WPC&/ 2BVVZZ#|Rle)8wC;,Xw P7XP"m'^;C]ddCCCdCCCCddddddddddCCȲY~~wCN~sk~CCCddCYdYdYCdd88d8ddddJN8ddddYYdYCdddddCCCCddddddddd8YYYYYY~Y~Y~Y~YC8C8C8C8ddddddddddYdddddsdYYYYYYYd~Y~Y~Y~YddddddddC8C8C8C8oNd~8~8~8~8~8dvddddJJJkNkNkNkN~8~8~8dddddddYYYd~8dJkN~8dddddCddCCC/NdddCYQQddddddFddddFCChhd44ddzzdddwooChdF"Ȑdhd岲dCCȐzȲxCddodȐȅdCdYdsȐ]ȐȐȧzȐUwŐdȐYYCCCCѐz~ozoY~NYdYC8YooYdYzsdzdd~YYzozzzzNd88YYYzYzzzzCCdddddddzzzzzzzzzzzzzzzzzzzNNNNNNNdddddddddddddddddddd888888888888YYYYYYYYYYYYYYYYYYYzzzzzzzzzzzzzzzzzzzzCs~CzCddYCxHP LaserJet IIIHPLASIII.PRSXw P7,\,TXP2y'  #Xw P7XP##|R"m'^ $155_Y$$$5_$$$$5555555555$$k_k5_MGMMGASS)3SGeMSASM;GMMkMMG$$$55$5;/;/&5;$;Y;5;;/)$;5M55/555k$ 55555$$$$55555555;M5M5M5M5M5oMM/G/G/G/G/))))M;S5S5S5S5M;M;M;M;M5M5M;S5S5M5M5A;M5M5M5M/M/M/M/M;G/G/G/G/S5S5S5S5S5S5S;S;))))a=3S;GGGG)GM;MUM;M;S5S5xOM/M/M/;);););)G$G$G$M;M;M;M;M;M;kMM5G/G/G/M;GM;M/;)G$M5M5M;S5M;$55$$$/kkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkNMMM555$5--55h555__5&5555_&$$==5_kk55AAQk5k5M5CYY;;_$=5e|&"kkkkMMkkk5Mkk=MM__kkk5__kkk_5$$kkkkkMAk_MMMMMMMMMMx$k5M5;5_kMkG5$5M/5=MSIkkMMMMMMMMMMkkkkk3kMMMMkkkkkkkkkkkkkkkkkkkMMkkkkkk]GASkMkkk-?MMMMkkkkkkMMM5MMkMMkkkMMM5MMMM5MMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMM$MMM$MMM$MMM$MMMMMMMMMMMMMMAG;A;G/G)G/S/M/)S/G;e;M/G5S/MAA5A55G/M/GAM;SAAAA)////A/AAAA$$5555555AAAAAAAAAAAAAAAAAAA)))))))///////////////////////////////////////AAAAAAAAAAAAAAAAAAAA$AMGS)SMA)M55/$x"m'^ $155_S$$$5_$$$$5555555555$$k_k/_MCIOC?MO$)KCaOM=MG9COMeMME$$$55$/5/5/$555S5555()55M55//5/k$55555$$$$555555555M/M/M/M/M/_GI/C/C/C/C/$$$$O5M5M5M5M5O5O5O5O5M5M/O5M5M5M5O5=5M/M/M/I/I/I/I/O5C/C/C/C/M5M5M5M5M5M5O5O5$$$$N;)K5CCCCCO5O?O5O5M5M5_MG(G(G(9)9)9)9)CCCO5O5O5O5O5O5eMM5E/E/E/O5CO5G(9)CM5M5O5M5O5$55$$$/kkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkNMMM555$/++55^555__5&5555_&$$775_kk55AAQk5k5M5?YY;;_$75e&"kkkkMMkkk5Mkk7MM__kkk5__kkk_5$$kkkkkMAk_MMMMMMMMMMx$k5M5;5_kMkG5$5M/5=MSEkkMMMMMMMMMMkkkkk1kMMMMkkkkkkkkkkkkkkkkkkkMMkkkkkkYGASkMkkk-?MMMMkkkkkkMMM5MMkMMkkkMMM/MMMM/MMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMM$MMM$MMM$MMM$MMMMMMMMMMMMMMAC;A;G/C)E/O5M/$K/G;a;O/G5M/MA=5A55C/M/GAM;SAAAA)5///A/AAAA$$5555555AAAAAAAAAAAAAAAAAAA)))))))55555555555555555555///////////////////AAAAAAAAAAAAAAAAAAAA$=MCO$MMA$M55/$x2"' Z,H#Xw P7[hXP#HP LaserJet IIIHPLASIII.PRSi P7,\,TP"m'^4;RXX;;;X;;;;XXXXXXXXXX;;Ooyoi;E|oev_os;;;XX;OXOXO;XX11X1XXXXAE1XXXXOOXO;XXXXX;;;;XXXXXXXXX1OOOOOvyOoOoOoOoO;1;1;1;1XXXXXXXXXXOXXXXXeXOOOyOyOyOyOXoOoOoOoOXXXXXXXX;1;1;1;1bE|Xo1o1o1o1o1XiXXXXvAvAvA_E_E_E_Eo1o1o1XXXXXXXsOsOsOXo1XvA_Eo1XXXXX;XX;;;/NXXX;OHHXXXXXX>XXXX>;;\\X..XXllXXXibb;\X>"X\X坝X;;lx;XXbXvX;XOXesRvlKiŀXOO;;;;рloblbvOoEsOXO;1|OvbbOvXOleXlXXoOOvlbllllEX11OOOlOllll;;XXXXXXXlllllllllllllllllllEEEEEEEXXXXXXXXXXXXXXXXXXXX111111111111OOOOOOOOOOOOOOOOOOOllllllllllllllllllll;eo;l;XXO;xCG Times (Scalable)CG Times Italic (Scalable)CG Times Bold (Scalable)2,T: '"m'^;C]ddCCCdCCCCddddddddddCCȲY~~wCN~sk~CCCddCYdYdYCdd88d8ddddJN8ddddYYdYCdddddCCCCddddddddd8YYYYYY~Y~Y~Y~YC8C8C8C8ddddddddddYdddddsdYYYYYYYd~Y~Y~Y~YddddddddC8C8C8C8oNd~8~8~8~8~8dvddddJJJkNkNkNkN~8~8~8dddddddYYYd~8dJkN~8dddddCddCCC/NdddCYQQddddddFddddFCChhd44ddzzdddwooChdF"Ȑdhd岲dCCȐzȲxCddodȐȅdCdYdsȐ]ȐȐȧzȐUwŐdȐYYCCCCѐz~ozoY~NYdYC8YooYdYzsdzdd~YYzozzzzNd88YYYzYzzzzCCdddddddzzzzzzzzzzzzzzzzzzzNNNNNNNdddddddddddddddddddd888888888888YYYYYYYYYYYYYYYYYYYzzzzzzzzzzzzzzzzzzzzCs~CzCddYCx"m'^4;UXX;;;X;;;;XXXXXXXXXX;;Xllvle;OvbvllXbllbb;;;XX;XXOXO1XX11O1XXXXEE1XOvOOEOXO;XXXXX;;;;XXXXXXXXX1lXlXlXlXlXvvOlOlOlOlO;1;1;1;1vXXXXXXXXXbOlXXXXbOXlXlXlXlXvOvOvOvOXlOlOlOlOXXXXXXXX;1;1;1;1XOvOb1b1b1b1b1vXvlvXvXXXvlElElEXEXEXEXEb1b1b1XXXXXXvbObEbEbEXb1vXlEXEb1bObOXXX;XX;;;/NXXX;XRRXXXXXX>XXXX>;;XXX11XXllXXX_XX;XX>"XXX坝X;;lx;XlXbXvX;XOXesRvlKiŀXOO;;;;lllblbvOlEbOXO;1vOvbbvOvXOllXlXXbObOvllbllllEX11OOOlOllll;;XXXXXXXlllllllllllllllllllEEEEEEEXXXXXXXXXXXXXXXXXXXX111111111111OOOOOOOOOOOOOOOOOOOllllllllllllllllllll;ell;bl;bXXO;x"D,. 6 ^ENluuNNNuNNNNuuuuuuuuuuNNu[pNNNuuNuhhRuANAuh[NuuuhuuuNuuuuuNNNNuuuuuuuuAuuuuuhhhhh[A[A[A[Auuuuuuuuuuuuuhhhhhhhhuuuuuu[A[A[A[ApAAA[Auuhhh[[[[NNNuhhhAh[NuuuNuuNNN/NuuuNuccuuuuuuRuuuuRNNu<<uuuuuNuR"uuuNNxNuuuuNuhupcuuuNNNNh[hhh[Ahhuhuuuhh[hAAhhhhNNuuuuuuu[[[[[[[hhhhhhhhhhhhhhhhhhhhAAAAAAAAAAAAhhhhhhhhhhhhhhhhhhhN[[uuhNx"m'^4;RXX;;;X;;;;XXXXXXXXXX;;XvvlEUvlbvv;;;XX;XbObO>Xb1;b1bXbbOE;bXXXOXXX;XXXXX;;;;XXXXXXXXb1XXXXXOvOvOvOvOE1E1E1E1bXXXXbbbbXXbXXXXlbXXXOOOObvOvOvOvOXXXXXXbbE1E1E1E1eUbv1v1v1vEv1bbbXXȃOOObEbEbEbEv;v;v;bbbbbbXvOvOvObv1bObEv;XXbXb;XX;;;/NXXX;XKKXXXXXX>XXXX>;;eeX..XXllXXXobb;eX>"XeX坝X;;lx;XXbXvX;XOXeyUvlKiŀXXX;;;;рlvblbvOvEvOOOE1OvbbOvXOllXlXXvOOvlbllllEO11OOOlOllll;;XXXXXXXlllllllllllllllllllEEEEEEEOOOOOOOOOOOOOOOOOOOO111111111111OOOOOOOOOOOOOOOOOOOllllllllllllllllllll;lvElEXXO;x2Z,R$/2i;4,ti P7P8wC;,[hXw P7XP60g;4,)g&_ x$&7X@NE,/!3_ p^71l;4,/kl_ p^7M?$ ,;@? P7@PMA$ ,/Q@A_ p^7@ a      X` hp x (#%'0*,.8135@8:8" ԌIn practice, the above statistical optimization will be slightly compromised because the subscales are modestly correlated (see Table 3). This is the price of imperfect simple structure and of weighting each item in the scale by 1, instead of by the item's more refined factor loading, when summing scores of subscales. Nunnally (1978) suggests, however, that such refined weighting makes little difference in practice. This contention is supported by the fact that the correlations between the subscales are much smaller than those between the items within the subscales (summarized by the substantial Cronbach alphas). These contrasting patterns of correlation generally support the convergent and discriminant validities of the GPA scales, and suggest that efficient statistical discrimination is highly likely. Furthermore, the predominantly orthogonal nature of the scales allows us to conjugate the varieties of abuse, affording us powerful heuristic strategies for theory development. A subsequent paper will elaborate upon the heuristic value of conjugating the GPA scales.   yPdddy  R # Table 3. Correlations Between Scales ă  RJ ` `  $ Sc1+Sc2hh3Sc3:Sc4AGPA   S  Sc1: Compliance  $1.00 .16hh3.36 :.33A.70  S  Sc2: Exploitation  $ .16 1.00hh3.21:.19A.63  Sf  Sc3: Mind Control  $ .36 .21 1.00:.27A.59  S   Sc4: Anxious Dep.  $ .33 .19 .27 1.00A.70  S  Group Psych. Abuse  $ .70 .63 .59 .70 1.00   ydddy  R7  Psychological Rationale for Factor Names  S  Compliance. The concept of compliance is nearly self-evident in the factor loadings: sex lives dictated (#01), sacrificing own goals (#04), intimacy dictated (#14), serving leaders (#18), leaders making decisions (#21), and consulting leaders on decisions (#28). The group's livingR8"  together (#13) is not overtly reflective of compliance but becomes so when one considers that such communal living is typically demanded by leadership and makes noncompliance to group norms more difficult. Most of the excluded but relevant items in Table 3 also describe compliance situations, ranging from the minimum needed to meet the definition of compliance (#275, leaders are obeyed) to fanatical devotion (#235, will die for leader). Items #231, #249, #262, and #263 (which relate to feelings of guilt, inadequacy, failure) describe the reactions of someone who believes in compliance but, for whatever reason, does not measure up.  S  Exploitation. At first we wanted to label this factor power but decided that the word was too general in meaning. Many people can seek power ethically. For example, a benign political or religious movement may seek to raise money (#19), recruit members (#27), or gain political power (#06). These items reflect the power dimension of exploitation. A cult, however, will tend to use unethical means to gain power. They manipulate, abuse, and use people"that is, cults exploit people. This sense of unethical means comes across rather clearly in other factor items, such as approving of violence against outsiders (#12), threatening outside critics (#20), advocating lawbreaking (#03), or directing women to use their bodies for the group (#02). The powerseeking dimension of exploitation is also reflected in some of the excluded items in Table 3: assets given to group (#222), lives at lower standard than leader (#237), and secret agency supports (#305). Most of the other items reflect the manipulation/abuse/use component of exploitation.  S  Mind Control. The items in the Compliance factor refer to behaviors that leadership values: serve leader, follow group guidelines, live with other members, and so on. The items in Factor 3 (including most of the  S excluded items in Table 3) refer to a particular type of relationship between the leaders and followers. Item #15 (people stay because deceived and manipulated) captures much of the meaning of this factor. Other items of the subscale and those included in Table 3 reflect methods by which leadership sustains the deception: the leader criticizes18"  members (#26), critical thinking is impaired (#22), psychological pressure from leader (#25), members feel they are part of elite (#09)"(this latter item is one of the few carrots among the many sticks). Even most of the excluded items describe techniques for sustaining the deception and dominance of the leader: questions are discouraged (#225), dissent is not tolerated (#243, #254), information and time are controlled (#261, #290), a childlike dependency is enforced (#301). We had considered naming this factor manipulation but decided the term was too general. The items in this factor seem to be pointing to a special kind of manipulation in which the manipulator seeks not only compliance on some dimension (as would an ordinary con man) but personal dominance as well. Given that the second and thirdhighest loadings were mind control used (#24) and group used coercive persuasion (#11), we concluded that mind control was an appropriate label for this factor.  S  Anxious Dependency. At first, the items in this factor seemed rather disparate, and we wondered if the factor had a coherent meaning. Terms that came to mind included exclusivity, dependency, isolation, totalism, and fear. Eventually, anxious dependency seemed most effectively to summarize this factor. This conclusion was based partly on clinical observations, partly on empirical findings (Martin, Langone, Dole, & Wiltrout, 1992), and partly on theoretical reasoning. An article by Craig and Weathers (1990) on archetypal dependency in cult members was particularly helpful in illuminating the psychological dynamics reflected in the simple and complex items of the subscales. Craig and Weathers pointed out that the dependency of the member on the group extends beyond physical levels to deep psychological and spiritual needs. This dependency resembles the unconditional archetypal bond between an infant and a parent. Upon leaving the group, the member is thrown into severe separation anxiety. The highestloading item in the subscale (leaving group means eternal damnation, #07) reflects both the absolute dependency on the group and the anxiety associated with that dependency. Although dependency is often associated with anxiety (because one tends to fear losing that on which-8"  one is dependent), this connection is not always necessary (e.g., one may realistically depend on a person for certain things and feel reasonably secure in that relationship). In a cult situation, however,  Sv dependency can be absolute and fear tends to color all experiences (except perhaps in the early honeymoon phase of recruitment/ seduction). As a way of consolidating absolute dependency, negative emotions toward the group (#08) are rejected, thus contributing to the numinosity of the archetypal possession (Jung, 1972, 1973; Stein, 1984). The archetypal inflation of the leader reaches its ultimate rarefied expression in the leader's supposed divinity (#23). Through apotheosis, the leader becomes like an allpowerful parent. The member, although mortal, is among the elite children of god. Between the divine and the elite mortal, however, stands an infinite competency gulf, untraversable by the evergroping mere mortal. Smaller distances are perhaps found when an abusing husband indoctrinates his wife to believe that she could never make it without him (Tobias & Lalich, 1994). The archetypal base of cultic dependency is further consolidated by the assumption that critics of the group are under an evil power (#10) and, therefore, are to be feared and avoided by those remaining dependent on the group. In extreme cases, not even the medical establishment is to be trusted (#05 and #15). Most of the excluded items also reflect an explicit or implicit anxious dependency"for example, items #257 (punishment for dissent), #315 (only leaders true), and #317 (avoid family and friends), which among the complex items have the three highest loadings on this factor.  S  An empirical definition of cult. Emerging from this analysis is an empirically based definition of cultism that seems to be concise and consistent with the views of most clinicians who have worked with cult victims:  S eCults are groups that often exploit members psychologically  S and/or financially, typically by making members comply with leadership's demands through certain types of psychological  S3 manipulation, popularly called mind control, and through the38"   S inculcation of deepseated anxious dependency on the group and its leaders.g It is important to remember that the conjugation of subscales may produce GPA profiles that vary from the set of related profiles that would fall within this definition of cult. Rational analysis and empirical investigations (e.g., discriminant analysis) of the spectrum of profiles can result in a classification system for the varieties of psychological abuse, with cult being one of many possible categories within this system.  R  Discrimination and Variations on the Theme of Abuse Although noncultic groups were not included in the study, a sense of the GPA's discriminatory power may be gleaned by comparing the means of the scales with the ranges of ratings. The means and standard deviations of the items and subscales (see Table 4) suggest that, on average, most of the cults score positively on the four types of abuse. We define positivity by reference to the midpoints of the possibilities. The midpoint of the items' rating range is 3. The subscales' midpoint is 3x7=21. The GPA summary scale midpoint is 21x4=84. A positive score implies that the means fall above their respective midpoints. That is, they indicate a rating of characteristic to very characteristic.  S The mean for the GPA scale was 110.70. A t-test comparing this mean  S with the midpoint of 84 was significant: t=31.98, df=264 (due to  SZ missing values), p<.0001. Thus, the summary scale falls well within the positive range of the GPA scores. The subscale means also fell above their midpoints, but they differed substantially from one another. In order to compare the subscale means with one another and their midpoints, the subscales were treated as hypothetical levels in an analysis of variance. The Duncan multiple-range test showed that the  SH scales differed significantly from one another: F=150.25, df=3, 1129,  S p<.0001. The Exploitation scale mean of 22.88 was the lowest. Its 95% confidence range was 22.18 to 23.59, placing it and the other scales above their midpoint of 21. The mean for Anxious Dependency was the next highest: m=26.48; its 95% confidence range was 25.83 to 27.13. The 88" Ԍ  ԙydddy  R  Table 4. Means, Standard Deviations, Standard Error of Means,  R | and Ranges of the GPA and Scales ă  RX  # Mean S.D. S.E.M. Range Description  GPA 110.70 13.42 .835 70-139 Group Psychological Abuse  S  (3.95 .48 .030 2.5!4.96)X01Í ÍX1*Í Í *ל S^ #i P7tP#э Figures in parentheses show point on characteristic scale adjusted for number of items.  R  Sc1: Compliance Sc1 29.27 5.30 .309 8-35 Compliance (4.18 0.76 .044 1.1!5)  S R01I* 4ל S ԍ R signifies a reversed item.I 4.22 1.01 .063 1-5 sex lives not dictated 04 4.19 1.00 .062 1-5 sacrifice own goals 13 3.82 1.34 .084 1-5 group lives together 14 4.24 1.05 .066 1-5 intimacy dictated 18 4.34 1.02 .064 1-5 serve leaders 21 4.35 1.04 .065 1-5 members make decisions 28 4.24 1.11 .070 1-5 members must consult  Rx  Sc2: Exploitation Sc2 22.96 5.91 .352 7-35 Exploitation (3.28 0.84 .050 1!5) 02 2.00 1.21 .076 1-5 women seduce for group 03 3.48 1.37 .085 1-5 breaking law okay 06 3.10 1.53 .095 1-5 politics major goal 12 2.69 1.40 .087 1-5 violence to outsiders 19 3.81 1.41 .088 1-5 money major goal 20 3.62 1.22 .076 1-5 critics threatened 27 4.38 1.06 .066 1-5 recruiting major goal P8" ԌTable 4. (continued)  R Sc3: Mind Control Sc3 31.64 3.43 .202 21-35 Mind Control (4.52 0.49 .029 3!5) 09 4.86 0.36 .024 2-5 members part of elite 11 4.52 0.81 .051 1-5 coercive persuasion 15 4.56 0.67 .042 2-5 stay because deceived R22 4.55 0.78 .049 1-5 can think critically 24 4.61 0.67 .042 1-5 mind control used R25 4.32 1.02 .064 1-5 little psychological pressure R26 4.19 1.14 .071 1-5 criticism rare  R  Sc4: Anxious Dependency Sc4 26.52 5.55 .326 8-35 Anxious Dependency (3.79 0.79 .046 1.1!5) 05 3.39 1.21 .075 1-5 medical help okay 07 4.03 1.26 .079 1-5 leaving is damnation 08 4.24 1.08 .067 1-5 no negative emotions 10 4.06 1.33 .083 1-5 criticism is evil 16 3.71 1.52 .095 1-5 exercises remove doubt 17 2.99 1.34 .084 1-5 no medical help 23 4.07 1.28 .080 1-5 leader is divine   ydddy Compliance scale mean was next highest: m=29.27; its 95% range was 28.66 to 29.87. The Mind Control scale produced the highest mean: m=31.66; its 95% range was 31.26 to 32.07. Consideration of the means and ranges for the specific items listed in Table 4 suggests that most but not all of the item means were above the midpoint of 3. As a group, the cults did not have women seducing for them (m=2.00) and did not advocate violence against outsiders (m=2.69). The factor analysis loadings, however, indicate that groups that advocate these items did tend to score higher on the general factor of exploitation. <8" ԌThe loadings reflect the meaning of the factor at all levels. Seduction may be rare in general, but it nonetheless falls into the pattern of exploitation. Thus, the fact that 2 of the 28 items were not generally characteristic of the cults does not invalidate the factorial meaningfulness of the two items. We would expect the more extremely exploitative groups to encourage seduction and violence. The notion of a continuum of abuse, as discussed for the items, may also be of relevance to the subscale means but with a very important difference. The ANOVA showed that the subscale means differed significantly from one another. The subscales, however, are different from the items in that they are generally uncorrelated. They are not points along an essential continuum. Nevertheless, factors do lend themselves to various nonessential but characteristic continuums. This point is potentially confused by the differences and correlations between the subscales, which may suggest a slightly prevalent continuum of factors. The hypothetical continuum ranges from Mind Control (m=31.66) to Compliance (m=29.27) to Anxious Dependency (26.48) to Exploitation (m=22.88). Similarly, Mind Control is most highly correlated with  S: Compliance (r=.36); while Compliance's next highest correlation is with Anxious Dependency (.33), which is followed by the remaining correlation linking Anxious Dependency with Exploitation (.19). Although each correlation is statistically significant, they do not collapse the general simple structure. At worst, the factor structure is slightly oblique. Thus, the hypothetical sequence of factors is certainly not the only sequence suggested by the data. Because the factors are orthogonal, the subscales and their combinations  S are best viewed as varieties of the species psychological abuse. Each item and subscale contributes to the overall level of abuse; but there are many potential paths that could be constructed through the matrix of varieties, each characterizing a different behavioral sequence or syndrome of abuse. Thus, in addition to the static (profile) perspective on the GPA discussed earlier, there is a dynamic (sequence) perspective that one can take. The delineation of sequences through a matrix of varieties can be useful in08"  explicating the dynamics of cultic abuse. For example, the sequence of Mind Control to Compliance to Anxious Dependency to Exploitation is consistent with Dole and DubrowEichel's (1985) findings. It is vital, however, that we recognize that cult leaders may freely conjugate the varieties of abuse and that none of these permutations provides an  S essential definition of cultic abuse. These permutations may, nonethe S less, prove very useful in elaborating the existential nature of cultic abuse. Thus, an unscrupulous, charismatic leader may follow the sequence above or may stumble upon a small group of already anxiously dependent individuals and take advantage of their psychological weakness to persuade them to comply to his exploitative demands without resorting to much mind control. This is a different permutation, or sequence, from that listed above. But both participate in the common existential nature of psychological abuse. Relating these ideas to the classification of profiles discussed earlier suggests that the varieties of psychological abuse implied by the GPA subscales may be viewed from both dynamic (sequence) and static (profile) perspectives.  Sl  Conclusions and Recommendations for Further Research ă Content validity for the GPA has been built from the Dole and Dubrow-Eichel (1985) Delphi study, the Langone and Chambers (1991) factor analysis of terms, and Langone's (1992) philosophical analysis of the respect/abuse continuum. Factorial validation of the GPA employed principal components analysis with varimax rotation. Items were selected in order to promote the simple structure of the GPA. Reliability coefficients for the GPA and its subscales support the use of the GPA in research contexts. Much additional research is needed, however, before we can claim definitive construct validity for the GPA scales. As alluded to earlier, discrimination is a crucial element in construct validity. Indeed, discrimination of cultic and noncultic groups would probably be the most common application of the GPA or of any other cult abuse scale. Steps toward establishing discriminant validity have already been taken. _8" ԌIn her master's thesis, Adams (1993) used the GPA subscales to contrast former members of the Cincinnati Church of Christ (CCC) with former members of InterVarsity Christian Fellowship (IVCF). The CCC is described by many as a cult, while the IVCF, a mainstream campus group, is not considered cultic. Adams found the CCC scored much higher on all four GPA subscales than did the IVCF. These results are encouraging and represent important steps in the establishment of discriminant validity. Additional studies of this nature will be conducted. Many applications of the GPA scales may be envisioned. As a standardized objective instrument, the GPA would lend considerable clarity to legal, philosophical, and psychological debates concerning the extent and manner of the abuse inflicted by cultic groups. Do therapy cults differ from religious cults in their emphasis on mind control? Do women report abuse in cults differently than men do? Do former members alter their ratings if they have participated in an exit counseling session? How do the ratings by former members compare with those of cult leaders, apologists, exit counselors, parents, and others? Does the convergence of former members' views with their parents' views of the group herald the reintegration of the family? Are there groups (academic, corporate, religious, professional) that would not ordinarily be considered cultic but that score high on one or more of the subscales? Could the GPA be used to monitor the transformation of an innocuous group into a cult or of a cult into an innocuous group? Literally hundreds of questions could be answered using the GPA or a similar scale. For now it is important that validation research with the GPA continue before the scale is used to make any conclusive judgments. More construct validity studies should be performed. The GPA factors should be cross-validated with different samples of subjects, including panels of experts. The sample used in this study must be broadened to include more exmembers outside the countercult network, current cultists, and members and former members of noncultic groups. It may be necessary to improve some of the items in order to simplify the factor structure. Test-retest reliability should be examined. Norms should eventually be gathered. The good news is that all of these steps toward validation are wellestablished procedures. It will take some work, but the fruits will be,8"  worth the labor. We will come to a more systematic understanding of the psychological abuse in cults. These possibilities more than warrant the investment of additional research with the GPA scales.  SH  ) References ă  S eAdams, D. (1993). The Cincinnati Church of Christ: How former  S members rate the group on the cultism scale. Unpublished master's thesis, Xavier University, Cincinnati, OH. eCraig, N. W., & Weathers, R. (1990). The false transformational  S9 promise of Bible-based cults: Archetypal dynamics. Cultic Studies  S Journal, 7(2), 160!173. eDole, A., & Dubrow-Eichel, S. (1985). Some new religions are  S dangerous. Cultic Studies Journal, 2, 17!30.  S eGhiselli, E. E., Campbell, J. P., & Zedeck, S. (1981). Measurement  S[ theory for the behavioral sciences. San Francisco: W. H. Freeman.  S/ eGorsuch, R. L. (1983). Factor analysis. Hillsdale, NJ: Lawrence Erlbaum.  S eJung, C. G. (1972). Two essays on analytical psychology. Princeton, NJ: Princeton University Press.  S{ eJung, C. G. (1973). Mandala symbolism. Princeton, NJ: Princeton University Press.  S! eKerlinger, F. N. (1973). Foundations of behavioral research. New York: Holt, Rinehart, & Winston. eLangone, M. D., & Chambers, W. V. (1991). Outreach to ex-cult  S members: The question of terminology. Cultic Studies Journal,  Sm 8(2), 134!150.  SA eLangone, M. D. (1992). Psychological abuse. Cultic Studies Journal,  S 9(2), 206!218. eLangone, M. D. (1993). Helping cult victims: Historical background.  S In M. D. Langone (Ed.), Recovery from cults: Help for victims of  S psychological and spiritual abuse. New York: W. W. Norton.  Sc eLifton, R. J. (1961). Thought reform and the psychology of totalism. New York: W. W. Norton. eMartin, P. R., Langone, M. D., Dole, A. A., & Wiltrout, J. (1992). Postcult symptoms as measured by the MCMI before and after  S residential treatment. Cultic Studies Journal, 9(2), 219!250.  S eNunnally, J. C. (1978). Psychometric theory. New York: McGraw-Hill.8" ԌeOfshe, R., & Singer, M. T. (1986). Attacks on peripheral versus central elements of self and the impact of thoughtreforming  S techniques. Cultic Studies Journal, 3(1), 3-24.  Sx eSchein, E., Schneier, I., & Baker, C. H. (1961). Coercive persuasion. New York: W. W. Norton. eSinger, M. T., Temerlin, M., & Langone, M. D. (1990). Psychothera S py cults. Cultic Studies Journal, 7(2), 101!125.  S eStein, M. (1984). Jungian analysis. Shambhala: London.  S eThurstone, L. L. (1947). Multiple-factor analysis. Chicago: University of Chicago Press.  S> eTobias, M. L., & Lalich, J. (1994) Captive hearts, captive minds;  Q Freedom and recovery from cults and abusive relationships. Alameda, CA: Hunter House.  S eYeakley, F. (Ed.). (1988). The discipling dilemma. Nashville, TN: Gospel Advocate.  S.   ****************************** ă  S  $ Appendix A: GPA Scale ă This inventory is designed to evaluate certain aspects of religious, psychotherapeutic, political, commercial, and other groups. Please rate, as best you can, the degree to which the following statements character S ize the group under consideration. Rate each item according to your  S experience and observations (in retrospect) of how the group actually functioned. If your group had different levels of membership (within which the group's dominant features differed), please apply the ratings to the level with which you have greatest familiarity. Circle the best answer, using the following ratings: 1 = not at all characteristic 2 = not characteristic 3 = can't say/not sure 4 = characteristic 5 = very characteristic 1.[R] The group does not tell members how to conduct their sex lives. 1` ` 2 $3+4hh35@8" Ԍe 2. Women are directed to use their bodies for the purpose of recruiting or of manipulation. 1` ` 2 $3+4hh35 e 3. The group advocates or implies that breaking the law is okay if it serves the interests of the group. 1` ` 2 $3+4hh35 e 4. Members are expected to postpone or give up their personal, vocational, and educational goals in order to work for the group. 1` ` 2 $3+4hh35 5.[R] The group encourages ill members to get medical assistance. 1` ` 2 $3+4hh35 6. Gaining political power is a major goal of the group. 1` ` 2 $3+4hh35 e 7. Members believe that to leave the group would be death or eternal damnation for themselves or their families. 1` ` 2 $3+4hh35 8. The group discourages members from displaying negative emotions. 1` ` 2 $3+4hh35 9. Members feel they are part of a special elite. 1` ` 2 $3+4hh35 Z8" Ԍe10. The group teaches that persons who are critical of the group are in the power of evil, satanic forces. 1` ` 2 $3+4hh35 11. The group uses coercive persuasion and mind control. 1` ` 2 $3+4hh35 e12. The group approves of violence against outsiders (e.g., satanic communists, etc.). 1` ` 2 $3+4hh35 13. Members are expected to live with other members. 1` ` 2 $3+4hh35 e14. Members must abide by the group's guidelines regarding dating and intimate relationships. 1` ` 2 $3+4hh35 e15. People who stay in the group do so because they are deceived and manipulated. 1` ` 2 $3+4hh35 e16. The group teaches special exercises (e.g., meditation, chanting, speaking in tongues) to push doubts or negative thoughts out of consciousness. 1` ` 2 $3+4hh35 e17. Medical attention is discouraged, even though there may be a medical problem. 1` ` 2 $3+4hh35 ,8" Ԍe18. Members are expected to serve the group's leaders. 1` ` 2 $3+4hh35 e19. Raising money is a major goal of the group. 1` ` 2 $3+4hh35 e20. The group does not hesitate to threaten outside critics. 1` ` 2 $3+4hh35 e21.[R] Members are expected to make decisions without consulting the group's leader(s). 1` ` 2 $3+4hh35 e22.[R] Members are just as capable of independent critical thinking as they were before they joined the group. 1` ` 2 $3+4hh35 23. The group believes or implies its leader is divine. 1` ` 2 $3+4hh35 e24. Mind control is used without conscious consent of members. 1` ` 2 $3+4hh35 e25.[R] Members feel little psychological pressure from leaders. 1` ` 2 $3+4hh35 e26.[R] The group's leader(s) rarely criticize members. 1` ` 2 $3+4hh35 ,8" Ԍe27. Recruiting members is a major goal of the group. 1` ` 2 $3+4hh35 e28. Members are expected to consult with leaders about most decisions, including those concerning work, child rearing, whether or not to visit relatives, etc. 1` ` 2 $3+4hh35 Note: [R] items are reversed in scoring by finding the absolute difference between the rating and the number 6. Do not include the [R] designations when administering the test. At this time the GPA Scale should be used only as a research instrument. We request that researchers wishing to use the GPA Scale contact Dr. Langone (AFF, P.O. Box 2265, Bonita Springs, FL 33959).  S   ****************************** ă  Si  William V. Chambers, Ph.D., is a statistical consultant and assistant professor at University College, Mercer University. He is coauthor with James Grice of CASPER, a software package for psychometric analysis.  S  Michael D. Langone, Ph.D., is editor of the Cultic Studies Journal and  S executive director of AFF. He is editor of Recovery from Cults: Help  SZ for Victims of Psychological and Spiritual Abuse (W. W. Norton, 1993).  S  ÞArthur A. Dole, Ph.D., is Professor Emeritus, Division of Psychology in Education, Graduate School of Education, University of Pennsylvania.  SI  James W. Grice, M.A., is a doctoral candidate in the Department of Psychology at the University of New Mexico. He was awarded the 1994 Mariani Award as the outstanding student by the New Mexico Psychological Association.