GPA Modifications and Development

1) The Group Psychological Abuse Scale (commonly referred as GPAS or simply GPA) was originally developed by Chambers, Langone, Dole & Grice (1994), owing to the work of Langone & Chambers (1991) and Langone (1992).

Thus, the primary reference is:

Chambers, W. V., Langone, M. D., Dole, A. A., & Grice, J. W. (1994). The Group Psychological Abuse scale: a measure of the varieties of cultic abuse. Cultic Studies Journal, 11, 88-117.

Secondary references are:

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

Langone, M.D. (2005). Psychological Abuse: Theoretical and Measurement Issues. ICSA E-Newsletter, 4 (3).

The GPA includes four subscales composed of seven items each: Compliance, about the degree to which an individual must agree to the group rules; Exploitation, about the degree to which the group uses the person in detriment of his/her own wellbeing; Anxious dependency, referring to the degree to which the person depends on the group; and Mind control, referring to the degree to which the group and its leaders use deceitful, manipulative strategies to keep their members.

2) People interested in the English version of GPAS, might want to read:

Almendros, C., Gámez-Guadix, M., Rodríguez-Carballeira, A., & Carrobles, J.A. (2011) Assessment of Psychological Abuse in Manipulative Groups. International Journal of Cultic Studies, 2, 61-76.

In this work, it is stated: “Almendros (2006) recently proposed a modified version of the GPA, which removes the negative wording of item 1 of the instrument: “The group does not tell members how to conduct their sex lives,” and modifies the response options, omitting the label of the central category. This is the version currently being used with US, Japanese, and Spanish participants (Group Psychological Abuse Scale-Modified—GPA-M) (Almendros, Gámez-Guadix, Carrobles, & Rodríguez-Carballeira, 2009). Using the responses of 138 English-speaking participants, self-identified as former members of abusive groups, Almendros (2011) reported support for the original four-factor solution of the GPA (Chambers et al., 1994) for the English version of the GPA-M, which yielded better-fitting statistical values through Confirmatory Factor Analyses when compared to a one-factor solution.”

3) We suggest that further applications of the instrument use this slightly modified version: Group Psychological Abuse Scale-Modified (GPA-M; Chambers el at., 1994; Almendros et al., 2009). Slight changes are highlighted in blue below at the end of this file.


An important reference for the English and Japanese version of this modified version is:

Almendros, C., Carrobles, J. A., Rodriguez-Carballeira, A., Gámez-Guadix, M., Martín-Peña, J., Porrúa-García, C. & Escartín, J. (2009, July). Psychological abuse reported by former members of manipulative groups across different cultural groups. Poster presented at the International Conference of The Norwegian Psychological Association. European Federation of Psychologists’ Associations (EFPA): XI European Congress of Psychology, Oslo, Norway.

This work provides an empirical cut-off value of >74 for the overall score of the GPA-M as an appropriate criterion value to discriminate abusive experiences among former group members.

4) As a result of e-mail communications with Jill Mytton, Almendros (January 16th, 2012) further suggested a modification in item 22 to read: “22 - Members are encouraged and/or allowed to use independent critical thinking” (highlighted in yellow below) instead of the original: “22 - Members are just as capable of independent critical thinking as they were before they joined the group.”

This, at least, when intending to use the GPA-M to be responded by former members who were born and/or raised in the groups.

Note.: No work has yet reported psychometrical characteristics of the instrument including this newer modification.

5) Scoring of the GPA-M.

Once removed the negative from the original item 1, for the GPA-M, this items doesn’t need to be recoded.

These are the steps:

Recode items 5, 21, 22, 25 and 26 so that (1=5) (2=4) (3=3) (4=2) (5=1).

Compliance: Sum items 1, 4, 13, 14, 18, 21reversed, 28

Exploitation: Sum items 2, 3, 6, 12, 19, 20, 27

Anxious Dependency: Sum items 5reversed, 7, 8, 10, 16, 17, 23

Mind Control: Sum items 9, 11, 15, 22reversed, 24, 25reversed, 26reversed

6) Possible spss syntax:

********Recode inverse items: 5, 21, 22, 25 and 26.



(1=5) (2=4) (3=3) (4=2) (5=1) INTO gpaa5r .

VARIABLE LABELS gpaa5r 'gpaa5rec inversa'.




(1=5) (2=4) (3=3) (4=2) (5=1) INTO gpac21r .

VARIABLE LABELS gpac21r 'gpac21rec inversa'.




(1=5) (2=4) (3=3) (4=2) (5=1) INTO gpam22r .

VARIABLE LABELS gpam22r 'gpam22rec inversa'.




(1=5) (2=4) (3=3) (4=2) (5=1) INTO gpam25r .

VARIABLE LABELS gpam25r 'gpam25rec inversa'.




(1=5) (2=4) (3=3) (4=2) (5=1) INTO gpam26r .

VARIABLE LABELS gpam26r 'gpam26rec inversa'.


********Compute totals for global score GPA and the four subscales.

COMPUTE sumagpa = gpac.1 + gpae.2 + gpae.3 + gpac.4 + gpaa5r + gpae.6 +

gpaa.7 + gpaa.8 + gpam.9 + gpaa.10 + gpam.11 + gpae.12 + gpac.13 + gpac.14 +

gpam.15 + gpaa.16 + gpaa.17 + gpac.18 + gpae.19 + gpae.20 + gpac21r +

gpam22r + gpaa.23 + gpam.24 + gpam25r + gpam26r + gpae.27 + gpac.28 .


COMPUTE sumagpac = gpac.1 + gpac.4 + gpac.13 + gpac.14 + gpac.18 + gpac21r + gpac.28 .


COMPUTE sumagpae = gpae.2 + gpae.3 + gpae.6 + gpae.12 + gpae.19 + gpae.20 + gpae.27 .


COMPUTE sumagpaa = gpaa5r + gpaa.7 + gpaa.8 + gpaa.10 + gpaa.16 + gpaa.17 + gpaa.23 .


COMPUTE sumagpam = gpam.9 + gpam.11 + gpam.15 + gpam22r + gpam.24 + gpam25r + gpam26r .


7) A caution ought to be made regarding the use of the GPA or the GPA-M.

The purpose of the works related to this instrument is to help researchers on group violence to increase our knowledge. It also intends to be useful for victims of group abuse, their relatives and helping professionals in gaining a better understanding of the abusive experiences they may have endured.

The GPA, and the further suggested GPA-M, are self-reports that share common flaws of these kind of instruments. It ought to be used with competence and responsibility at interpreting and deriving conclusions.

In no way GPA or GPA-M scores can be used to classify groups.

8) So far, either the GPA or the GPA-M have been used in different countries and/or translated to different languages (Spanish, Japanese, Swedish, German, Italian, and Icelandic). These works are at different stages of reporting psychometrical properties of the respective versions.

Psychological Abuse Experienced in Groups Scale (PAEGS)

9) You might be interested in a more recent instrument developed by Rodriguez-Carballeira’s research team: Psychological Abuse Experienced in Groups Scale (PAEGS).

Primary references for the English version are:

Rodríguez-Carballeira, A., Saldaña, O., Almendros, C., Martín-Peña, J., Escartín, J., & Porrúa-García, C. (2015). Group psychological abuse: Taxonomy and severity of its components. The European Journal of Psychology Applied to Legal Context, 7, 31-39.

Saldaña, O., Rodríguez-Carballeira, A., Almendros, C., & Escartín, J. (2017). Development and validation of the Psychological Abuse Experienced in Groups Scale. The European Journal of Psychology Applied to Legal Context, 9, 57-64. PAEGS Scale.