Post-Cult Symptoms as Measured by the MCMI

Cultic Studies Journal, Volume 9, Number 2, 1992, pages 219-250

Post-Cult Symptoms As Measured by the MCMI Before and After Residential Treatment

Paul R. Martin, Ph.D.

Wellspring Retreat and Resource Center

Michael D. Langone, Ph.D.

American Family Foundation

Arthur A. Dole, Ph.D.

University of Pennsylvania

Jeffrey Wiltrout

Department of Psychology

Ohio University

Abstract

Two groups of former cultists were administered the MCMI and other psychological tests in order to assess the nature and magnitude of post-cult psychological problems. One group consisted of 13 former cultists who attended an educational conference on cults. The other group was made up of Ill former cultists who attended a residential treatment center. In order to evaluate the effectiveness of this treatment, 66 members of the latter group participated in a six-month followup. Post-cult distress at initial testing was high and did not differ between groups. Pre-post evaluations of the treatment group indicated a strong treatment effect. Marked personality configuration shifts in the MCMI profile after treatment occurred. Results appear to support the view that dissociative processes are central to the cult experience.

Cults are exploitatively manipulative groups that utilize thought reform programs (Ofshe & Singer, 1986; Singer & Ofshe,1990) to subordinate members' well-being to the goals of leaders (Langone, in press). Clinical investigations of former members of cultic and related groups indicate that cult involvement results in a significant level of distress for this population (Clark, 1979; Goldberg & Goldberg, 1982; Hochman, 1984; Schwartz, 1985; Singer, 1978, 1987; Spero, 1982; Swartling & Swartling, 1992; Temerlin & Temerlin, 1982; West & Singer, 1980). According to clinicians, the most common symptoms ex- cultists experience are emotional volatility, dissociative symptoms such as “floating” (a phenomenon similar to drug flashbacks), depression, loneliness, guilt, inability to concentrate, indecisiveness, difficulty communicating, fear of retribution, fatigue, a sense of a spiritual-religious-philosophical void, career confusion, and conflicts with family.

The handful of studies that have collected statistical data on ex-cultists have, by and large, supported clinical observations. Among the statistical findings bearing on distress are the following:

Conway, Siegelman, Carmichael, & Coggins, 1986 (The sample consisted of 353ex-cultists from 48 different groups. Because subjects reported on “lasting effects,” symptom reports may reflect in-cult as well as post-cult difficulties.)

• 75% depression

• 68% loneliness

• 68% anger toward group leader

• 59% guilt feelings regarding leaving the group

• 59% feelings of humiliation/embarrassment

• 52% suicidal tendencies

• 49% fear of physical harm by the group

• 48% nightmares

• 42% inability to break rhythms of chanting, meditation, etc.

• 37% hostile feelings toward family

• 31% sleeplessness

• 25% memory loss

• 22% menstrual dysfunction

• 20% physical punishment while in group

• 19% sexual dysfunction

• 18% abnormal weight loss

• 17% violent outbursts

• 17% bewildering, psychic phenomena

• 16% abnormal weight gain

• 15% hallucinations and delusions

• 5% sex with leaders while in group (60% in Children of God)

Galanter, 1983 (The sample consisted of 66 Unification Church dropouts.)

• 36% reported serious emotional problems after leaving

• 24% sought professional help after leaving

• 3% were hospitalized after leaving

• 61% felt Rev. Moon had negatively impacted on members

Knight, 1986 (The sample consisted of 58 former members of a psychotherapy cult.)

• 97% were verbally abused in therapy sessions in the group

• 86% felt harmed by the group exposure

• 82% were shoved at least occasionally in therapy sessions

• 78% were hit at least occasionally in therapy sessions

• 75% sought therapy

• 52% anxiety

• 48% depression

• 48% trouble making decisions

• 41% confusion

• 40% disoriented

• 33% lonely

• 25% had sex with their therapist when in the group

• 18% menstrual cessation

Langone, Chambers, Dole, & Grice (in press) (The sample consisted of 308 former cultists from 101 groups.)

• 83% reported feeling anxiety/fear/worry

• 76% anger toward the group leader

• 72% low self-confidence

• 71% vivid flashbacks to group experience

• 70% received counseling after leaving

• 67% depression

• 67% difficulty concentrating

• 61% despair/hopelessness/helplessness

• 56% guilt about what they did in the group

• 55% “floating” among very different states of mind

• 51% felt as though they lived in an unreal world

• 46% had conflicts with loved ones

• 44% reported that the experience was very harmful

• 42% reported that the group experience was very unsatisfying

• 38% feared physical harm by the group

• 34% severe anxiety attacks after leaving

• 11% were sexually abused in the group

Solomon (1981) has suggested that negative effects reported by former cult members reflect the influence of contact with the anti-cult movement, which, she maintains, seems “to serve a reference group function, influencing an ex-member's attitudes and values, regardless of whether this group had been chosen by the subject or had been imposed primarily through deprogramming and/or rehabilitation” (p. 287). Solomon appears to overlook the possibilities that (a) contact with the “anti-cult movement” could reflect a self-selection process that screens out those ex-cultists who are inclined positively toward their group and (b) the “anti-cult movement” may help ex-cultists become aware of factors that have much objective significance.

Post-cult distress may at least in part reflect pre-cult psychopathology. This suggestion is bolstered by findings concerning the percentage of cultists reporting pre-cult psychological counseling: 7% (Barker, 1984), 30% (Spero, 1982), 30% (Galanter, Rabkin, Rabkin, & Deutsch, 1979), 38% (Galanter & Buckley, 1978), 42% (Langone et al., in press), 59% (Knight, 1986), and 62% (Sirkin & Grellong, 1988). Averaged out, these studies indicate that approximately one third of former cultists had had counseling before joining the cult, a finding that is very similar to Sirkin & Grellong's (1988) non-cult comparison group of Jewish youth, 33% of whom had had counseling. It is certainly possible that cult joiners tend to be somewhat more troubled psychologically than nonjoiners. However, since a comprehensive National Institute of Mental Health epidemiological study found that approximately 20% of the general population suffer from at least one psychiatric disorder (Freedman, 1986), the level of psychopathology in the cult joiner population may not be much greater than that of the population as a whole. Cult joiners may simply be more willing to seek help, which could possibly contribute to their susceptibility to cultic recruitment. Moreover, even if on the whole cultists are more disturbed psychologically, a majority appear to have been within the normal range psychologically before they joined their group.

Whatever the nature, magnitude, and causes of cult-related harm, helping professionals must try to treat it as best they can. Their effectiveness is difficult to evaluate, for only three studies have investigated treatment/intervention outcomes, and two of those examined exit-oriented interventions. None looked at treatment effectiveness with regard to symptomatology or personality dimensions.

Langone's (1984) survey of parents who had attempted deprogrammings of a cult-involved child found that 63% of the deprogrammings were “successful,” that is, the cultist left his/her group. Of the 37% who returned to the group, 25% later left on their own. Dubrow Eichel, Dubrow Eichel, and Eisenberg (1984) reported a 67% success rate with 17 cases at an outpatient facility specializing in the treatment of cult-related problems. “Success” was defined either as the client's deciding to leave the group or, if cult involvement was deemed secondary to a more primary destructive problem, as the client's taking “clear and sincere action to remedy this problem” (p. 158). Spero (1982) completed psychodynamically oriented psychotherapy with 51 of 65 former cultists. A number of indices of improvement were noted. IQ scores increased. Bender Gestalt scores demonstrated a “significant trend toward increased perceptual openness" (p. 338). Responses to the Draw-a-Person Test indicated a decline in emotionality suggestive of conflict. Of the 51 subjects who completed therapy, 5 consulted their therapist prior to the scheduled six-month followup, in which 42 subjects participated.

In this study we add to the limited amount of systematically collected data bearing on post-cult distress and its treatment. More specifically, the study's purposes are to:

1. examine the nature and level of distress in former cult members

2. evaluate changes in distress variables occurring between initiation of a post-cult residential treatment intervention and a six-month follow-up

A future report will examine the question of whether distress is related to other factors, in particular certain demographics, pre-cult experiences with parents, pre-cult indicators of psychopathology, mode of exit from the group, and physical and/or sexual coercion within the group.

Methods

Sample

Two groups of subjects were studied. Group one consisted of Ill clients of Wellspring Retreat and Resource Center. Wellspring provides a "bed-and-breakfast” type of setting where clients receive daily outpatient counseling and workshops. Referrals come from cult educational organizations throughout the world, pastors and mental health professionals, as well as a significant number of self-referrals. Clients enter voluntarily and are free to leave at any time.

Group two consisted of 13 former cultists who attended an educational workshop sponsored by FOCUS, a support organization for ex-cultists.

Demographic data on the two groups are provided in Table 1.

Table I

Demographics - All Subjects

Variable Wellspring Focus

Frequency (%) Frequency (%)

Sex

Male 44 (40%) 8 (62%)

Female 67 (60%) 5 (38%)

Marital Status

Single 87 (79%) 6 (46%)

Married 17 (15%) 6 (46%)

Divorced 6 ( 5%) 1 ( 8%)

Unspecified 1 ( 1%) 0

Race

Caucasian 104 (94%) 13 (100%)

Asian 4 ( 4%) 0 ( 0%)

Unknown/other 3 ( 3%) 0 ( 0%)

Age (Mean=26.6) (Mean=29.8)

<14 1 ( 1%) 0

14-18 9 ( 8%) 0

19-25 53 (48%) 3 (23%)

26-35 35 (31%) 8 (62%)

36-50 8 ( 7%) 2 (15%)

50+ 2 ( 2%) 0

Unspecified 3 ( 3%) 0

Table 1 (Continued)

Demographics: All Subjects

Variable Wellspring Focus

Frequency (%) Frequency (%)

Religious Upbringing

Protestant Fund. 23 (21%) 6 (46%)

Protestant Liberal 29 (26%) 7 (54%)

Catholic 31 (28%) 0

Jewish 4 ( 4%) 0

Eastern 8 ( 7%) 0

Other 10 ( 9%) 0

Unspecified 6 ( 5%) 0

Relationship to Mother

Good 55 (49%) 5 (38.5%)

Fair 34 (31%) 5 (38.5%)

Poor 17 (15%) 3 (23%)

Unspecified 5 ( 5%) 0

Relationship to Father

Good 42 (38%) 5 (38.5%)

Fair 34 (31%) 5 (38.5%)

Poor 25 (22%) 3 (23%)

Unspecified 10 ( 9%) 0

Years in the Group (Mean=3.6 yrs.) (Mean=5.5 yrs.)

<1 28 (25%) 3 (23%)

1-3 45 (40%) 2(15.5%)

4-7 16 (14%) 5 (38%)

8-12 5 ( 5%) 2 (15.5%)

13+ 5 ( 5%) 1 ( 8%)

Unspecified 12 (11%) 0

College

<3 years 65 (59%) N/A

Bachelor’s Degree 36 (33%)

Master’s Degree 5 ( 4%)

Unspecified 5 ( 4%)

Method of Exit

Exit Counseling 49 (44%) 2 (15.4%)

Deprogramming 28 (25%) 2 (15.4%)

Walkaway 25 (23%) 9 (69.2%)

Unspecified 9 ( 8%) 0

Group Type

Fundtl.Aberrational 54 (49%) 4 (31%)

Charismatic 21 (19%) 4 (31%)

New Age/Eastern 8 ( 7%) 2 (15%)

Other Bible-based 10 ( 9%) 2 (15%)

Other 18 (16%) 1 ( 8%)

Pre-Cult Counseling

No 13 (10%) N/A

(22% of those who answered question)

Yes 47 (37%)

(78% of those who answered the question)

Unknown 66 (52%)

Measures

Milton Clinical Multiaxal Inventory. The Million Clinical Multiaxial Inventory ([MCMI-I], Choca, Shanley & Van Denburg, 1992) is a self-report inventory designed to assess personality and clinical symptoms of psychiatric patients. It is currently considered to be an extremely popular instrument of this genre (Piotrowski & Keller, 1989, Piotrowski & Lubin, 1989, 1990). The score used for the MCMI is the Base Rate (BR), A BR of 35 was established as the median score for normal or nonpsychiatric populations. A BR of 60 was set as the median for psychiatric populations. BRs of 75 or higher were considered anchor points for the presence of a particular disorder. A BR of 85 or higher was defined as representing the most predominant characteristic. Thus, any score over a BR of 75 would indicate a high probability that a subject would indeed possess the disorder represented by the scale in question. On the MCMI-1, all BR scores are adjusted to take into consideration the likelihood that subjects would tend to either deny or inflate reporting of certain personality or emotional symptoms.

The MCMI is distributed in other countries and has been translated into several languages (Luteijn, 1990, Simonsen & Mortensen, 1990). Test-retest reliabilities range from the low 60s to the low 90s (Choca et al., 1992). As to validity, Gibertini, Brandenburg, and Retzlaff (1986) noted that the MCMI-I scales varied widely as to their usefulness in accurately pinpointing diagnosis. Positive predictive power (PPP) ranged from 19% to 84%. The PPP index is determined

by the magnitude of the sensitivity and specificity of the test and the prevalence of the disorder in the population. When the sensitivity and specificity of the test are very high (e.g., 90%), the PPP and the NPP [negative predictive power] indexes are optimal. However, as prevalence decreases, so does PPP. In populations with very few disordered cases, even tests with high specificity and sensitivity can have low predictive power. Overall diagnostic power, an index representing the proportion of correct classifications, also varies in its usefulness as disorder prevalence rates vary. (Choca et al., 1992, p.38)

The MCMI includes the following clinical symptom scales: Anxiety, Somatoform, Hypomanic, Dysthymic, Alcohol Abuse, Drug Abuse, Psychotic Thinking, Psychotic Depression, and Psychotic Delusions. The following scales reflect more enduring personality traits than the symptom scales: Schizoid (Asocial), Avoidant, Dependent (Submissive), Histrionic (Gregarious), Narcissistic, Antisocial (Aggressive), compulsive (Conforming), and Passive-Aggressive (Negativistic).

With the MCMI-1, Gibertini et al. (1986) found that eight scales were rated as having good PPP, that is, 70% or more: Avoidant, Dependent, Histrionic, Negativistic, Borderline, Anxiety, Dysthymia, and Drug Abuse. Nine scales were rated fair (50% - 69%): Schizoid, Narcissistic, Antisocial, Compulsive, Schizotypal, Paranoid, Somatoform, Hypomania, and Alcohol Abuse. Three scales were rated poor (below 50%): Psychotic Thinking, Psychotic Depression, and Psychotic Delusions. The scales were also rated according to their ability to predict the 'most predominant syndrome' and the results showed that five scales were rated as good: Avoidant, Schizotypal, Paranoid, Anxiety, and Dysthymia.

Other studies have shown the scale to have rather high negative predictive power (NFP), that is, low test scores were rarely false negatives (Gibertini et al., 1986). These investigators also found that the MCMI has fairly high diagnostic power. Choca et al.'s (1992) review of the MCMI found a number of studies that are critical of the MCMI and its correspondence to DSM-III diagnosis. As a result the MCMI-II was revised according to clinical judgments based on DSM-III-R diagnostic classifications (Millon, 1987). Accordingly, Wellspring has been using the MCMI-II since shortly after its publication. The results from the MCMI-II will be presented in another study. Nevertheless, the bulk of studies on the MCMI-I, which was used in this study, do support it as reliably measuring personality traits and symptomatology.

A review of the literature found no study which used the MCMI as a measure of distress among former sect or cult members, although Weiss and Comrey (1987) reported strong compulsive traits among members of the Hare Krishna group. Several studies, most done on psychiatric inpatients, have reported changes on MCMI scores after a treatment program, including increases in the narcissistic and histrionic scales on the MCMI following treatment (Libb, Stankovic, Sokol, Freeman, Houck & Switzer, 1990; McMahon, Davidson, & Flynn, 1986; McMahon, Flynn, & Davidson, 1985; Stankovic, Libb, Freeman, & Roseman, 1992). A number of studies have reported on correlations for the MCMI-I prior to treatment and after treatment (Hyer, Woods, Bruno, & Boudewyns, 1989; Libb, et al., 1990; McMahon et al., 1985; Millon, 1987; Overholser, 1990; Piersma, 1986). Typically, stability coefficients for the clinical scales range between r = .60 and r = .70, and for the personality scales more than r = .70.

Beck Depression Inventory. The Beck Depression Inventory ([BDII; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) is one of the most commonly used self-report methods of assessing depression. The BDI is symptom focused. Split-half reliabilities range from .58 to .93; test-retest ranges from .69 to .90. Concurrent validity has been demonstrated with diverse measures. BDI scores range from 0 to 63, with a score of 10 or more considered to be beyond the normal range and a score of 17 or more suggestive of depressive disorder. The BDI has been used with a wide variety of psychiatric and normal populations but has been criticized for being a measure of a social undesirability response set and for correlating more closely to an anxiety measure than to another depression measure (Langevin & Stancer, 1979, cited in Beckham & Leber, 1985).

Hopkins Symptom Checklist. The HSCL is a frequently used, self-administered rating scale for assessing psychiatric symptomatology (Derogatis, Lipman, Rickels, Uhlenhuth, & Covi, 1974). Reliability and validity studies have consistently shown the HSCL to be an instrument possessing high degrees of both reliability and validity. Internal consistency measures are uniformly high for all five factoral dimensions, that is, .84 to .87. Test-retest reliability studies show uniform and high consistency (Rickels, Lipman, Park, Covi, Uhlenhuth, & Mock, 1971). Validity studies demonstrating clinical sensitivity have shown the USCL to be an efficient and dependable measure of anxiety and other clinical psychiatric symptoms (Derogatis et al., 1974; Rickels, Lipman, Garcia, & Fisher, 1972). The HSCL is a 58-item inventory with scores ranging from 58 to 232. One hundred is considered a cutoff score indicating the need for psychiatric treatment.

Staff Burnout Scale. The SBS-HP scale is based on the conceptual framework developed by Maslach and Pines (1979). However, the SBS-HP also includes behavioral and physiological indices of burnout. The SBS-HP scale is published by London House management consultants. It was designed to measure acute stress episodes and was normed on populations that have been traditionally considered high-stress professions, namely health professionals. The scale is a self-report measure consisting of 30 questions. Ten of the questions were designed to detect lying. With the remaining 20 questions a reliability coefficient of .93 was obtained (Jones, 1980). Three validity studies showed significant correlations with absenteeism, tardiness, job turnover, physical illness, patient neglect, employee theft, drug and alcohol abuse, job dissatisfaction, and perceived levels of stress (Jones 1980, 1981). Martin (1983), found the mean SBS-HP score to be 48.75 (N = 488, SD = 18.61) for religious workers in high schools and colleges, for example, Campus Crusade for Christ, Young Life, and so forth. Scores greater than 70 indicate burnout; scores greater than 140 reflect severe burnout.

Clinical Interviews. Two semi-structured interviews developed at Wellspring were employed in this study. The first inquires into family of origin, general relationships, occupational history, sexual attitudes and practices, mental status, suicide potential, depression, intelligence, and memory function. The second interview deals with the cult experience: personality changes, why the person left, changes in affect related to cult involvement, changes in family relationships, years in group, and so forth.

Procedures

The interview and test battery described in the previous section are given routinely to new clients at Wellspring. Because some clients are especially confused or exhausted, not everyone completes the battery during the first day of their stay. As part of its followup with clients, Wellspring administers the MCMI approximately six months after a client completes the Wellspring program. Sixty-six of the 111 subjects completed the followup test battery.

Thirteen members of FOCUS, the ex-member support group, completed the MCMI at an ex-member workshop. No other measures were administered to this population, with the exception of the demographics questionnaire.

Analysis

Three MANOVA's were run on the Millon scales to determine: (1) differences, if any, between ex-cultists who sought treatment (Wellspring clients) and those who did not (FOCUS group); (2) differences in scores on the MCMI between those Wellspring clients who responded to the followup survey and those who did not; (3) differences between the MCMI scores at the beginning of the Wellspring treatment program (Time 1) and the retest that occurred approximately six months after treatment (Time 11). Means, medians, and standard deviations were also calculated and tabulated for the variables under study, by group. Data were analyzed by Statistical Package for the Social Sciences (SPSS, Inc. Staff, 1990).

Results

Comparison of Subject Samples

The MANOVA showed no significant difference on MCMI Time I scores between those who received treatment at Wellspring and the FOCUS members [F = (1,11) = 1.00, p = .450]. There was also no significant difference at Time I between those who did and those who did not respond to the followup survey [F (1,11) = .695, p = .741].

Table 2 provides the post-cult means and standard deviations of the Wellspring and FOCUS groups on the MCMI and the Wellspring group's scores on the Beck Depression Inventory, Hopkins Symptom Checklist, and Staff Burnout Scale for Time I. Both groups showed elevations on Anxiety, Dysthymia, and Somatoform. The Wellspring group had an elevated score on Dependent. The apparently lower Dependent score of the FOCUS group is intriguing, and univariate analysis found a statistically significant difference [F = (1,122) = 4.43, p = .037].

Nature and Level of Distress

Table 3 gives the means and standard deviations on all of the MCMI scales for Time I and Time II for those Wellspring subjects who participated in both testings. Anxiety has a Time I mean score of 74, Dysthymia 69, and Dependent 72. Two scales, Dependent and Anxiety, have Time I population median scores with BR's above 75 (both have median scores of 79). Dysthymia has a Time I BR median of 73.

Millon (1987) assessed the accuracy of the cutoff score of BR 75 on a sample of 256 patients that closely matched the sample used in constructing the MCMI-1. He found that the percent of correct diagnostic classifications ranged from 82% to 94%. Thus, a majority of this sample of ex-cultists could be classified as having, with a high degree of probability, a dependent personality disorder and an anxiety disorder.

Another general interpretive rule with regard to the MCMI-I is that a BR score above 60 may have value in providing diagnostic information, while scales below 60 as a rule are not useful in providing diagnostic information, although low scores do not mean the absence of the particular characteristic or symptom. Thus, this sample of ex-cultists can be characterized as having abnormal levels of distress in several of the personality and clinical symptom scales. Of those subjects completing the MCMI-I, 89% had BR's of 75 or better on at least one of the first eight scales. Furthermore, 106 out of the 111 subjects (95%) who completed the MCMI at Time I had at least one BR score of 75 or comp higher on one of the MCMI scales.

Table 2

Post-cult Means and Standard Deviations: Wellspring (n = 111) vs. FOCUS (n = 13)

Variable Wellspring FOCUS

Mean S.D. Mean S.D.

Schizoid 44 26 41 25

Avoidant 54 27 44 24

(1) Dependent 72 26 56 28

Histrionic 61 22 58 25

Narcissistic 55 21 57 18

Antisocial 43 24 50 17

Compulsive 62 18 65 26

Negativistic Aggress 51 30 47 31

Schizotypal 55 16 52 14

Borderline 62 19 63 18

Paranoid 54 17 61 16

(1) Anxiety 76 24 81 19

Somatoform 71 17 78 16

Hypomania 45 30 41 32

(1) Dysthymia 72 24 74 19

Alcohol Abuse 45 22 31 24

Drug Abuse 47 24 47 20

Psychotic Thinking 54 14 51 13

Psychotic Depression 52 16 49 14

Psychotic Delusions 52 17 53 19

BDI (n = 98) 14 20 N/A

SBS-HP (n = 46) 72 30

HSCL (n = 42) 102 30

1. Median scores of the Wellspring group for Dependent, Anxiety, and Dysthymia were 79, 79, and 75, respectively.

Table 3

Means and Standard Deviations: Pre- and Post-Treatment, Wellspring (n = 66)

Variable Pre-Treatment Post Treatment

Mean S.D. Mean S.D.

Schizoid (Time 1) 43 26 30 20**

Avoidant 52 27 35 26**

(1) Dependent 72 27 52 27**

Histrionic 62 27 69 21**

Narcissistic 58 21 64 21*

Antisocial 46 23 53 20**

Compulsive 63 17 65 13

Negative Aggress 48 28 37 26**

Schizotypal 53 14 44 18**

Borderline 59 18 48 19**

Paranoid 55 18 55 14

(1) Anxiety 74 22 59 24**

Somatoform 69 17 62 16**

Hypomania 46 29 41 27

(1) Dysthymia 69 24 55 24**

Alcohol Abuse 49 20 34 19**

Drug Abuse 48 22 48 21

Psychotic Thinking 54 12 42 20**

Psychotic Depression 51 15 41 19**

Psychotic Delusions 52 17 48 17

*= Univariate F-tests with (1,65) D.F.significant at .05 level (pre-post)

**= significant at .01 level or higher

1. Dependent, Anxiety, and Dysthymia had pre-treatment median scores of 79, 78, and 73, respectively

The contention that this population of former cultists is indeed distressed is further buttressed by the sample's mean score of 102 on the HSCL. Typically, scores of over 100 are indicative of the need for professional psychiatric care. Moreover, these ex-cultists had a mean of 72 on the SBS-HP burnout scale, which is suggestive of burnout and more than one standard deviation above the mean from Martin's (1983) sample of parachurch workers.

Table 4 shows the percentage of individuals who actually scored BR's of 75 or higher. Over 50% of the population had a ER of 75 or higher on either Dependent, Anxiety, or Dysthymia.

Effects of Abuse

In addition to the variables listed in Table 4, the investigators inquired into whether or not clients had been physically and/or sexually abused in the group. Of 54 respondents to these questions, 7 subjects were sexually abused, 5 physically abused, and 5 others both physically and sexually abused. Subjects' scores on the eight highest of the MCMI scales (Dependent, Histrionic, Narcissistic, Obsessive-Compulsive, Aggressive-Negativistic, Anxiety, Somatoform, Dysthymia) were examined to determine if there was any effect from physical or sexual abuse. The MANOVA was not significant, F (2, 16) = .416, p < .975. Neither did univariate ANOVAs approach significance.

Pre-Post Changes

Table 3 reveals that there were statistically significant differences between Time I and Time 11 scores on 15 Millon scales. The MANOVA (effect-time) did reveal highly significant mean differences between the MCMI-I pre- and post-treatment scores, Hotellings F (1, 20) = 4.11, p<.001. The only scales that did not show a significant effect change, were Obsessive-Compulsive, Paranoid, Hypomania, Drug Abuse, and Psychotic Delusions. All scales showing significant change moved downward from Time I to Time 11, except for Histrionic, Narcissistic, and Antisocial, which moved up. The elevation in these three scales in a treatment study is not unusual in that successful treatment tends to increase self-esteem, self-expressiveness, and to disinhibit inappropriate self-restraints (McMahon et al., 1986). Among those scales that declined, the decline ranged from .4 to 1.0 standard deviations. Psychotic Thinking declined from 54 to 42, a drop of one standard deviation. Dependent declined from 72 to 52, or .74 SD. Somatoform and Negativistic-Aggressive declined about .4 SD and Schizoid about .5 SD. The other scales all declined between .6 and .7 SD. The effect size measures (partial eta squared) on scales with the highest BRs at Time I were as follows: Dependent (.339), Somatoform (.100), Anxiety (.183), Histrionic (.138), Narcissistic (.079), Compulsive/Conforming (.014), Passive Aggressive/Negativistic (.152), Dysthyrnia (.172). All ANOVAs were significant, except for Compulsive/Conforming.

Table 4 Percentage of Subjects with a Base-rate Score of 75 or Higher on MCMI-I Scales for Pre- and Post-treatment (n = 66)

MCMI-I Scale Group

Pre-Treat. Post-Treat.

Schizoid-Asocial 17.9% 1.5%

Avoidant 23.9 7.5

Dependent-Submissive 58.2 28.4

Histrionic-Gregarious 32.8 43.3

Narcissistic 16.4 31.3

Antisocial-Aggressive 3.0 1.5

Compulsive-Conforming 10.4 14.9

Psv. Agg.-Negatvst. 25.4 11.9

Schizotypal-Schizoid 3.0 1.5

Borderline-Cycloid 13.4 9.0

Paranoid 13.4 3.0

Anxiety 52.2 26.9

Somatoform 35.8 19.4

Hypomanic 7.5 6.0

Dysthymia 47.8 25.4

Alcohol Abuse 6.0 1.5

Drug Abuse 9.0 4.5

Psychotic Thinking 4.5 0.0

Psychotic.Depression 0.0 4.5

Psychotic Delusion 4.5 0.0

Table 4 also presents the frequency and percent of subjects with BR'S above 75 on the MCMI-I scales prior to treatment and after treatment. With the exception of the Histrionic, Narcissistic, Antisocial-aggressive, and Psychotic Depression scales, the percentage of subjects with BR's above 75 drops by as much as over 50%. The rise in the above listed personality scales is not unusual and has been documented in other studies as a positive treatment effect (McMahon et al., 1986).

Profile Stability

Of the 66 subjects who took the MCMI-I at Time I and Time II, 47 out of 66 subjects (71.2%) demonstrated a personality profile shift, as indicated by changes on the two highest points on the first eight scales from Time I to Time II. At Time I, 32 of 66 (48.4%) subjects had Dependent as the highest scale. At Time II, 26 of the 32 subjects (81.2%) with Dependent as the highest scale showed entirely different profiles, that is, the two highest scales were not in the same rank order after treatment. Indeed, in the majority of cases the highest Time I scales were no longer the highest scales at Time II. For example, of those subjects for whom Dependent was the second highest personality scale score at Time 1, 5 out of 5 (100%) had entirely different profiles on the two highest scales at Time II. The overall stability of the two highest scales for Time I to Time II was only 22%. This contrasts to Hyer et al.'s (1989) study showing pre- to post-treatment stability of 84% on the two highest personality scales.

In the current study only one scale, Drug Addiction, had a Time I to Time II coefficient above r = .60 (see Table 5). Six of the first eight personality scales had the lowest coefficients reported in the literature on the MCMI. On the clinical scales the current study reported the lowest coefficients in five of the nine scales. The overall average for both the personality, clinical, and total MCMI scales was the lowest reported in the literature thus far.

Table 5

Test-Retest Correlations (Stability Coefficients) of MCMI-I Scores Compared to Those of Other Studies

Scales Investigators This Study

Piersma Libb Over- Millon McMahon Hyer

(1986) et al. holser (1872) et al.

(1990) (1990) (1983) (1989)

1 56 77 71 86 82 69 65 82 63 38

2 56 65 76 80 84 70 70 80 63 56

3 65 49 67 45 79 58 61 79 75 50

4 75 81 80 86 85 83 87 82 62 53

5 61 51 87 80 81 61 71 88 71 50

6 55 62 85 79 83 63 72 79 79 42

7 56 36 44 66 77 70 70 77 71 30

8 48 19 61 55 81 61 54 69 72 53

9 57 40 75 83 78 65 74 76 65 49

C 27 27 64 54 77 50 42 70 33 30

P 46 66 51 91 77 66 32 38 67 58

A 31 15 68 44 61 72 43 2

H 21 37 62 40 45 72 53 14

N 75 78 65 67 62 78 61 58

D 32 18 66 44 57 70 58 16

B 54 54 76 57 22 74 46 56

T 75 69 74 70 41 74 68 62

SS 40 69 68 63 51 72 60 48

CC 52 35 61 55 61 61 38 21

PP 58 73 66 69 44 61 67 53

Average (all scales)

52 49 78 61 61 72 61 39

Average (scales 1-8)

59 55 71 72 82 67 69 80 69 47

Averages (scales A-PP)

49 44 67 57 49 69 55 36

Reliability of Measures

Some may argue that the MCMI scores on Anxiety and Dysthymia may not be entirely valid and, consequently, not capable of measuring levels of distress in a clinically sensitive manner. However, the Wellspring respondents to the MCMI also took the Beck Depression Inventory and the Hopkins Symptom Checklist, which correlated significantly and moderately with MCMI levels of Anxiety and Dysthymia (See Table 6).

Table 6 Correlations Between tire MCMI and Other Measures

Measure Anxiety Somatoform Dysthymia

BDI(N = 98) .50 (p< .000) .35 (p<.000) .54 (p<.000)

HSCL (N = 42) .59 (p < .000) .43 (p<.002) .58 (p<.000)

Discussion

Nature and Level of Distress

The former cultists in this study clearly experienced clinical levels of distress, as evidenced by their scores on the MCMI-I, the BDI, the HSCL, and the SBS-HP. That the Wellspring population, which was a selected sample of people seeking assistance, scored high on distress measures is not surprising. That the FOCUS group scored at a comparable level is unexpected and may have important implications for the treatment of ex-cultists (to be discussed later).

The FOCUS and Wellspring populations' mean scores were between one-half and one standard deviation above the inpatient median on the MCMI for Anxiety, Dysthymia, and Somatoform. Dependent, Histrionic, Compulsive, and Borderline scales were at or above the median scores for inpatient populations. Overall, this population of ex-cultists is as distressed as some psychiatric inpatients. Nevertheless, the investigators were surprised that scores on the MCMI clinical scales were not higher, in part because 25% reported physical and or sexual abuse while in the cult. Yet these abused persons were no more distressed, as measured by the MCMI clinical scales, than those not experiencing physical or sexual abuse. Consequently, these researchers suspect that unique aspects of distress may be more accurately detected by using measures of dissociation in conjunction with clinical measures such as the MCMI.

Indeed, the collective profiles for these populations appear to be sufficiently dissimilar to established clinical profiles as to warrant labeling as a distinct syndrome. Moreover, subjects' profiles looked at individually regularly show distinct elevations of Dependent and Compulsive personality disorder scales.

The evidence of high symptom distress and personality disorder is consistent with clinical reports about this population and inconsistent with reports that portray cultic involvement as essentially benign or, as Levine (1984) says, “in the end therapeutic, although excesses can occur” (p. 27).

Research alluded to earlier suggests that in the years before joining a cult approximately one third of cult joiners had been sufficiently distressed to seek counseling at some time. Clinical reports (Clark, 1979; Singer 1978) indicate that although a sizeable minority of cultists were psychologically troubled before joining cults, a majority appear only to have been experiencing normal developmental crises prior to joining. Several studies found that current cultists appear to be psychologically normal (Galanter, 1989; Levine, 1984; Ross, 1983; Ungerleider & Wellisch, 1979), although MMPT results in some studies evidenced elevated Lie Scales (Ross, 1983; Ungerleider & Wellisch, 1979). Wright (1987) and Skonovd (1983) reported that former cultists were subjected to intense pressure to remain in their groups. Yeakley's (1988) studies with the Myer-Briggs Type Indicator (MBTI) suggest that cultists are subjected to great pressures toward conformity. Yet this study, clinical reports (Ash, 1985; Clark, 1979; Singer, 1979), and Galanter's (1983) research strongly suggest that the level of post-cult distress is quite high. These findings cause Langone to comment:

And yet the majority eventually leave [Barker, 1984]. Why? If they were unhappy before they joined, became happier after they joined, were pressured to remain, left anyway, and were more distressed than ever after leaving, what could have impelled them to leave and to remain apart from the group?

The inescapable conclusion seems to be that the cult experience is not what it appears to be (at least for those groups that deem it important to put on a “happy face”), either to undiscerning observers or to members under the psychological influence of the group. Clinical observers, beginning with Clark (1979) and Singer (1978), appear to be correct in their contention that dissociative defenses help cultists adapt to the contradictory and intense demands of the cult environment. So long as members are not rebelling against the group's psychological controls, they can appear to be ”normal,” much as a person with multiple personality disorder can sometimes appear to be “normal.” However, this normal-appearing personality, as West (1992) maintains, is a pseudopersonality. When cultists leave their group, the flood gates open up and they suffer. But they don't generally return to the cult 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. (Langone, in press)

This study supports the clinical observations indicating that powerful cultic environments distort personality (or, as West says, create pseudopersonalities) and induce dependency. The disparity between the benign appearance of many cults and the harsh underlying reality appears in bold relief when one examines certain individual cases. Subjects in this study experienced various forms of abuse. Several subjects were repeatedly beaten. Others had guns held to their head. Some were put on starvation diets. Some adults were held in their rooms as punishment. Several women were raped. Several leaders, who presented as “pastors,” “trained" some female subjects to be “good women” or “good wives” by subjecting them to a program of diverse sexual “exercises,” which included group sex, oral sex, baring one's breasts, intercourse, and, in one case, being urged to have sex with an animal. One man reported that when he was a boy his pastor told him that God willed that the boy should please the “pastor” by regularly performing fellatio on him.

Pre-Post Changes

The pre-post MCMI changes in this study further support the contention that dissociation is central to the cult experience. The striking change in the personality scales and the improvement in MCMI symptom scales between admission and followup of Wellspring clients is inconsistent with the hypothesis that post-cult personality profiles and distress levels reflect long-standing, pre-cult psychological difficulties. Treatment evaluation studies using the MCMI simply do not indicate both marked personality changes and marked symptom reduction. Indeed, this study reported the lowest test-retest correlations of any treatment study using the MCMI. Rather than concluding that Wellspring's treatment program is miraculous, or that the MCMI scales are very unstable, we conclude that the admission personality profile and distress level do not, in a general sense, reflect longstanding personality traits.

On the personality scales of the MCMI, a majority of Wellspring clients have Dependent as the highest or second highest personality scale. Thirty-two of 66 (48.4%) clients had Dependent as the highest scale. Another 5 clients had Dependent as the second highest scale (7.5%). And yet after treatment only 14 of 66 clients had Dependent as the highest (21.2%) and 3 had Dependent as the second highest (4.5%) scale. Thus, after treatment 17 (25.8%) of the clients had Dependent as the first or second highest score on the first eight scales of the MCMI-I, compared to 37 (56.1%) before treatment. In comparison, 24 of 66 (36.4%) of the clients had Compulsive-Conforming as the first or second highest scale at Time I, whereas at Time II 25 of the 66 subjects (37.8%) still had Compulsive-Conforming as the first or second highest scale.

There are only three possible explanations for these elevations: (1) cultic groups and/or helping organizations tend to attract dependent-compulsive types, who for some unknown reason achieve remarkable decreases in Dependent after 10-14 days of treatment; (2) cultic groups tend to produce and/or exacerbate dependent-compulsive types; or (3) cultic groups tend to harm dependent-compulsive types, who as a result of being harmed are more likely to seek help from cult educational organizations. Regardless of which of these three explanations or combination thereof is true, something sets the cultic environment apart from other environments.

The change from pre- to post-treatment personality configuration, particularly the striking drop in Dependent, supports explanation two. If the admission Dependent score reflected a true personality dimension, one would not expect the dramatic change that occurs after a treatment program averaging 10 to 14 days. One would expect such a change, however, if the clinically derived theories of induced dependency as the distinguishing feature of cultic environments were correct. Elevated Dependent scores at admission would reflect an artificially induced personality change that, as clinical lore suggests, would respond readily to intervention. This point of view is supported by Yeakley's (1988) finding that cult members' scores on the MBTI converge toward a single type. The distribution of type scores in non-cultic groups, on the other hand, was normal and showed no tendency toward convergence, whereas in cultic groups “present and future distributions deviate increasingly from the normal distribution” (p. 34). Therefore, it seems reasonable to conclude that pressures toward increasing dependency may exist in the cult population.

The overall lack of personality profile stability as measured by test-retest correlations (see Table 5) suggests that other aspects of cultists' personalities are also likely to change while in the cult and to reorganize dramatically after leaving and seeking treatment. Further study is clearly needed.

Implications for Treatment

That symptom scales such as Anxiety and Depression decrease substantially after treatment is gratifying but not surprising. The Wellspring program is intensive, consisting of 4-6 hours daily of individual counseling, group counseling, and didactic workshops. In addition there are many informal discussions with other former cultists and the staff (of which a majority themselves are former members), as well as time for independent study, reading, and viewing or listening to Wellspring's collection of video- and audiotapes related to cultic phenomena. Typically the program lasts 14 days. In addition to the therapy and didactic hours, there is adequate time for rest and recreation, such as swimming hiking, fishing, using the Ohio University athletic facilities, movies, malls, and restaurants.

The program is intensive, but also specialized. Much of the early segment of the treatment program is geared to increasing ex-members' understanding about the dynamics of thought reform and helping them explore how their group practiced thought reform and how the thought reform program affected their personality and their relationship to the world outside the cult. Thus, Wellspring uses a model that assumes that the cultic environment adversely affects members. At its core it is educational, even though much therapeutic activity occurs.

One might argue that Wellspring's employment of a thought reform model as a conceptual tool for understanding cult involvements may produce an artificial decrease in Dependent scores at Time 11. However, other clinical populations in which Dependent is high do not show so large a drop as the Wellspring clients did. Moreover, for the FOCUS group, Dependent and years out of group varied inversely (a subsequent report will go into greater detail on this point).

The implications for treatment are noteworthy. Some of the issues suggested by this study relate to the nature of dissociation itself and how it is associated with trauma. If indeed these former cultists are in a partial state of dissociation as a result of their cult involvement, then this dissociation must be treated. Our clinical impressions of most of the clients indicate that they are split off from their emotional responses to the various traumas described to the clinician.

Wellspring's approach to treating the dissociation begins therapy by reconstructing the client's experiences in terms of a systems model of thought reform. Typically, clients' awareness of what happened to them is restricted because they lack a conceptual framework that can adequately attach meaning to their experiences. In this regard, Wellspring's treatment, at least in the beginning phases, is similar to certain methods of treating victims of trauma or sexual abuse. In treating trauma and abuse, the nature of the experience must be explained in order to prevent victims from blaming themselves. Similarly, former cultists are able to free themselves of the cult-imposed tendency to blame their own inadequacy for personal problems and to recognize the role of forces within the cult environment. As Langone (1992) argues, behind such treatment lies an unavoidable ethical imension. In order to regain their former level of psychological well-being, which includes their sense of right and wrong, former cultists must come to understand not only what was done to them and how it was done but also why it was wrong. To ignore this ethical dimension is to ignore one of the central elements of self.(Ofshe & Singer, 1986) that cults assault and that is dissociated from consciousness.

Very much tied to the sense of right and wrong are religious concerns, which often preoccupy many former cultists, in part because, especially for those who were in religious cults, they have used religious frameworks to understand the world since at least the time of joining the cult. Addressing these concerns can be a daunting task for many mental health clinicians, and even for pastoral counselors. Frequently, former cultists need to adopt new religious frameworks, or return to their pre-cult religious belief systems, in order to recapture dissociated parts of themselves, understand the cult experience, cope with present challenges, and tie all of these issues together. Typically, psychological and theological analyses cannot be cleanly separated.

As ex-cultists come to understand the mechanisms operating in the cultic environment, they become more capable of effectively grieving the loss of friends, time, career pursuits, idealism, and other aspirations that were lost as a result of spending time in and leaving the cult. Supportive counseling reinforced by educational programs regarding grief and loss appears to be effective in helping ex-cultists deal with grief-related problems.

Graduates of the Wellspring program frequently say that the daily treatment regime in a quiet, rural milieu with understanding staff is a source of healing and comfort. This temporary, supportive time of reflection enables former cultists to reconnect to dissociated elements of self and to assess their lives in more detail than would be possible in a hectic environment. The clarity thus gained permits them to plan more effectively for their time after Wellspring.

Future Research

That the FOCUS group exhibited as much distress as the Wellspring group at Time I suggests that the need for treatment among ex-members may be greater than most people realize. It is possible, of course, that the FOCUS group may not he representative of ex-members who don't receive treatment and that Wellspring clients may turn out to be more distressed than other ex-cultists. On the other hand, ex-cultists who do not receive an intense educational-therapeutic intervention as occurs at Wellspring may not be able to recover as rapidly and as thoroughly as Wellspring clients. Longitudinal studies of ex-cultists from a variety of groups and exposed to various treatments, as well as no treatment, must be conducted.

The field could also benefit from longitudinal studies (using the measures of this study) that examine changing distress levels during cultists' time in and out of the cult. Pre-cult studies would also be very useful, although these may be extremely difficult to conduct. It win also be useful to conduct studies over time of various ex-cultist populations exposed to different amounts and kinds of treatment, for example, walkaways with no treatment vs. exit counseled subjects with no treatment vs. exit counseled subjects with treatment, etc.

Researchers should also explore the differential effectiveness of various components of the Wellspring program, for example, two weeks of peace and quite in the woods, intensive psychotherapy over a two-week period, didactic groups, etc.

It would also be useful to compare subjects participating in various treatment activities, for example, unstructured ex-member support groups, didactic seminars on thought reform, individual psychotherapy with therapists who do not address thought reform issues, and psychotherapy with cult-aware psychotherapists.

A followup to the current study will examine how certain variables (e.g., pre-cult counseling, years in cult) relate to MCMI scales and other measures. Is there, for example, a relationship between level of dependency and time spent in a cult?

Although this study has produced intriguing findings, it is just a beginning. We know much more than we did 10 years ago. But we still have much to learn.

References

Ash, S. (1985). Cult-induced psychopathology, part I: Clinical picture. Cultic Studies Journal, 2(l), 31-91.

Barker, E. (1984). The Making of a Moonie: Choice or brainwashing? Oxford: Basil Blackwell.

Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561-571.

Beckham, E. E., & Leber, W. R. (Eds.). (1985). Handbook of depression: Treatment, assessment, and research. Homewood, IL: Dorsey Press.

Choca, J., Shanley, L. & Van Denberg, E. (1992). Interpretive guide to the Millon Clinical Multiaxial Inventory. Washington, DC: American Psychological Association.

Clark, J. G. (1979). Cults. Journal of the American Medical Association, 242, 279-281.

Conway, F., Siegelman, J. H., Carmichael, C. W., & Coggins, 1. (1986). Information disease: Effects of covert induction and deprogramming. Update: A Journal of New Religious Movements, 10, 45-57.

Derogatis, L. R., Lipman, R. S., Rickels, K., Uhlenhuth, E. H., Covi, L. (1974). The Hopkins Symptom Checklist (H Sc L): A self-report symptom inventory. Behavioral Science, 19(l), 1-15.

Dubrow Eichel, S. K., Dubrow Eichel, L., & Eisenberg, R. C. (1984). Mental health intervention in cult-related cases: Preliminary investigation of outcomes. Cultic Studies Journal, ](2), 156-166.

Freedman, D. X. (1986). Psychiatric epidemiology counts. Archives of General Psychiatry, 41, 931-933.

Galanter, M. (1983). Unification Church ("Moonie”) dropouts: Psychological readjustment after leaving a charismatic religious group. American Journal of Psychiatry, 140, 984-989.

Galanter, M. (1989). Cults and new religious movements. Washington, DC: American Psychiatric Association.

Galanter, M., & Buckley, P. (1978). Evangelical religion and meditation: Psychological effects. Journal of Nervous and Mental Disease, 166,685-691.

Galanter, M., Rabkin, R., Rabkin, I., & Deutsch, A. (1979). The “Moonies”: A psychological study of conversion and membership in a contemporary religious sect. American Journal of Psychiatry, 136, 165-170.

Gibertini, M., Brandenburg, N. A., & Retzlaff, P. D. (1986). The operating characteristics of the Millon Clinical Multiaxial Inventory. Journal of Personality Assessment, 50, 554-567.

Goldberg, L., & Goldberg, W. (1982). Group work with former cultists. Social Work, 27,165-170.

Hochman, J. (1984). latrogenic symptoms associated with a therapy cult: Examination of an extinct “new psychotherapy” with respect to psychiatric deterioration and "brainwashing.” Psychiatry, 47, 366-377.

Hyer, L., Woods, M. G., Bruno, R., & Boudewyns, P. (1989). Treatment outcomes of Vietnam Veterans with PTSD and the consistency of the MCMl. Journal of Clinical Psychology, 46(3), 2,85-291.

Jones, J. (1980). The staff burnout scale for health professionals. Park Ridge, IL: London House Management Press.

Jones, J. (1981). Staff burnout and employee counterproductivity. In Jones, J. W. (Ed.), The burnout syndrome. Park Ridge, IL: London House Management Press.

Knight, K. (1986). Long-term effects of participation in a psychological "cult" utilizing directive therapy techniques. Unpublished master's thesis, University of California, Los Angeles.

Langevin, R., & Stancer, H. (1979). Evidence that depression rating scales primarily measure a social undesirability response set. Acta Psychiatrica Scandinavica, 59, 70-79.

Langone, M. D. (1984). Deprogramming: An analysis of parental questionnaires. Cultic Studies Journal, 1, 63-78.

Langone, M. D. (1992). Psychological abuse. Cultic Studies Journal, 9(2), 206-218.

Langone, M. D. (Ed.). (in press). Recovery from cults.

Langone, M. D., Chambers, R., Dole, A., & Grice, J. (in press). Results of a survey of ex-cult members. Cultic Studies Journal.

Levine, S. (1984, August). Radical departures. Psychology Today, 21-27.

Libb, J. W., Stankovic, S., Sokol, R., Freeman, A., Houck, C., & Switzer, P. (1990). Stability of the MCMI among depressed psychiatric outpatients. Journal of Personality Assessment, 55, 209-218.

Luteijn, F. (1990). The MCMI in the Netherlands: First findings. Journal of Personality disorders, 4, 297-302.

Martin, P. (1983). An analytical study of the burnout syndrome as it occurs among parachurch professionals. (Doctoral dissertation, University of Pittsburgh, School of Education). University Microfilms International, 300 N. Zeeb Rd., Ann Arbor, MI 48106 (#8411737).

Maslach, C., & Pines, A. (1979). Burnout: the loss of human caring. In A. Pines & C. Maslach (Eds.), Experiencing social psychology (pp. 246-2-52). New York: Knopf (Random House).

McMahon, R. C., Davidson, R. S., & Flynn, P. M. (1986). Psychological correlates and treatment outcomes for high and low social functioning alcoholics. The International Journal of the Addictions, 2](7), 819-835.

McMahon, R. C., Flynn, P. M., & Davidson, R. S. (1985). Stability of the personality and symptom scales of the Milton Clinical Multiaxial Inventory. Journal of Personality Assessment, 49(3), 231-234.

Milton, T. (1983). The Millon Clinical Multiaxial Inventory Manual (3rd edition). Minneapolis: National Computer Systems.

Milton, T. (1987). Manual for the MCMI-II (second edition). Minneapolis, MN: National Computer Systems.

Ofshe, R., & Singer, M. T. (1986). Attacks on peripheral versus central elements of self and the impact of thought reforming techniques. Cultic Studies Journal, 3(l), 3-24.

Overholser, J. C. (1990). Retest reliability of the Millon Clinical Multiaxial Inventory. Journal of Personality Assessment, 55, 202-208.

Piersma, H. L. (1986). The stability of the Millon Clinical Multiaxial Inventory for psychiatric inpatients. Journal of Personality Assessment, 50(2), 193-197.

Piotrowski, C., & Keller, J. W. (1989). Psychological testing in outpatient mental health facilities: A national study. Professional Psychology. Research and Practice, 20, 423-425.

Piotrowksi, C., & Lubin, B. (1989). Assessment practices of Division 38 practitioners. Health Psychologist, 11, 1.

Piotrowski, C., & Lubin, B. (1990). Assessment practices of health psychologists: Survey of APA Division 38 clinicians. Professional ,Psychology. Research and Practice, 21, 99-106.

Rickels, K., Lipman, R. S., Park, L. C., Covi, L., Uhlenhuth, E. H., & Mock, J. E. (1971). Drug doctor warmth and clinic setting in the symptomatic response to minor tranquilizers. Psychopharmacologia, Berl. 20: 128-152.

Rickels, K., Lipman, Garcia, C. R., & Fisher, E. (1972). Evaluating clinical improvement in anxious outpatients. A.M.S. Psychiatry, 128, 119-123.

Ross, M.(1983). Clinical profiles of Hare Krishna devotees. American Journal of Psychiatry, 140, 416-420.

Schwartz, L. L. (1985). Leaving the cults. Update: A Quarterly Journal of New Religious Movements, 9(4), 3-12.

Simonsen, E., & Mortensen, E. L. (1990). Difficulties in translation of personality scales. Journal of Personality Disorders, 4, 290-296.

Singer, M. T. (1978). Therapy with ex-cult members. Journal of the National Association of Private Psychiatric Hospitals, 9, 15-18.

Singer, M. T. (1979, January). Coming out of the cults. Psychology Today, 12, 72-82.

Singer, M. T. (1987). Group psychodynamics. In R. Berkow (Ed.), The Merck Manual of diagnosis and therapy (15th ed.) (pp. 1467-1471). Rahway, NJ: Merck.

Singer, M. T., & Ofshe, R. (1990). Thought reform programs and the production of psychiatric casualties. Psychiatric Annals, 20, 188-193.

Sirkin, M., & Grellong, B. A. (1988). Cult vs. non-cult Jewish families: Factors influencing conversion. Cultic Studies Journal, 5, 2-23.

Skonovd, N. (1983). Leaving the cultic religious milieu. In D. G. Bromley & J. T. Richardson (Eds.), The brainwashing/deprogramming controversy. Sociological, psychological, legal and historical perspectives (pp. 106-121). Lewiston, NY: Edwin Mellen Press.

Solomon, T. (1981). Integrating the “Moonie” experience: A survey of ex-members of the Unification Church. In T. Robbins & D. Anthony (Eds.), In god we trust: New patterns of religious pluralism in America (pp. 275-295). New Brunswick, NJ: Transaction.

Spero, M. (1982). Psychotherapeutic procedure with religious cult devotees. The Journal of Nervous and Mental Disease, 170, 332-344.

SPSS, Inc. Staff. (1990). SPSS reference guide.:SPSS, Inc.

Stankovic, S. R., Libb, J. W., Freeman III, A. M., & Roseman, J. M. (1992). Posttreatment stability of the MCMI-2 personality scales in depressed outpatients. Journal of Personality Disorders, 6(l), 82-89.

Swartling, G., & Swartling, P. G. (1992). Psychiatric problems of ex-members of Word of Life. Cultic Studies Journal, 9(l), 78-88.

Temerlin, M. K., & Temerlin, J. W. (1982). Psychotherapy cults: An iatrogenic perversion. Psychotherapy: Theory, Research, and Practice, 19, 131-141.

Ungerleider, T. J., & Wellisch, D. K. (1979). Coercive persuasion (brainwashing), religious cults and deprogramming. American Journal of Psychiatry, 136, 279-282.

Weiss, A. S., & Comrey, A. L. (1987). Personality characteristics of Hare Krishnas. Journal of Personality Assessment, 51(3), 399-413.

West, L. J., & Singer, M. T. (1980). Cults, quacks, and nonprofessional psychotherapies. In H. I. Kaplan, A. M. Freedman, & B. C. Saddock (Eds.), Comprehensive textbook of psychiatry. Baltimore: Williams & Wilkins.

Wright, S. A. (1987). Leaving cults: The dynamics of defection. Society for the Scientific Study of Religion Monograph Series, Number 7. SSSR: Washington, D.C.

Yeakley, F. (Ed.). (1988). The discipling dilemma. Nashville, TN: Gospel Advocate.

Acknowledgements

The authors wish to thank the American Family Foundation, which funded the data analysis and report-writing phases of this study, the Ohio University Computer Services, and the directors, staff, and clients of Wellspring Retreat and Resource Center for their diligence and cooperation.

* * * * *

Paul R. Martin, Ph. D. is Director of Wellspring Retreat and Resource Center, Chair of the American Family Foundation’s Victim Assistance Committee, and author of Cult-proofing your kid (Zondervan).

Michael D. Langone, Ph. D. is Editor of the Cultic Studies Journal, Executive Director of the American Family Foundation, coauthor of Cults: What parents should know and Satanism and occult-related violence and editor of the forthcoming Recovery from Cults.

Arthur A. Dole, Ph. D. is Emeritus Professor, Psychology in Education Division, Graduate School of Education, the University of Pennsylvania.

Jeffrey Wiltrout is a doctoral candidate in Organizational Psychology at Ohio University.

Cultic Studies Journal, Vol. 9, No. 2, 1992, Page