Some theorists (e.g., Langone, 1996; Lifton, 1987, 1991; Singer & Lalich, 1995) have proposed that cults use a systematic program of psychological manipulation to convince people to join cults and remain in them. These manipulations may take a severe toll. Nevertheless, not all cults exercise the same level of abusiveness. If one accepts that cults use deception in the recruitment and indoctrination process (Andersen, 1985), one cannot rely on their self-portrayal or any comparisons one may draw from such self-portrayals. Additionally, comparisons with benign groups are useful, especially for educational purposes.

Personality and Psychiatric Symptomatology

Langone (1996) concluded that cult members’ personality profiles fall within the normal range. Walsh et al. (1995) found that although ex-cult members had elevated scores on the neuroticism scale of the Eysenck Personality Questionnaire, their scores approached normal as a function of time since leaving the cult. Yeakley (1988) gave the Myers-Briggs Type Indicator to members of the BCC. He asked them to respond to each item one time as they would have responded before their conversion, a second time as they perceived themselves at the time the study was conducted, and a third time as they imagined themselves answering after five more years of discipling. Nearly all respondents tended to change their psychological type scores across the three versions. According to Yeakley, the direction in which these changes occurred was toward the personality of the leader.

Perceptions of Cult Members and Ex-members

Singer (1978) and Clark (1979) first recognized that dissociative defenses are a mechanism by which cult members (much like political captives and hostages) adapt to the intense demands of the environment in which they find themselves. As we have mentioned, the syndrome associated with this stressful adaptation has since been called pseudo-identity (West & Martin, 1994). While members are under the group’s psychological control and are not in a state of questioning or rebellion, the pseudo-identity can appear to be normal and well adjusted. However, if an event or outside influence is strong enough to fracture the pseudo-identity, which until then has enveloped the original self, the underlying pain and psychological harm become apparent. Langone (1995) insightfully remarked that the reason why cult members generally do not return to the cult after the floodgates of recognition and emotion have opened is "because the suffering they experience after leaving [the cult] is more genuine than the ‘happiness’ they experienced while in it. A painful truth is better than a pleasant lie. . . . If this analysis is correct, ex-members may indeed provide more accurate information about cults than would current members" (pp. 8-9).

Former members, however, are often reluctant to participate in scientific studies. Several factors may motivate this reluctance, such as fear of retaliation from the cult, embarrassment at having been manipulated, ridicule from those who lack understanding, and a need for closure. For this reason, the methodology of the present study was designed with an awareness of these potential sensitivities.

Background: Church Universal and Triumphant

Church Universal and Triumphant (CUT) is a New Age apocalyptic cult with roots in Theosophy and various splinters of the I AM Movement. Its core belief is that a pantheon of Gods, Goddesses, Angels, and Ascended Masters communicate via their messenger (Elisabeth Claire Prophet) to teach, prophesize, reprimand, and direct devotees. Ascended Masters are considered enlightened heavenly beings who have "graduated from earth’s schoolroom" by balancing their karma, thereby having escaped the cycle of reincarnation. They are seen as exemplars and teachers to assist their followers to do the same. Obedience to the dictates of these heavenly beings is seen as a guaranteed path to the individual’s ascension. Although Prophet is still "inhabiting a physical body," years ago it was announced that she has balanced all of her personal karma and voluntarily remains in embodiment. Her mission is one of "balancing world-karma," of being the one and only mouthpiece of God, and of being a guru and spiritual teacher in the flesh. Thus, her word and position are beyond reproach, and her power over CUT members is absolute.

One of the major identifying practices used by CUT is that of decreeing, which is also referred to as the "science of the spoken word." It is a fast and repetitious form of chanting affirmations and commands in the name of the individual’s own Christ Self and by the authority of any divine being whose power the individual seeks to invoke. Decreeing can last many hours each day, and it can be done in solitude, in small groups, or in very large groups. It is seen as the all-powerful tool to influence not only present and future events but also personal and global history, including events from past incarnations. Experts on thought reform have noted that fast, repetitive chanting, such as decreeing, may serve as a thought-stopping technique (Conway & Siegelman, 1995; Hassan, 1988). In the context of Lifton’s (1987, 1991) theory of ideological totalism, CUT not only qualifies as a totalistic environment but also exemplifies all eight criteria identified by Lifton.

Research Questions

For a selected group of former members of CUT, we investigated the following four main questions:

Instruments

The present survey used several instruments: (a) two copies of the Group Psychological Abuse scale (Chambers et al., 1994); (b) the revised Symptom Checklist 90 (Derogatis, 1994); (c) the revised Eysenck Personality Questionnaire (Eysenck & Eysenck, 1994); and (d) a 29-item background questionnaire, an adaptation and expansion of a form used by Langone (1995) in a study investigating the BCC.

Group Psychological Abuse Scale. We used the Group Psychological Abuse (GPA) scale to assess ex-members’ current and pre-involvement (retrospective) perceptions of CUT. The two GPA copies were identical except for the introductory paragraphs.

One copy asked respondents to respond according to "your current understanding of CUT," whereas the other requested that they respond according to "the way you would have responded when you first joined the group." To control for order effects, half of the surveys were sent out with the post-involvement version to be filled out before the pre-involvement (retrospective) one, and the other half of the surveys had the two versions stapled together in reverse order. The instructional paragraphs to both versions requested that responses reflect respondents’ personal experiences and observations. The GPA scale consists of 28 items, 7 on each of four factors (subscales), which identify Compliance, Exploitation, Mind Control, and Anxious Dependency as specific dimensions of group psychological abuse. Scores on the Compliance subscale reflect beliefs about the degree to which an individual must accede to group norms (e.g., "Members must abide by the group’s guidelines regarding dating and intimate relationships"). Scores on the Exploitation subscale reflect beliefs about the degree to which a cult manipulates, abuses, and uses people (e.g., "The group advocates or implies that breaking the law is okay if it serves the interests of the group"). Scores on the Mind Control subscale reflect beliefs about the degree to which the cult leadership uses deceptive and manipulative means to sustain membership (e.g., "People who stay in the group do so because they are deceived and manipulated"). Scores on the Anxious Dependency subscale reflect beliefs about the degree to which members depend in an absolute and possibly also anxious way on the group (e.g., "Members believe that to leave the group would be death or eternal damnation for themselves or their families"). On each item, respondents rated the target group on a 5-point Likert scale from 1 (not at all characteristic) to 5 (very characteristic). Thus, the range for each subscale is 7-35, and the range for the overall (composite) score is 28-140. Scores above the midpoint (21 for each subscale, 84 for the composite) indicate that the respondent rated the group as being abusive (Langone, 1994). Previously obtained Cronbach’s alpha (reliability) coefficients for the GPA ranged from .70 to .81 on the four subscales (Chambers et al., 1994).

Symptom Checklist 90, Revised. The Symptom Checklist 90, Revised (SCL-90-R), a multidimensional self-report inventory widely used in psychiatric screening to measure psychological distress levels, offers four normative scoring versions for both males and females. The norms have been established for psychiatric inpatients, psychiatric outpatients, nonpatient adults, and nonpatient adolescents. The present study used nonpatient adult norms. A total of 90 items measure the severity of symptoms on nine distinct dimensions: (a) somatization, (b) obsession-compulsion, (c) interpersonal sensitivity, (d) depression, (e) anxiety, (f) hostility, (g) phobic anxiety, (h) paranoid ideation, and (i) psychoticism. The instrument includes seven additional items that are reflected only in the three global scores: the Global Severity Index, the Positive Symptom Distress Index, and the Positive Symptom Total. Their inclusion is based on their clinical significance. The function of the global measures is to express the level or depth of psychological distress in a single score.

Because the SCL-90-R is used to screen various non-psychiatric populations (e.g., Derogatis & DellaPietra, 1994; Derogatis & Lazarus, 1994), an operational definition of what constitutes a positive case seems appropriate and helpful. Derogatis (1994) defined caseness by an operational rule which states that the individual is considered a positive risk if the person has a Global Severity Index score on the nonpatient norm greater than or equal to a standardized score of 63 (or scores 63 or above on two primary dimensions). Previously obtained Cronbach’s alpha coefficients for the SCL-90-R ranged from .77 to .90 on the various symptom dimensions and from .83 to .90 for test-retest coefficients (Derogatis, Rickels, & Rock, 1976; Horowitz, Rosenberg, Baer, Ureno, & Villasenor, 1988). Criterion validity studies, in particular with MMPI scales, showed high convergent validity and peak correlations on eight of the nine scales of the SCL-90 (Derogatis et al., 1976).

Eysenck Personality Questionnaire Revised. The Eysenck Personality Questionnaire, Revised (EPQ-R) is a self-report instrument that uses 73 items to determine the strengths of three dimensions of personality: psychoticism or "tough-mindedness," neuroticism or "emotionality," and extroversion. A fourth subscale, the Lie scale, uses 21 additional items to measure dissimulation or social desirability. Norms for the EPQ-R have been established for American males and females (see Results section). Previously obtained Cronbach’s alpha coefficients for the subscales range from .66 to .86 (Eysenck & Eysenck, 1994).

Procedure

From a mailing list for a newsletter that is primarily sent to former members and some families of current members, 90 former members of CUT were identified. The survey materials were sent to all of them. In order not to violate people’s trust by revealing their names and addresses, the investigator never had access to the mailing list. Instead, the newsletter’s editor affixed address labels to the sealed and stamped envelopes and delivered them to the post office. Survey materials were also sent to 6 former members who had not been aware of the newsletter, but who heard of the study by word of mouth and requested a copy of the survey materials.

We did not ask survey recipients for their names or other information that would have violated their anonymity. To obtain their approximate ages, we asked them their birth year. Respondents were informed that by completing and returning the forms, they consented to have their responses included in the study.

Respondents

A total of 61 individuals returned questionnaires that were useable for statistical analysis. The response rate (63.5%) is unusually high for survey research, which may be partially attributable to the fact that survey recipients were encouraged to pass on the materials to other former members if they themselves did not wish to respond. (We excluded one questionnaire because more than 50% of the items on each instrument were unanswered.) Of the 61 returned surveys, 60 included both GPA forms, 58 included the EPQ-R, 55 included the SCL-90-R, and 60 included the background questionnaire.

A total of 35 (57.4%) of the respondents were female, and 26 (42.6%) were male. In addition, 57 (93.4%) respondents were White, 2 (3.3%) were Black, and 2 (3.3%) were of mixed ethnicity.

Respondents’ religious upbringing was as follows: 31.1% Roman Catholic, 24.6% Protestant liberal, 16.4% Protestant fundamental, 9.8% CUT, 3.3% Jewish, 1.6% Mormon, and 1.6% New Age (other than CUT). An overwhelming 80.3% reported having no specific religious affiliation at the time the study was conducted, 8.2% identified themselves as being Roman Catholic, 6.6% as fundamental Protestant, and 4.9% as liberal Protestant.

The median annual household income was $50,000. The mean number of years of education was 15, and 87% had completed at least one year of higher education.

The age of respondents ranged from 18 to 79 years, with a mean of 45.1 (SE = 1.4). Age at the time of joining CUT ranged from 0 (born into the group) to 58 years, with a mean of 25.7 (SE = 1.4). Nearly 60% had joined between the ages of 18 and 28.

The length of time respondents reported having spent in CUT ranged from 1.8 to 22.2 years, with a mean of 11.4 (SE = 0.7). The level of involvement varied widely, ranging from "Keeper of the Flame," which may be considered the lowest level of affiliation or commitment, to attending the leader and her family as their personal staff. The length of time between exiting CUT and completing the survey ranged from 0.6 to 20.0 years, with a mean of 7.8 (SE = 0.7).

Four individuals (6.6%) reported that they had received outpatient counseling before their involvement with CUT. Three (4.9%) reported that they had received counseling during their involvement. A total of 21 (34.4%) reported that they received counseling after leaving the group, and 38 (62.3%) reported that they had never received any counseling. (Percentages add up to more than 100% because some individuals had received therapy during more than one of these periods.)

Thirty-eight (62.3%) respondents were involved in a formal or informal support group or network at some time since their exit. Most comments relating to that involvement affirmed the benefits of receiving validation that come from sharing with those who understand the effects of cult involvement and the benefits of reducing their isolation.

Thirty-five respondents (57.4%) rated the overall CUT experience as very harmful, 16 (26.2%) as harmful, 4 (6.6%) as neutral, 3 (4.9%) as mildly beneficial, and another 3 (4.9%) as very beneficial. Several individuals falling into the latter two categories qualified their responses by stating that they had learned to recognize abuse for what it was, learned that they had the right and power not to subject themselves to it any longer, or both.

Table 1 shows the mean and standard error for each subscale, as well as for the composite score, on both versions of the GPA. It also shows corresponding alpha coefficients. In Table 1 and hereafter, the GPA filled out retrospectively, responding as if at the time of joining CUT (pre-involvement), is referred to as GPA1, and the GPA responding as to current perceptions of abuse (post-involvement) is referred to as GPA2.

Paired-sample t tests revealed that compared to the GPA1, reported abusiveness was greater on each of the GPA2 subscales, as well as on the composite score, all t(59) > 12.2, p < .001.

Comparison of GPA ratings for CUT and BCC.

With only means and standard deviations of previous studies available for statistical analysis, a comparison of the GPA2 (post-involvement) ratings of former CUT members with those of former BCC members was conducted using one-sample t tests for each of the subscale means, as well as the composite mean. The means assumed according to the null hypothesis for these t tests were those from Langone’s (1995) investigation of the BCC. Although Langone published the means from two separate samples of BCC ex-members, we used the higher values to make a more conservative comparison. As shown in Table 1, GPA2 means were higher than the corresponding BCC means on each subscale, as well as on the Composite, all t (60) > 7.7, p < .001. A comparison with other groups investigated by Langone was unnecessary, because their mean scores were significantly lower than those for BCC. A comparison using the means of the original GPA study (Chambers et al., 1994) was also significant for each subscale rating and the composite rating, all t > 6.48, p < .001.

Evidence of discriminant validity for the GPA scale comes from the nonsignificant correlations between the lie scale of the EPQ-R and GPA1 and GPA2 Composite scores, both p > .05.

Psychiatric Symptomatology of Ex-CUT Members (SCL-90-R)

We assessed the reported psychiatric symptomatology of ex-members by calculating two measures derived from responses to the SCL-90-R. The overall severity of reported psychiatric symptomatology was assessed by the typically used measure, the Global Severity Index (GSI). To assess the breadth of elevated scales in addition to their overall severity, we also created another variable, the number of high scores. It is the number of SCL-90-R subscales on which an individual scored a standardized value of 63 or above. Numerical values for this variable ranged from 0-9 (9 subscales). We then investigated correlations between these two measures and other measures concerning pre-involvement, involvement, and post-involvement periods (see Table 2). We also created multiple regression models to determine whether any other variable predicted the severity of psychiatric symptomatology.

Note. Scores above the midpoint (21 for each subscale, 84 for the composite) indicate that the respondent rated the group in the abusive range. Standard error (SE) of each mean is shown in parentheses. BCC data are from Langone (1995), who did not provide alpha coefficients. All means in the same row differ at p < .001.

Responses to the SCL-90-R revealed that 27 (49.1%) of the respondents met Derogatis’s (1994) definition of caseness or positive risk. Thus, nearly half of the respondents reported clinically relevant psychiatric symptomatology. Responses were internally consistent: Cronbach’s alpha coefficients for the nine subscales of the SCL-90-R ranged from .75 to .95, and for the entire instrument (all 90 items) alpha was .97.

Group Psychological Abuse Ratings

As shown in Table 2, there was no significant relationship between either pre-involvement (GPA1) or post-involvement (GPA2) ratings of psychological abuse and either measure of psychiatric symptomatology. Thus, respondents who perceived CUT as most abusive were not necessarily those who reported the greatest psychiatric symptoms.

Duration of Involvement and Post-involvement Periods

Although the GSI was not significantly correlated with time in CUT, it was negatively correlated with time out of CUT. However, the number of high scores was positively correlated with time in CUT and negatively correlated with time out of CUT. Thus, reported psychiatric symptoms tended to increase with the length of time during which a person was a member of CUT and to decrease with the length of time since the person had become an ex-member of CUT.

Relationship with Spouse

*p < .05. ***p < .001.

Multiple Regression Models

When the GSI was used as the dependent variable in a multiple stepwise regression model, time out of CUT and quality of the relationship with spouse during CUT were significant predictors, p < .001 and p < .017, respectively. Neither time in CUT nor the relationship with one’s spouse after CUT entered into the model. The overall R² = .46 was significant, F(2, 40) = 13.49, p < .001, and the standardized regression coefficients (b s) were -.55 for time out of CUT and -.33 for relationship with spouse during CUT. A second multiple regression model was constructed to predict the number of high scores. An initial stepwise multiple regression also revealed that both time out of CUT and the reported quality of the relationship with one’s spouse after CUT involvement were significant predictor variables, p = .005 and .011, respectively. Neither time in CUT nor the relationship with one’s spouse during involvement with CUT entered into the model. Next, each other variable was combined with time out of CUT, using a simultaneous entry procedure, in order to determine whether any increased the proportion of variance explained (R²). Again, only the quality of the relationship with one’s spouse after CUT and time out of CUT were significant predictors. The overall R² = .45 was significant, F(2, 40) = 14.67, p < .001, and the standardized regression coefficients (b s) were -.40 for relationship with spouse after CUT and -.39 for time out of CUT. In short, these analyses both show that respondents who reported a relatively good relationship with their spouse after CUT involvement and who had left CUT a relatively long time ago tended to report fewer psychiatric symptoms.

Comparisons of Ex-Members’ Personalities with Established Norms (EPQ-R)

Responses to the EPQ-R showed good internal consistency: Alpha coefficients for the four subscales ranged from .67 to .89. One-sample t tests were used to compare respondents’ mean score with the normative mean on each of the subscales (as provided in the manual for the EPQ-R; Eysenck & Eysenck, 1994). As shown in Table 3, neither male nor female ex-CUT members differed from the norm on the Psychoticism scale or on the Neuroticism scale. Scores on the Extroversion scale, however, were significantly lower than the norm for females and marginally lower for males. On the Lie (dissimulation) scale, there was no significant difference for females. For males, the difference was significant: Male respondents scored lower (exhibited less dissimulation) than the norm.

Note. Standard error of each mean is shown in parentheses. Means in the same row that do not share subscripts differ at p < .05. For former CUT members, n = 33 females and 25 males.

The correlation between extroversion and time out of CUT was positive, r = .31, p = .019. Neuroticism and time out of CUT were negatively correlated, r = -.55, p < .001. The correlation between neuroticism and extroversion was also negative, r = -.62, p < .001. Extroversion was negatively correlated with dissimulation, r = -.33, p < .016. There were negative correlations between dissimulation and measures of psychological distress on the SCL-90-R for both the GSI, r = -.41, p = .002, and the number of high scores, r = -.35, p = .009.

Reported Quality of Personal Relationships

Table 4 shows the mean reported quality of personal relationships before, during, and after CUT involvement. Paired-sample t tests across all relationships revealed higher ratings before than during CUT involvement. In addition, all after ratings were higher than corresponding during ratings. For spouse relationships, ratings were higher before and after than during CUT involvement, and for other relationships, ratings were higher before than after; during ratings were the worst. In short, the reported quality of all relationships was consistently worse during CUT involvement.

This study investigated a sample not examined by previous research: former members of CUT. It used the first, and as yet only, objective measure designed to assess the level of group abusiveness, the GPA Scale (Chambers et al., 1994). A novel use of the instrument made it possible to measure reported discrepancies between initial and post-involvement perceptions of abusiveness, thereby assessing the extent to which individuals’ interpretation of events may have changed or to what extent misrepresentation on the part of the group may have influenced their perceptions. Although the study used no comparison groups—a design feature that may be regarded as a weakness—whenever it was appropriate and possible, we made use of established normative values and comparative data from previous research to evaluate the findings. They clarify the major four research questions posed earlier.

Perceived Abusiveness and Deception

No previous study using the GPA scale has found mean values for the composite score and each subscale score that were as high as those found here. The difference between the overall GPA means for another cultic group, the BCC (Langone, 1995), and the CUT sample exceeded two standard deviations. The overall GPA mean in the initial study (110.7), which surveyed former members of 101 different cults, was only 2.2 points higher than the mean score of the BCC sample (Langone, 1995). The present findings suggest that among groups perceived as abusive by their former members, CUT is at the high end of reported abusiveness. Even if sampling bias is taken into account and one assumes that the mean score of former members who do not subscribe to the newsletter were two standard deviations below that of subscribers, ex-CUT members’ GPA means would be similar to those obtained from former members of other allegedly cultic groups.

A comparison of retrospective GPA scores reflecting ex-members’ recollections of their perceptions when they first joined the group with their post-involvement perceptions revealed profound differences (see Table 1). Several studies have investigated retrospection biases (Marcus, 1986; for a review, see Dawes, 1988). These studies argue that recollections of past beliefs are biased toward current beliefs. The large discrepancies between retrospective perceptions and current perceptions of former CUT members suggest that any such bias did not eliminate the differences found here. The perceptions these ex-members remember having had at the time they joined CUT match the image CUT tries to portray, which is clearly a nonabusive one. Respondents’ current perceptions reflect a highly abusive environment. If a retrospection bias toward current views operated in this study, the retrospective scores were conservative. If we had been able to measure abusiveness while respondents were actually joining CUT, the reported abusiveness might have been even lower. Other studies suggest that peoples’ investment (e.g., in time, money, or emotion) in something biases them to exaggerate the differences between pre- and post-involvement if there is an unconscious desire to justify the investment by viewing the change as a positive one. In the present study, the change in perception was negative. Thus, given the conditions and findings of this particular study, neither of these conclusions offers an adequate explanation.

If false recollections are not a likely cause of the discrepancies between the two sets of scores, one might conclude that CUT misrepresents itself and deceives its prospective and current members. Support for this interpretation comes from the finding that the greatest discrepancy between pre- and post-involvement was on the Mind Control subscale, and the next greatest was on the Compliance subscale. These two subscales also showed the highest post-involvement means, and each of them was near the highest possible score (35) on the subscale.

Another possible interpretation for the discrepancies between GPA1 and GPA2 is that individuals changed their interpretation of reality when they first came in contact with CUT and then again when they left it. While a person is still in a cultic group and subscribes to the group’s interpretation of reality, he or she may not perceive abusive actions or attitudes as abusive. Instead, a person may interpret what outsiders might consider being abuse as a need to atone for past transgressions or deserving chastisement for imperfections. In CUT terms, this is an individual’s karma, and "bad karma" may stem from past deeds of a previous or several previous embodiments. This belief allows the leader to interpret current events, including being abused, by attributing them to a past of which the follower may have no knowledge and which may be impossible to verify. Additionally, if one subscribes to the belief that "God chastises those most whom he (or she) loves most," abuse may be interpreted as a privilege of "the chosen." A belief that a follower’s devotion or unquestioning commitment must be tested so that his or her spiritual attainment may be assessed also provides a formidable tool to change the meaning of events. Clinicians who are familiar with issues of domestic abuse may find interesting parallels between what is referred to as mystical manipulation (Lifton, 1987, 1991) in a cult context and the mind games with which victims of domestic abuse are manipulated.

The high discrepancies between GPA1 and GPA2 may also be attributable to individuals’ changes in the interpretation of events that took place during the conversion process, throughout involvement, and since exiting CUT, as well as misrepresentations on the part of CUT. These explanations may be two sides of the same coin. To an outsider, a reinterpretation of reality may appear to be deceptive (and not necessarily abusive), but to the new recruit and long-term follower it may be a "higher form of knowledge." Former members who have shed their pseudo-identity (West & Martin, 1994), however, are not merely outsiders. Their personal experiences may cause them to view the former interpretation of their realities as deceptive and abusive. They may often prefer even stronger terms, such as spiritual betrayal.

Psychological Distress

A large percentage of former CUT members (almost 50% of the respondents) reported experiencing such high levels of psychological distress as to be categorized as positive risks (Derogatis, 1994). However, only 6.6% of the respondents had received psychotherapy or counseling prior to their involvement in CUT. This low percentage may partly be attributable to the relatively young age at which many of them joined (57.4% joined before the age of 25 and 72.1% before the age of 30).

Although the percentage of ex-members experiencing such high levels of psychological distress to qualify as a psychiatric risk is large, as time away from CUT increased, distress levels decreased. Time out of CUT explained a significant proportion of the variability in both regression models (one using Global Severity Index scores as dependent variable, the other using the number of subscales on which an individual had scored 63 or above). Two additional variables that had significant predictive value in one or the other of the two regression models were the quality of spousal relationship during and after CUT involvement. Individuals who reported a higher quality of relationship with their spouse after exiting CUT also reported lower levels of psychological distress than those who had poor spousal relationships. The same is true for the quality of spousal relationship during CUT involvement. This is not surprising if one considers that a relationship which can be maintained at a qualitatively high level while the spouses undergo high levels of environmental stress may be a source of strength for the partners. This will reduce the overall stressfulness of the experience and the psychological harm inflicted. Although time in CUT was excluded from the multiple regression model, it was positively correlated with psychological distress. This suggests that there is a positive relationship between the time an individual spent in CUT and the psychological problems experienced later.

Most theorists think that people who are in pain (psychologically or otherwise) try to attribute that pain to an outside source, which then leads to an unfair negative evaluation of that source. What if those not in pain share the negative evaluations? Nearly half of the present respondents met the criteria for being a positive risk. Some reported high levels of psychological distress, but others were virtually symptom-free. The lowest individual post-involvement GPA rating given to CUT in this study, however, was 109. According to Langone (1995), ratings suggestive of non-abusiveness range from 28 to 84. Moreover, there was no relationship between the reported level of psychological distress and respondents’ perceptions of group abusiveness.

There was a negative relationship between reported psychological distress and the extent to which respondents portrayed themselves in a socially desirable light: Those who reported higher levels were less inclined to answer according to what is considered socially desirable. Respondents who scored relatively high on the lie scale may have underreported psychological distress, perhaps because they did not want to be seen as having problems.

Personalities of Former CUT Members

The present evidence suggests that former members of CUT do not differ from the norm on the psychoticism and neuroticism dimensions. Females did not differ from the norm in their desire to be socially acceptable (as measured by the lie scale of the EPQ-R), but males appeared to be less concerned with their social desirability than one expects of males in the general population. This may be a reaction to the manipulation they experienced during their involvement in CUT. If these scores are, however, representative of a condition present before cult involvement, they may reflect a stronger than average desire for honest self-evaluation. This desire, in turn, may have contributed to the vulnerability of these individuals to manipulation. Honest people often assume honesty in others. If either or both of these proposed explanations are valid, it is unclear why females in this sample did not differ from the norm. Is it possible that females in general feel more pressure in this society to represent themselves in a more socially desirable light?

Extroversion was the only personality dimension on which both males and females differed from the norm, although the difference was slightly greater for females than males. If one takes into account the intense alienation from the outside world former cultists experienced during their involvement and the fact that many, upon their exit, think they were betrayed by the thought reform program to which they were subjected, it makes sense that they became more introverted. The significant positive correlation between extroversion and years out of CUT suggests that with increased time away from the cult environment, extroversion scores approach the norm.

Although the sample’s neuroticism scores did not significantly differ from established norms, there was a significant negative correlation between neuroticism and time since leaving CUT: Respondents who were out longer scored lower on neuroticism.

Changes in Close Personal Relationships

The ratings respondents gave to the quality of their close relationships reveal a clear pattern across all relationship categories. The rated quality of relationships was consistently lowest during CUT involvement. A comparison of before and after ratings showed higher ratings for the quality of post-CUT parental relationships. This was not the case for spouse relationships, which were about the same before and after CUT, nor for other relationships, which were worse after than before CUT. Based on additional comments that respondents made on the survey, it appears that some individuals had rated their relationships with friends who had joined the group with them or someone they had met during their involvement. In either case, these relationships may, as a result of their exit, be strained at best and non-existent at worst. This may be the reason why the post-involvement mean for the category other is lower than the pre-involvement mean.

Several respondents commented that divorces were related to their cult involvement. Some indicated that they were completely cut off from their former spouse and therefore did not give a post-involvement rating for the spouse category. Others who remarried since their exit rated their current marriage. Thus, the relatively high mean for post-cult ratings of spousal relationships appears to be partly attributable to post-CUT marriages. Unfortunately, we did not ask respondents to indicate the year they divorced. This prevented an exact determination of how many marriage dissolutions might have been related to CUT involvement, which marriages dissolved prior to CUT involvement, and which dissolved considerably later.

Limitations, Unanswered Questions, and Suggestions for Future Research

One weakness of the present study was the sample itself, because the majority of the respondents were recipients of a newsletter primarily sent to former CUT members. In this newsletter, grievances about the group and its leadership are freely expressed. Additionally, each issue of the newsletter includes a list of recommended publications about CUT as well as more general cult educational materials. Thus, the sample’s representativeness of ex-members may be questioned. Sixty-one individuals represent a small proportion of the hundreds who have left CUT during the 29 years since it was founded. The majority of those who participated in the study had been exposed to the newsletter’s anti-CUT and anti-cult stance. Although this is a legitimate concern, a lack of understanding about cult-related issues does not necessarily mean that people’s perceptions are more valid. Put differently, more information and education, which is generally regarded as an advantage in discovering the truth, may be a desirable quality in respondents. The present sample offered just that. Additionally, one cannot assume that former members who may not be aware of the newsletter or fear being on its mailing list perceive the group as any less abusive. Representativeness with regard to the level of past group involvement and commitment to the group, as well as the time spent in the group, was remarkably good.

As with most surveys, it is impossible to ascertain what distinguishes respondents from non-respondents. Are non-respondents likely to be more or less distressed than those who participated in the study? Do they hold more or less favorable views of CUT? These questions remain unanswered. Also, research that focuses on children who have left cults is virtually nonexistent. The development of instruments and methodologies that permit such investigation without causing additional harm should be among the goals for future work.

Adams, D. (1993). The Cincinnati Church of Christ: How former members rate the group on a cultism scale. Unpublished master’s thesis, Xavier University, Cincinnati, Ohio.