Paranormal phenomena are those which, if genuine, would violate basic limiting principles of science. (Tobacyk, 1988, p. 3), and paranormal belief is belief in paranormal phenomena. The research on paranormal belief can be traced back to the interest on investigating the concept of superstition. For example, Scheibe and Sarbin (1965) tried to conceptualize superstition theoretically, and the variables correlated with superstition/supernatural belief such as conservatism (Boshier, 1973), gender (Bhushan & Bhushan, 1986; Blum, 1976), surgical stress (Shrimali & Broota, 1987), and locus of control (Randall & Desrosiers, 1980; Scheidt, 1973) had also been studied. In addition, the relationship between paranormal belief and the Barnum effect has been investigated (Tobacyk, Milford, Springer & Tobacyk, 1988).

One of the correlates with paranormal belief that attracts much attention is religiosity or  religious belief, since religion is an important part of human culture and the presence of God/gods have yet to be proved scientifically. A number of studies have shown that paranormal belief correlated with religiosity (e.g. Buhrmann & Zaugg, 1983; Orenstein, 2002; Thalbourne & Hensley, 2001). However, other studies did not support this association (e.g., Ellis, 1988; Rice, 2003). The fisrt goal of the present investigation therefore was to replicate previous findings. Furthermore, the relationship between paranormal belief and religiosity has not been explored in Chinese society as well. In the recent decade, although there have been studies in finding the personality or psychopathology variables correlating with paranormal belief (e.g., Dag, 1999; Rattet & Bursik, 2001; Wiseman, Greening &

Smith, 2003; Wolfradt, 1997), many studies showed an inconsistent result. For example, empirical studies have shown extraversion to have a high association between paranormal belief and alleged paranormal experience (Honorton, Ferrari & Bem, 1992; Schmeidler, 1982). The sheep.goat effect was found in individuals with higher paranormal belief scores (sheep) to be more extraverted than disbelievers (goats) (Thalbourne, 1981). However, some studies indicated extraversion was not associated with paranormal belief (Rattet & Bursik, 2001; Windholz & Diamant, 1974). Consequently, extraversion seems to be not as obvious a predictor of belief in paranormal phenomena. One possible explanation for the failed replications is the limited reliability of this psychological trait. Therefore, one of our attempts in this study is to find a stable psychological indicator of paranormal belief.

Intelligent or highly educated participants have been shown to have less paranormal belief (e.g., Blum & Blum, 1974; Jahoda, 1970; Killen, Wildman & Wildman, 1974), and this relation did not seem to be accounted for by context effects (Smith, Foster & Stovin, 1998). Moreover, Irwin (1993) contended that no correlation between paranormal belief and intelligence was found from several studies. Musch and Ehrenberg (2002) suggested that general cognitive ability might be a critical underlying variable correlating with paranormal belief and, then, they found general cognitive ability negatively correlated with belief in the paranormal. Nevertheless, intelligence or general cognitive ability correlates with education (Kaufman, 1990), which in turn appears to be related to cognitive complexity. According to the Personal Construct Theory proposed by George Kelly and later elaborated further by his student James Bieri, a cognitive complex person has a personal construct system in which the constructs are clearly differentiated, whereas a cognitive simple person has a personal construct system containing constructs that are poorly differentiated (Potkay & Allen, 1986). The measurement of cognitive complexity indicates the degree of differentiation of the personal constructs construed by the participant, that is, the degree of non-overlapping of these constructs. People who had a higher cognitive complexity related positively to their degree of confidence (Adams-Webber, 2003). We suspected that cognitive complexity might be an important predictor of paranormal belief. The effect of cognitive complexity on paranormal belief has yet to be determined. Therefore, this study has two purposes.

First, the Chinese version of the Revised Paranormal Belief Scale was constructed. Second, the relationship among paranormal belief, religiosity and cognitive complexity was examined. It was hypothesized that cognitive complexity negatively correlated with both paranormal belief and religiosity respectively.