Cultic Studies Journal, Vol. 14, No. 2, 1997, page 80
Table 2. Group-related Demographic Information (Former Cult Members Only)
Mean SD Range
Age when entered group
Cult 1
Support-
seeking
23.9 8.1 5-39
Cult 2
Non-support-
seeking
25.1 7.1 15-49
Years in group
Cult 1
Support-
seeking
6.6 6.3 .02-25
Cult 2
Non-support-
seeking
5.4 5.5 .08-22
Years since leaving group
Cult 1
Support-
seeking
7.5 5.7 .42-24
Cult 2
Non-support-
seeking
6.7 5.3 1-17
Procedures
Participants in this study were volunteers who were contacted through fliers distributed at
support meetings for former cult members (located throughout the United States), and at a
religiously sponsored psychotherapy clinic in Southern California. Volunteers were invited to
“participate in a research project about the psychological experiences of former members of
philosophical and religious groups ...that will help educate mental health professionals
about the experiences and needs of those who have been involved in abusive groups.”
Interest in participating in the study was expressed by 159 people, who were then sent
questionnaire packets through the mail.
All participants received the same set of guidelines for selecting a group experience on
which to base survey responses. If they had belonged to more than one cult, former cult
members referred to their most distressing cult experience. Those who had not been cult
members referred to their most distressing group experience, either past or current. All
participants clearly indicated whether or not they perceived their reference group to have
been cultic. The surveys took about 90 minutes to complete, and were returned by 104
people (65%). Those who returned a survey were paid 10 dollars. For Cult 1, the return rate
was 74% for Cult 2, the return rate was 50% and for the Comparison group, the return
rate was 59%. The rates of return for Cult 1 and Cult 2 were significantly different (X2 =
12.22 a =.01).
Results
Analysis of the ICE Index was conducted with the primary sample group (Cult 1). One
survey was excluded from analysis as the participant did not complete the ICE Index,
reducing the sample to 76. Item response frequencies were calculated for the 47 items on
the ICE Index. Three items, which were missing 14 or more responses, were subsequently

Cultic Studies Journal, Vol. 14, No. 2, 1997, page 81
excluded. For the remaining 44 items, the average number of missing responses per item
was .89, with a range of 0 to 4. In all 23 cases in which there were missing item values, the
values were replaced with the mean for the item.
Reliability Analyses
Item analysis conducted on the 44 items revealed good internal consistency (Cronbach‟s
alpha =.89). The average item-total correlation was .37. Thirty-eight items had item-total
correlations above .25 and 33 of those had item-total correlations of .30 or more.
Validity Analyses
Using the K-S (Lilliefors) statistic (SPSS, 1992), ICE Index scores were found to be normally
distributed for each group, and were found to meet assumptions for homogeneity of
variance as per Levene‟s test for equality of variances (SPSS, 1992). Therefore, additional
planned analyses were conducted. The ability of the ICE Index to distinguish between cult-
involved and non-cult-involved participants was tested with a 1-way ANOVA, which revealed
significant differences between mean scores for the three groups (F =18.76 df =2 p
.001). Mean ICE Index scores for each group are shown in Table 3. Post-hoc analyses with
t-tests showed significant differences between mean scores for Cult 1 and the Comparison
group (t =6.09 p .001), and between Cult 2 and the Comparison group (t =5.14 p
.001).
Table 3. Extent of Exposure to Cult Experiences, by Sample Group
ICE Index Scores
Sample
Group
Mean SD Range
Cult 1
Support-
seeking
22.4 7.8 3-41
Cult 2
Non-support-
seeking
22.4 7.9 10-38
Comparison 8.6 5.8 2-22
Cult 1 Comparison (t =6.09 p .001).
Cult 2 Comparison (t =5.14 p .001).
Using a cutoff score of 15 to discriminate between cult-involved and non-cult-involved
individuals (see Table 4), the ICE Index correctly identified 84% of the cult involved cases
as having been cult involved, and correctly identified 85% of the non-cult-involved cases as
not having been cult involved. For this sample, the ICE Index had an overall correct
classification rate of 84%.
Mean LASC scores for each group were X =40.0 for Cult 1 (SD =25.1) X =35.9 for Cult 2
(SD =25.3) and X =52.7 for the Comparison group (SD =21.1), with higher scores
indicating greater levels of distress. Despite differences in help-seeking behavior between
the three groups, an analysis of variance revealed no statistically significant difference
between distress scores.
For Cult 1 (support-seeking former cult members), the relationship between scores on the
ICE Index and the LASC were examined using Pearsons product-moment correlation. A
significant positive correlation was found (r =.45 p .001)