin, 1994).

In future work, it would be worthwhile to include men and adults from the general population as participants, as we used only women college students in the present studies because of their relative availability. Also, there is a need to increase the number and type of paranormal tasks used in studies to test the generality of our findings and to increase the reliability of our illusory-control-factor scores. Finally, the present work (and the other work we cited in this article) has not addressed empirically the underlying nature of the relations we have identified. Does depression (or do the other types of disorders noted in this article) predispose someone to magical thinking? Does magical thinking predispose someone to develop a mood disorder? Is there a bidirectional relation? Does a third variable predispose someone to both depression and magical thinking? Answers to such questions await further inquiries that will require different research methods, samples, and analytical approaches from those which we used in the present work.

Alloy, L. B., & Abramson, L. Y. (1979). Judgment of contingency in depressed and nondepressed students: Sadder but wiser? Journal of Experimental Psychology: General, 108, 441-485.

Alloy, L. B., & Clements, C. M. (1992). Illusion of control: Invulnerability to negative affect and depressive symptoms after laboratory and natural stressors. Journal of Abnormal Psychology, 101, 234-245.

American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. 4th ed. Washington: Author.

Ayeroff, F., & Abelson, R. P. (1976). ESP and ESB: Belief in personal success at mental telepathy. Journal of Personality and Social Psychology, 34, 240-247.

Benassi, V. A., & Belli, R. F. (1989). Statistics and replication experiments: Comparing interaction effects. Cognitive Therapy and Research, 13, 283-294.

Benassi, V. A., & Mahler, H. I. (1985). Contingency judgments by depressed college students: Sadder but not always wiser. Journal of Personality and Social Psychology, 49, 1323-1329.

Benassi, V. A., Sweeney, P. D., & Drevno, G. E. (1979). Mind over matter: Perceived success at psychokinesis. Journal of Personality and Social Psychology, 37, 1377-1386.

Bryson, S. E., Doan, B. D., & Pasquali, P. (1984). Sadder but wiser: A failure to demonstrate that mood influences judgments of control. Canadian Journal of Behavioral Science, 65, 145-153.

Chapman, L. J., Chapman, J. P., Kwapil, T. R., Eckblad, M., & Zinser, M.C. (1994). Putatively psychosis-prone subjects 10 years later. Journal of Abnormal Psychology, 103, 171-183.

deCharms, R. (1968). Personal causation. New York: Academic Press.

Eckblad, M., & Chapman, L. J. (1983). Magical ideation as an indicator of schizotypy. Journal of Counseling and Clinical Psychology, 51, 215-225.

George, L., & Neufeld, R. W. (1987). Magical ideation and schizophrenia. Journal of Consulting and Clinical Psychology, 55, 778-779.

Irwin, H. J., & Green, M. J. (1998-1999). Schizotypal processes and belief in the paranormal: A multidimensional study. European Journal of Parapsychology, 14, 1-15.

Kwapil, T. R., Miller, M. B., Zinser, M. C., Chapman, L. J., Chapman, J., & Eckblad, M. (1994). A longitudinal study of high scorers on the hypomanic personality scale. Journal of Abnormal Psychology, 109, 222-236.

Langer, E. J. (1975). The illusion of control. Journal of Personality and Social Psychology, 32, 311-328.

Langer, E. J., & Roth, J. (1975). Heads I win, tails it's chance: The illusion of control as a function of the sequence of outcomes in a purely chance task. Journal of Personality and Social Psychology, 32, 951-955.

Martin, D. J., Abramson, L. Y., & Alloy, L. B. (1984). The illusion of control for self and others in depressed and nondepressed college students. Journal of Personality and Social Psychology, 46, 125-136.

Peltzer, K. (2002). Paranormal beliefs and personality among black South African students. Social Behavior and Personality, 30, 391-398.

Plutchik, R., Platman, S. R., Tilles, R., & Fieve, R. R. (1970). Construction and evaluation of a test for measuring mania and depression. Journal of Clinical Psychology, 26, 499-503.

Presson, P. K., & Benassi, V. A. (1996). Illusion of control: A meta-analytic review. Journal of Social Behavior and Personality, 11, 493-510.

Thalbourne, M. A. (1998). Technical note: The level of paranormal belief and experience among psychotics. Journal of Parapsychology, 62, 79-81.

Thalbourne, M. A., Bartemucci, L., Delin, P. S., Fox, B., & Nofi, O. (1997). Transliminality: Its nature and correlates. The Journal of the American Society for Psychical Research, 91, 305-331.

Thalbourne, M. A, & Delin, P. S. (1994). A common thread underlying belief in the paranormal, creative personality, mystical experience, and psychopathology. Journal of Parapsychology, 58, 3- 38.

Thalbourne, M. A., & French, C. C. (1995). Paranormal belief, manic-depressiveness, and magical ideation: A replication. Personality and Individual Differences, 18, 291-292.

Thalbourne, M. A., Keogh, E., & Crawley, S. E. (1999). Manic-depressiveness and its correlates. Psychological Reports, 85, 45-53.

World Health Organization. (1992). The International Statistical Classification of Diseases and Related Health Problems. 10th revised ed. Geneva: Author.

Zusne, L., & Jones, W. H. (1989). Anomalistic psychology: A study of magical thinking. (2nd ed.) Hillsdale, NJ: Erlbaum.

Appreciation is due to reviewers including Dr. Michael Thalbourne

Department of Psychology

Room 518 Hughes Building

University of Adelade

SA5005

Australia.

Victor A. Benassi, Department of Psychology, Conant Hall, University of New Hampshire, Durham, New Hampshire 03824, USA. Phone: 603- 862-3180; Fax: 603-862-4896; Email: [email protected]