Measuring Social Climate in German Prisons Using the Essen Climate Evaluation Schema

Abstract

Sociotherapeutic prison units have been criticized for their close operational connection with traditional prison systems. This research compared the social climate of German therapeutic prison units with that of traditional prison units, using the Essen Climate Evaluation Schema (EssenCES). The EssenCES was administered in conjunction with the shortened Correctional Institutions Environment Scale. Results indicate significant differences between therapeutic and traditional units’ social climates, with the former demonstrating more supportive, safe, and therapeutic climates. Findings further confirm the psychometric properties of the EssenCES and support its use as a social climate measurement instrument.

Keywords

EssenCES social climate prison climate therapeutic prison climate

Up until the Middle Ages and later, the punishment of criminals in Europe was predominantly effected through physical means (e.g., whipping, mutilation, or capital punishment) or different forms of forced servitude. During the colonial era, convicts were often exiled to remote, isolated penal colonies. In the early stages of the development of the modern European/North American prison system, punishment, control, and incapacitation stood out as the primary goals of incarceration (Foucault, 1977). With these primary objectives, the quality of the institutional atmosphere was of a little concern to the architects of the penal system. The origins of the German prison system could be “ . . . characterized by hunger, overcrowding, and conflict” (Wachsmann, 2002, p. 411). In the 18th century, however, prison reform became a “hot button topic” for debate, and an emerging perspective argued that humane “correctional” institutions could provide moral instruction and foster proper behavior in their inmates (Lewis, 2009).

While the current operation of many high security prisons may suggest differently, the rehabilitation of prisoners remains a primary goal of correctional agencies in most Western countries (Day, Casey, Vess, & Huisy, 2011). In addition to humanitarian reasons, evidence supportive of the positive effects of rehabilitative correctional programming underpins this perspective (Andrews & Bonta, 2010). Nevertheless, as in forensic psychiatric institutions (Davies, 2004), meeting the support and rehabilitation needs of inmates while maintaining a secure and functioning corrections system can be a challenge. Day et al. (2011) refer to this challenge of addressing conflicting interests as the “implementation gap.” Referencing Shefer’s (2010) work on rehabilitation in the modern correctional institutions, Day et al. (2011) point to the environmental and interpersonal contexts of prison settings as the source of constraint on the effectiveness of program delivery.

The nature of the interaction of individuals and their environments is a focus of social climate research. In characterizing a correctional or treatment setting, it is not sufficient to describe only its objective features. The physical, social, and emotional conditions of an institutional setting interact in a specific way and constitute a condition that can be called “social climate or atmosphere,” which may over time influence the mood, behavior, and self-concept of the people involved (Schalast & Groenewald, 2009).

Assessing Social Climates

Beginning in the 1960s, there is extensive research examining the effects of institutional climate in clinical and correctional settings. Many of the earliest significant contributions to this line of research were made by American psychologist Rudolf H. Moos. Moos (1997) discussed the social climate of an environment in terms of its “personality.” The social climate of an institution could be characterized in a number of ways including, but not limited to, friendly, supportive, restrictive, controlling, involved, encouraging, clear, or well organized (Timko & Moos, 2004). As with any discussion of human “personality,” many variations on different climate traits might be enumerated and investigated.

Moos developed a number of social climate measurement scales designed to produce a more objective measure of the climate of an institution. Differing slightly across variations, these instruments were of comparable structure and consisted of similar content, among them the more widely known Ward Atmosphere Scale (WAS; Moos, 1996; Moos & Houts, 1968) and the Correctional Institutions Environment Scale (CIES; Moos, 1975). Moos’s research and the measurement instruments he developed have served as a foundation for social climate research; however, the lengthy questionnaires (90-100 items) have also been criticized for methodological and pragmatic reasons. The complex dimensionality of the Moos’s scales, consisting of three primary scales and nine to 10 subscales, has not been confirmed through adequate statistical analysis (Denny, Costello, & Cochran, 1984; Pedersen & Karterud, 2007; Røssberg & Friis, 2003; Wright & Boudouris, 1982). A second source of criticism derives from the scales’ length and impractical resource requirements for administration, which may cause participant dropout and missing data, especially among less motivated individuals or populations for whom such a lengthy endeavor may prove challenging (Middelboe, Schjodt, Byrsting, & Gjerris, 2001). This may be particularly the case when dealing with correctional populations, with low incentive to complete these types of surveys and often consisting of individuals with mental or learning disabilities. Shortened versions of these social climate measurement instruments have been suggested by many authors over time (e.g., Røssberg & Friis, 2003), but have yet to achieve standard implementation.

Moos actually developed a shortened 36-item version of the CIES by selecting the four items from each subscale demonstrating the greatest item-to-subscale correlation (Moos, 1974, 1975). Dahle and Steller (1990) administered a German translation of the shortened CIES in their research examining the effects of incarceration on adolescents. Based on factor and item analyses, the authors removed seven of the 36 items. The remaining 29 items were assigned to five climate dimensions with differing weights across items and dimensions: “rehabilitative orientation” (nine items), “order and cleanliness” (three items), “openness of communication between inmates and staff” (seven items), “discipline and control” (five items), and “openness of communication between inmates” (five items). This abbreviated and modified version of CIES demonstrated high reliability and validity, but its structure was far from a replication of the trait structure suggested by Moos (1975).

The Essen Climate Evaluation Schema (EssenCES)

In the 1990s, Schalast (1995) began developing a brief measurement instrument to assess the basic aspects of social climate in forensic psychiatric institutions. In a series of studies, different sets of questionnaire items were evaluated, intending to identify a small number of relevant and coherent climate dimensions that would remain stable and facilitate replication across different settings and samples. From this process, the 15-item (17 including two nonscored items) EssenCES was developed. The EssenCES measured social climate across three dimensions: Therapeutic Hold, Patients’ Cohesion and Mutual Support, and Experienced Safety (against threat of aggression and/or violence), each assessed by a distinct scale consisting of five items (Schalast, Redies, Collins, Stacey, & Howells, 2008). Scales scores relay

. . . the extent to which the climate is perceived as supportive of therapy and therapeutic change, whether mutual support of a kind typically seen as characteristic of therapeutic communities is present, and the level of tension and perceived threat of aggression and violence. (Day, Casey, Vess, & Huisy 2012, p. 158)

The EssenCES has grown in popularity and utilization internationally. Recently, the U.K. Department of Health included climate evaluation using the EssenCES in their Commissioning for Quality and Innovation (CQUIN) framework (Milsom, Freestone, Duller, Bouman, & Taylor, 2014).

The instrument’s statistical veracity has been examined in a number of studies, which have repeatedly confirmed the factor structure and reported adequate internal consistency and subscale validity (Howells et al., 2009; Milsom et al., 2014; Schalast et al., 2008; Tonkin et al., 2012). A corrections-modified version of the EssenCES was developed for application in prisons, with slight variations on the wording used in the original instrument designed for use in forensic psychiatric wards. The subscales’ names “Therapeutic Hold” and “Patients’ Cohesion” were changed to “Hold and Support” and “Inmate Cohesion,” and items were modified to reflect inmate/correctional staff respondents. The corrections-modified version was tested and validated in numerous studies (Day et al., 2011, 2012; Tonkin et al., 2012).

Specific Aspects of the Social Climate in German Prisons

As mentioned above, a supportive environment that is conducive to the social rehabilitation of inmates is a lauded goal of incarceration in most Western countries (Day et al., 2012). This is characteristic of the correction’s system that is common to Germany, as well. Section 3 of the German Execution of Sentences Act requires prisons to mitigate to the extent possible the negative impact of incarceration on prisoners and to facilitate social readjustment of inmates postrelease. In cases where an individual is sentenced to 2 or more years for a sex-related offense, priority consideration for transfer to a sociotherapeutic prison is given, and said transfer is mandated where rehabilitation is recommended by the Court (German Execution of Sentences Act, § 9). This requirement also holds for violent offenders.

Besides these social therapeutic prison units, there are units specialized to foster young prisoners. These units were designed to provide vocational training, educational opportunities, and specially trained staff. Originally envisioned as independently functioning facilities, specialized therapeutic units operating in the wings of traditional prisons have developed in the face of economic and practical constraints. Criminal justice practitioners and academics have criticized this compromise, some pointing to it as detrimental to the rehabilitative goals of the corrections system. For instance, Rehn (2008) argued that many of these units have lost focus on their initial goals and become difficult to differentiate from the prisons to which they are attached.

This study adds to the existing literature further evidence of the reliability and validity of the EssenCES as a measurement instrument for social climate in the prison setting. The following research questions were relied on throughout the research process:

Research Question 1: Is the EssenCES a reliable and valid instrument to measure the social climate of prisons?

Research Question 2: What is the relationship between staff and inmate perceptions of social climate as measured by the EssenCES?

Research Question 3: What is the relationship between German “treatment units” and traditional units in terms of social climate?

Method

The current research measured the social climate of several German prisons characterized by different inmate populations, treatment levels, and rehabilitative goals. The comparison of measures between different types of units serves to further validate the ability of the EssenCES to provide a consistent and reliable measure of social climate, quantifying particular elements of the climate along three dimensions: Hold and Support, Inmate Cohesion, and Experienced Safety. The sampling design, data collection strategy, and analytic procedures are discussed here.

Sampling

Fourteen medium to maximum security units were sampled from five different prisons in North Rhine-Westphalia. Five of these units functioned as sociotherapeutic units, one as a specialized young offender unit, and eight as traditional units. To facilitate a comparison of therapeutic versus nontherapeutic units in evaluation, the young offender unit was considered together with the sociotherapeutic units, as they have similar settings and rehabilitative goals. The final evaluation compared six therapeutic units with eight traditional units.

A total of 271 individuals participated in the survey, 188 prisoners and 83 staff. Participation was anonymous and no incentives were offered to respondents. Approximately half of the sample could be characterized as coming from a therapeutic unit and half from a traditional unit. The sample was also divided into inmate and staff subsamples for supplemental analysis. The participation rate among prisoners was considerably higher at 86% than among staff, of which 64% participated. In all, 85% of prisoners were male, 26% were below the age of 21, and 6% were above the age of 50. And, 74% of staff participants were male, 16% were 30 years of age or younger, and 23% above the age of 60.

Data Collection

All subjects completed the corrections-modified EssenCES, designed to measure social climate along three dimensions: Inmate Cohesion, Hold and Support, and Experienced Safety. Identical versions of the instrument were administered to both inmates and staff. In addition, a German-translated short version of the Correctional Institutions Environment Scale (S-CIES) was administered as an external validity check for the EssenCES (Dahle & Steller, 1990; Moos, 1975). The S-CIES is a 29-item institutional atmosphere assessment instrument. Items are grouped into five dimensions: rehabilitative orientation, order and cleanliness, communication between inmates and staff, discipline and control, and communication between inmates (Dahle & Steller, 1990).

Analysis

The dimensionality of the EssenCES was examined using factor analysis, specifically principle components analysis followed by varimax rotation. Subscale Homogeneity was examined using the coefficient Cronbach’s α. To assess the validity of the EssenCES as a measurement instrument for social climate, correlations between the EssenCES and S-CIES subscales were examined.

Results

Dimensionality of the EssenCES and Subscales’ Internal Consistency

The Kasier–Meyer–Olkin (KMO) measure of sampling adequacy (MSA) criterion and Bartlett’s test of sphericity were used to evaluate the appropriateness of these data for factor analysis. The items of the EssenCES feature substantial intercorrelation, MSA = .85; χ²(105, N = 230) = 1,250.3, p < .001. The sample is sufficiently large (n ≥ 150), and five items are expected to load on each factor. Factor loadings after varimax rotation resulting from factor analysis are displayed in Table 1.

Table 1.

Dimensionality of the EssenCES.

Item	Factor loadings
Item	Factor 1Staff support	Factor 2Experienced safety	Factor 3Inmates cohesion
Inmates’ Cohesion 1			.72
Inmates’ Cohesion 2
Inmates’ Cohesion 3			.57
Inmates’ Cohesion 4			.56
Inmates’ Cohesion 5			.78
Experienced Safety 1		.60
Experienced Safety 2		.71
Experienced Safety 3		.63
Experienced Safety 4		.45
Experienced Safety 5		.67
Hold and Support 1	.66
Hold and Support 2	.74
Hold and Support 3	.84
Hold and Support 4	.60
Hold and Support 5	.74

Note. Principle components analysis followed by varimax rotation (n = 271). Loadings <.40 are omitted. EssenCES = Essen Climate Evaluation Schema.

Item “Inmate Cohesion 2” does not reach a substantial loading on any of the three factors. The loading score for “Experienced Safety 4” is rather low as well. All other items demonstrated sufficient loadings on their respective expected factors. No item demonstrated unexpected loading on any factor. The pattern of loadings may therefore be interpreted as supportive of the conceptual validity of the instrument: “Hold and Support” (explaining 28.7% of the items’ variance), “Experienced Safety” (11.0%), and “Inmate Cohesion” (8.7%). Factor analysis of the inmate subsample (n = 188) data produced results similar to the overall sample, but are omitted here to accommodate space considerations.

In addition, the factorial structure of the S-CIES, as posited by Dahle and Steller (1990), was examined. The five-dimensional social climate suggested by their results was not confirmed in this analysis. Instead, principal components analysis yielded eight factors. Only three of the subscales suggested by Dahle and Steller (1990) were somewhat maintained. Five of the eight items of “Treatment Orientation” loaded clearly on the first resulting factor, four of the five items of “Communication Between Inmates” loaded clearly on the second factor, and two of the three items of “Order and Cleanliness” loaded on the third factor. Items comprising the other dimensions of the S-CIES loaded nonsystematically across multiple factors (e.g., “Communication Between Staff and Inmates” Cohesion loaded across four factors).

The internal consistency of the EssenCES subscales was examined using Cronbach’s α. This coefficient corresponds with the average item intercorrelation within a scale and is also used as an indicator of a scale’s reliability (Bortz & Döring, 1995). Cronbach’s α >.70 is generally considered sufficient for demonstrating internal consistency for individual assessments, although lower values might be acceptable when making group comparisons (Bortz & Döring, 1995). Cronbach’s α values for the three dimensions of the EssenCES were sufficiently high, as measured in the total sample: “Experienced Safety” = .77, “Inmate Cohesion” = .85, and “Hold and Support” = .85.

Correlations Between the EssenCES and the S-CIES

Although factor analysis failed to support the dimensionality of the S-CIES reported by Dahle and Steller (1990), the S-CIES and the EssenCES do measure related phenomena. For this reason, correlations between the two instruments may serve as an external check on the validity of the EssenCES subscales. Table 2 presents correlation coefficients between the five dimensions posited by Dahle and Steller (1990) and the three dimensions of the EssenCES in the current analysis.

Table 2.

Correlations Between EssenCES and S-CIES Subscales.

S-CIES subscale	EssenCES subscale
S-CIES subscale	Inmates’ cohesion	Experienced safety	Hold and support
Treatment Orientation	.34**	.29**	.69**
Order and Cleanness	−.14*	−.16**	−.00
Openness of Communication Between Inmates and Staff	.27**	.14*	.63**
Discipline and Control	−.12*	−.02	−.20**
Openness of Communication Between Inmates	.66**	.20**	.50**

Note. EssenCES = Essen Climate Evaluation Schema; S-CIES = short version of the Correctional Institutions Environment Scale.

p < .05. **p < .01.

There were several strong and significant correlations between the subscales of the two instruments: “Inmate Cohesion” and “Communication Between Inmates” (r = .66), “Hold and Support” and “Treatment Orientation” (r = .69), and “Hold and Support” and “Communication Between Inmates and Staff” (r = .63). The “Experienced Safety” dimension of the EssenCES failed to demonstrate any comparable correlation with the subscales of the S-CIES.

Comparison of Traditional and Treatment Units Across EssenCES Dimensions

As stated above, six of the participating units were designated “treatment units” (i.e., therapeutic units). The remaining eight were designated “traditional units.” The following analysis excludes female and adolescent prisoners. Supplemental analyses including these individuals did not significantly vary from the results displayed below.

The comparison between treatment units and traditional units is presented below for each dimension of the EssenCES. Figure 1 displays findings of this comparison for the “Inmate Cohesion” dimension. The type of unit (treatment vs. traditional) or type of respondent (prisoner vs. staff) did not reach statistical significance as main effects when conducting ANOVA. However, the interaction term (Unit Type × Respondent Type) was statistically significant (F = 6.3, p < .05), with prisoners’ perception of inmate cohesion slightly higher in traditional units as opposed to treatment units. Perceptions of inmate cohesion changed in the opposite direction among staff, with staff perceiving greater inmate cohesion in the treatment units.

Figure 1.

Comparison of traditional units and treatment units—Inmates’ cohesion.

Figure 2 illustrates findings of the comparison between treatment and traditional units on the “Experienced Safety” dimension of the EssenCES. ANOVA yielded a statistically significant difference between prisoners and staff on this dimensions (F = 5.6, p < .05), with staff reporting lower levels of “Experienced Safety” than prisoners overall. Scores on this dimension were significantly lower among staff in traditional units than those in treatment units.

Figure 2.

Comparison of traditional units and treatment units—Experienced safety.

Figure 3 displays findings of the comparison between traditional and treatment units on the “Hold and Support” dimension of the EssenCES. On this dimension of social climate, between-group differences are even more distinct than on the other two dimensions. First, staff perceptions of the level of hold and support were significantly higher than prisoners’, regardless of the type of unit (F = 9.9, p < .01). Second, and perhaps more importantly, the level of hold and support perceived by both groups was significantly higher in the treatment units than in the traditional units (F = 36.5, p < .001). The difference in perceptions of this dimension from traditional to treatment unit between staff and prisoners did not reach statistical significance.

Figure 3.

Comparison of traditional units and treatment units—Hold and support.

Discussion of Findings

This research project was intended to examine the differences in social climate between traditional units and treatment units in German prisons, using the EssenCES. The findings of this research lend further evidence-based support to the use of this instrument as a measure of social climate. The trait structure of the questionnaire was largely confirmed by factor analysis. Almost all items reached medium to high factor loadings on their expected dimension. The exception here was Item 2 of the “Inmate Cohesion” dimension. This item was worded negatively in the version of the instrument administered in this study: “Inmates begrudge each other everything.” Subsequently, this item has been modified to a positive wording: “Inmates care about their fellow inmates’ problems.”

Item 4 of the “Experienced Safety” dimension also failed to load as strongly as the other items in this dimension. The original wording of this item may have affected respondents’ answers: “Staff are afraid of some inmates.” Not only in the prison setting but also in the forensic psychiatry settings where this instrument was initially administered, staff seem reluctant to identify themselves as “afraid” of inmates/patients, due to the negative connotation of the word. This item has since been revised too: “At times, members of staff feel threatened by some of the inmates.”

While statistical analyses revealed certain psychometric flaws within the EssenCES, data obtained from subsequent research using a modified English version of the instrument have indicated that the adjusted items improved and stabilized the instrument’s factor structure (Day et al., 2012; Milsom et al., 2014; Tonkin et al., 2012). Despite the weaknesses of the unrevised version used here, the EssenCES’s structure still performed better than the five-trait-structure of the S-CIES proposed by Dahle and Steller (1990).

The construct validity of the corrections-modified version of the EssenCES was confirmed by the current research. This held most true for the subscales “Inmate Cohesion” and “Hold and Support,” which showed strong correlations with their S-CIES subscale counterparts. The subscale “Experienced Safety” demonstrated less correlation with the S-CIES, a finding consistent with previous research using the EssenCES in forensic psychiatry wards (Schalast et al., 2008). In that project, “Experienced Safety” failed to demonstrate strong correlation with any subscale of the WAS; however, it did correlate strongly with the number of previous disciplinary events in a ward. It would seem that the Moos scales were not designed to measure the threat of aggression or violence with the same emphasis as the EssenCES. While the number of previous disciplinary events in each unit was not part of the data collected, the comparison of traditional and treatment units does lend some additional support to the validity of the “Experienced Safety” subscale. As would be expected, staff and inmates in treatment units reported greater perceptions of this dimension.

Conclusion and Limitations

Overall, these findings do not support the criticisms such as those made by Rehn (2008) that the rehabilitative goals originally envisioned for the German prison system have been case aside. The data suggest a more positive social climate, conducive to rehabilitation, in treatment units. This observation was particularly true for the “Hold and Support” dimension, especially among staff. The latter finding is consistent with previous research indicating that staff generally has a more positive perception of the climate than inmates or patients (Moos, 1974). In the current survey, staff reported higher levels of all three dimensions than did inmates.

This research provides further evidence for the validity of the corrections-modified EssenCES as a measure of the three dimensions of a prison’s social climate. Results from subsequent studies, using the modified items discussed above, suggest even more promising psychometric properties of the instrument. In addition, the economic advantage of the EssenCES, at a brief 17 items, over more lengthy, complex instruments relied on in the past is considerable. This is especially true in cases where multiple measurements of the social climate of an institution are needed over time.

The prison units that participated in this study should not be considered as representative of all “traditional” German prison units or all “treatment” German prison units. Future research may benefit from the inclusion of a wider sampling of units from across the country. Future research may also benefit from comparing units of traditional prisons, treatment units attached to traditional prisons, and independently functioning treatment prisons.

Our findings are tentatively supportive of the atmosphere in modern German prison treatment units. However, as is generally the case in social climate research, readers should be cautious in attempting to draw causal conclusions from these findings, which are correlative in nature. Social therapeutic units may benefit from better staffing, from staff’s skills and education, from the organization of everyday living together, and from the therapeutic program. The selection of prisoners accepted for the specialized units may also play a role. Differences of social climate ratings do not only reflect differences of institutions (e.g., units, wards, etc.) but also differences of individuals rating these institutions (Moos, 1974). Nonetheless, the findings presented here allow for the preliminary conclusion that treatment units in the German prison system are perceived as more supportive and rehabilitative than traditional prison units by inmates and staff. The EssenCES is also evidenced as a valid scale to measure these differences in social climate.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Norbert Schalast, PhD, is a psychologist and researcher, Institute of Forensic Psychiatry, University of Duisburg & Essen, Germany. He has been at the Institute of Forensic Psychiatry since 1992. He received his PhD at the Technical University of Braunschweig.

Jacob Michael Laan is currently the Research and Statistics Coordinator for the Arkansas Department of Correction. He is also a fourth year doctoral student at the University of Arkansas at Little Rock.

References

Andrews

D. A.

Bonta

(2010). The psychology of criminal conduct (5th ed.). Albany, NY: Anderson.

Bortz

Döring

(1995). Forschungsmethoden und Evaluation [Methods of research and evaluation, book]. Berlin, Germany: Springer. (In German)

Dahle

K. P.

Steller

(1990). Coping in prison. Zeitschr Exp Angew Psychol, 37, 31-51. (In German)

Davies

(2004). Toxic institutions. In Campling

Davies

Farquharson

(Eds.), From toxic institutions to therapeutic environments: Residential settings in mental health services (pp. 20-31). London, England: Gaskell.

Day

Casey

Vess

Huisy

(2011). Assessing the social climate of prisons. Canberra, Australia: Criminology Research Council. Retrieved from www.criminologyresearchcouncil.gov.au/reports/02-0910.pdf

Day

Casey

Vess

Huisy

(2012). Assessing the therapeutic climate of prisons. Criminal Justice and Behavior, 39, 156-168.

Denny

Costello

R. M.

Cochran

M. L.

(1984). Factor-implied scales of ward atmosphere. Evaluation & the Health Professions, 7, 181-192.

Foucault

(1977). Discipline and punish: The birth of the prison. New York, NY: Pantheon Books.

Howells

Tonkin

Milburn

Lewis

Draycot

Cordwell

. . . Schalast

(2009). The EssenCES measure of social climate: A preliminary validation and normative data in UK high secure hospital settings. Criminal Behaviour and Mental Health, 19, 308-320.

10.

Lewis

W. D.

(2009). From Newgate to Dannemora: The rise of the penitentiary in New York, 1796-1848. Ithaca, NY: Cornell University Press.

11.

Middelboe

Schjodt

Byrsting

Gjerris

(2001). Ward atmosphere in acute psychiatric in-patient care: Patients’ perceptions, ideals and satisfaction. Acta Psychiatrica Scandinavica, 103, 212-219.

12.

Milsom

S. A.

Freestone

Duller

Bouman

Taylor

(2014). Factor structure of the Essen Climate Evaluation Schema measure of social climate in a UK medium-security setting. Criminal Behaviour and Mental Health, 24, 86-99.

13.

Moos

R. H.

(Ed.). (1974). Evaluating treatment environments: A social ecological approach. New York, NY: Wiley.

14.

Moos

R. H.

(1975). Evaluating correctional and community settings. New York, NY: Wiley.

15.

Moos

R. H.

(1996). Ward Atmosphere Scale: Developments, applications, research. Redwood, CA: Mind Garden.

16.

Moos

R. H.

(1997). Evaluating treatment environments: The quality of psychiatric and substance abuse programs. New Brunswick, NJ: Transaction.

17.

Moos

R. H.

Houts

P. S.

(1968). Assessment of the social atmospheres of psychiatric wards. Journal of Abnormal Psychology, 73, 595-604.

18.

Pedersen

Karterud

(2007). Associations between patient characteristics and ratings of treatment milieu. Nordic Journal of Psychiatry, 61, 271-278.

19.

Rehn

(2008). The social therapeutic institution—The different prison? Kriminologisches Journal, 40(1), 42-53. (In German)

20.

Røssberg

J. I.

Friis

(2003). A suggested revision of the Ward Atmosphere Scale. Acta Psychiatrica Scandinavica, 108, 374-380.

21.

Schalast

(1995). Stress and strain in general and forensic psychiatry (Doctoral disseratation). Technical University Braunschweig, Germany. (In German).

22.

Schalast

Groenewald

(2009). Ein Kurzfragebogen zur Einschätzung des sozialen Klimas im Strafvollzug [A short questionnaire to assess the social climate in prisons] (Erste Befunde auf Abteilungen des Regelvollzugs und der Sozialtherapie). In Haller

Jehle

J. M.

(Ed.), Drogen – Sucht – Kriminalität (pp. 329-352). Mönchengladbach, German: Forum. (In German)

23.

Schalast

Redies

Collins

Stacey

Howells

(2008). EssenCES, a short questionnaire for assessing the social climate of forensic psychiatric wards. Criminal Behaviour and Mental Health, 18(1), 49-58.

24.

Shefer

(2010). Doing rehabilitation in the contemporary prison—The case of one-wing therapeutic communities (Doctoral dissertation). Cambridge University, Cambridge, MA.

25.

Timko

Moos

R. H.

(2004). Measuring the therapeutic environment. In Campling

Davies

Farquharson

(Eds.), From toxic institutions to therapeutic environments: Residential settings in mental health services. London: Gaskell.

26.

Tonkin

Howells

Ferguson

Clark

Newberry

Schalast

(2012). Lost in translation? Psychometric properties and construct validity of the English Essen Climate Evaluation Schema (EssenCES) social climate questionnaire. Psychological Assessment, 24, 573-580.

27.

Wachsmann

(2002). Between reform and repression: Imprisonment in Weimar Germany. The Historical Journal, 45(2), 411-432.

28.

Wright

K. N.

Boudouris

(1982). An assessment of the Moos Correctional Institutions Environment Scale. Journal of Research in Crime & Delinquency, 19, 255-276.