The International Journal of Psychosocial Rehabilitation
 
Perceptions of a Norwegian clubhouse among its members:
A psychometric evaluation of a user satisfaction tool

 
Lisa S. Uhrmann1, MSc student
Hege Nordli2, PhD, Acting Director
Orsolya R. Fekete3, MSc, PhD student
Tore Bonsaksen4, 5, MSc, Professor

 

 
1 Munich University of Applied Sciences, Department of Applied Social Science, Munich, Germany
2 Fontenehuset i Oslo, Oslo, Norway
3 Western Norway University of Applied Sciences, Faculty of Health and Social Sciences, Department of Welfare and Participation, Bergen, Norway.
4 Oslo Metropolitan University, Faculty of Health Sciences, Department of Occupational Therapy, Prosthetics and Orthotics, Oslo, Norway
5 VID Specialized University, Faculty of Health Studies, Sandnes, Norway

 

Citation:
Uhrmann LS, Nordli H, Fekete OR & Bonsaksen T. (2017) Perceptions of a Norwegian clubhouse among its members:
A psychometric evaluation of a user satisfaction tool  . I
nternational Journal of Psychosocial Rehabilitation. Vol 21 (2) 82-91


 

Corresponding author: Tore Bonsaksen
 Oslo Metropolitan University
Department of Occupational Therapy
Prosthetics and Orthotics
Faculty of Health Sciences
 Postbox 4 St. Olavs Plass
0130 Oslo, NORWAY.
E-mail tore.bonsaksen@hioa.no

Abstract
Survey tools to assess user satisfaction have gained increased attention as meaningful sources of information for quality assurance and development of health care services. Several tools are developed, but these mostly refer to clinical settings and therefore address patients receiving treatment within these settings. This study aimed to investigate the psychometric properties of a survey instrument used to assess user satisfaction among members of a clubhouse and possibly to improve the instrument. Principal Components Analysis was used to assess the instrument’s factor structure, and Cronbach’s Alpha was used to assess internal consistency of the scale items. The results showed that the scales improved their psychometric properties if two items were removed from the tool, and that two different factor structures may be applied. In conclusion, the survey tool may serve as a means to assess clubhouse members’ satisfaction with various aspects of the clubhouse.

Keywords: clubhouse, factor analysis, psychometric properties, psychosocial rehabilitation, user satisfaction, validity


Introduction:
The clubhouse model is an international movement for people with mental illness (Clubhouse International, 2017). It is a work-based psychosocial rehabilitation model built on the ideas of health promotion. Furthermore, it can serve as a bridge between mental illness and active participation in society (Fekete, Kinn, Larsen, & Langeland, 2018; Stimo et al., 2015). The aim is to support people to return to ordinary work, or other forms of productive activity, through taking part in meaningful activities at the clubhouse. As a systematic review showed that clubhouses are effective in promoting employment, reducing hospitalizations and improving quality of life (McKay, Nugent, Johnsen, Eaton, & Lidz, 2016), it appears they are successful in reaching that aim.

Clubhouse participants are referred to as members rather than patients or clients (Doyle, Lanoil, & Dudek, 2013; Stimo et al., 2015). The concept of membership is fundamental to the model (Propst, 2016), and serves as a means to generate shared ownership, involvement and responsibility (Raeburn, Halcomb, Walter, & Cleary, 2013). At the clubhouse, members and staff support each other work side by side on shared tasks (Stimo et al., 2015). The emphasis is on identifying and developing personal strengths, not on remediating clinical symptoms (Raeburn et al., 2013). Taking part in meaningful activities in their own pace, and with regard for their personal interests, is considered an important aspect (McKay et al., 2016; Stimo et al., 2015). Members are valued, productive colleagues and an important part of the success of the clubhouses (Bonsaksen et al., 2016; Clubhouse International, 2017). Hence, the clubhouses depend on their members’ involvement and participation. User participation is also acknowledged as necessary for the development and delivery of health care services, and users’ involvement and active participation have been shown to contribute to better treatment results (Almeida, Bourliataux- Lajoinie, & Martins, 2015). Ensuring that people are involved in their own treatment is therefore considered an important component of high quality healthcare services (Almeida et al., 2015). The fact that rehabilitation at clubhouses operate outside the public health care sector does not make the argument for user participation any less relevant.

Although there is no gold standard for how quality of services should be evaluated, monitoring the users’ own satisfaction with the services is highly relevant (Al-Abri & Al-Balushi, 2013). The concept of satisfaction and its measurement appear to be widely used, as they are important for healthcare administrators, health professionals and users. Observing user satisfaction is essential for quality assurance purposes and because satisfaction can affect health outcomes. It has been shown that satisfied patients are more optimistic about their situation, more compliant and more likely to take an active part in their recovery (Crow et al., 2002). In accordance with this view, the national stepping-up program for mental health in Norway considered user participation essential to secure the outcome of health services (Sverdrup, Myrvold, & Kristofersen, 2005). Standardized questionnaires appear useful and have been frequently used in patient satisfaction studies (Almeida et al., 2015). Several tools for assessing user satisfaction have been developed and used, but these mostly refer to clinical settings and therefore address the patients receiving treatment within these settings (Crow et al., 2002; Thørrisen, Nordli, Fekete, & Bonsaksen, 2018). According to Al-Abri and Al- Balushi (2013), such instruments developed for healthcare contexts may not be valid nor reliable in other settings of care.

Currently, there are over 300 clubhouses in over 30 countries worldwide, including 13 in Norway (Clubhouse International, 2017; Fontenehus Norge, 2017). As the number of clubhouses in Norway is growing, developing a suitable instrument for measuring user satisfaction is important. There are no perfect tools and psychometric studies of relevant instruments are scarce (Almeida et al., 2015). Boyer and colleagues (2009) postulated that the validation of a satisfaction measurement tool is challenging. Systematic reviews have shown that instruments used to evaluate satisfaction, mainly demonstrated a lack of validity or reliability. Only eleven studies, constituting 6% of the 181 included studies, included measures of the instrument’s validity and reliability (Sitzia, 1999). To improve the instruments used for measuring user satisfaction, and thereby to ensure their credibility, it is therefore essential to examine their psychometric properties. One previous study has indicated that the instrument currently used for measuring user satisfaction among clubhouse members in Norway is a useful tool (Thørrisen et al., 2018). However, the authors concluded that the results should be interpreted carefully, and that further studies with other samples were needed. This study continues the efforts in establishing a valid and reliable tool for measuring user satisfaction among clubhouse members in Norway.

Study aim
The aim of the study was to investigate the psychometric properties of the clubhouse instrument for measuring user satisfaction, and potentially to improve the instrument.

Methods
Design and data collection
The study had a cross-sectional design. All data were self-reported by the members of one clubhouse in Norway, and were collected by clubhouse staff and members during the winter of 2016.
Sample
All members of the clubhouse were eligible participants in the study, and there were no exclusion criteria. Active members were defined as those who had visited the clubhouse at least once during the last 90 days. At the time of the data collection, there were 151 active members of the clubhouse, of which 94 (62.3 %) opted to participate. Basic characteristics of the sample are displayed in Table 1.

Table 1 Sociodemographic characteristics of the participants (n = 94)

 

n

%

Age group

94

100

21-30 years

11

11.7

31-40 years

25

26.6

41-50 years

25

26.6

51-60 years

23

24.5

61 years and above

10

10.7

Gender

94

100

Men

45

47.9

Women

47

50.0

Transperson

2

2.1

Education (highest completed level)

93

98.9

Elementary school or high school

58

62.4

College or university (BSc level or higher)

35

37.2

Work and income

94

100

In paid work

19

20.2

Time-limited disability pension

36

38.3

Disability pension

41

43.6

Note. On variables with some missing responses (i.e., n < 94), the valid percent is reported.


Measurement
The standard and annually employed member survey contains a range of questions and statements to which respondents are asked to answer and/or rate their level of agreement. Sections of the member survey concern sociodemographic characteristics, duration of membership and use of the clubhouse, sources of income, work and education, and perceived impact from using the clubhouse.

Finally, one section with eight statements, all of which rated from 1 (totally disagree) to 5 (totally agree), constitute the instrument for measuring user satisfaction. This study focuses on the user satisfaction measure alone. A factor analytic study based on the 2015 member survey found that the eight statements “loaded” on two different factors, and they might therefore constitute two different scales (Thørrisen et al., 2018). The scales were labeled “personal outcome” and “user involvement”. Factor loadings for the items on the personal outcome scale were between 0.65 and 0.75, and between 0.48 and 0.80 on the user involvement scale. There were no cross-loadings. In total, 52.2 % of the data variance was explained by the two factors.

Data analysis
As there has been only one previous factor-analytic study of the user satisfaction instrument, the starting point of the current study was to replicate the psychometric investigation of it. When assessing latent factors, an exploratory Principal Components Analysis (PCA) was performed. The Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy (Kaiser, 1974) and Bartlett’s Test of Sphericity (Bartlett, 1954) were used to assess whether factorization was adequate. KMO values should exceed 0.60 in order to proceed (Cerny & Kaiser, 1977; Kaiser, 1974). Factor extraction was determined by visually inspecting the scree-plots, and by assessing the Eigenvalue (λ) estimates and the data variance explained by the factors. According to statistical convention, factors with λ > 1 and/or factors explaining more than 10 % the variables’ variance proportions were retained.

In addition, Parallel Analysis was used (Horn, 1965) – this analysis is known to be more restrictive with a view to how many factors to extract, and is therefore used as a means to ascertain that one does not overestimate the number of extracted factors (Zwick & Velicer, 1986). The method proposes that one should retain factors with an actual λ exceeding the randomly generated λ of the corresponding factor in a random dataset with the same number of variables and respondents.

Statistical measures reported from the factor analyses also include communalities, indicating the variance proportion of each variable explained by the factors together, and factor loadings as estimates of the impact from a given variable on each factor. Factor loadings > 0.40 were considered high. Internal consistency of the scales was examined with Cronbach’s α, and coefficients exceeding 0.70 were considered satisfactory. In cases of cross-loadings (loadings > 0.40 on more than one factor) or low factor loadings or communalities, theoretical considerations about the conceptual content of each factor would assist in deciding whether items should be discarded from the relevant scale or not.

Ethics
All members of the clubhouse were informed about the survey by the clubhouse staff and research-active members. Participation was voluntary, the data were collected anonymously, and completing and returning the survey implied informed consent to participate. As the collected data was anonymous and not related to health and/or illness, formal approval from the Regional Ethics Committee for Healthcare Research or the Data Protection Official for research was not required.

Results
Internal consistency and factor structure of the original scales
When investigating the internal consistency of the items constituting the personal outcome scale (items 1-4), this was found to be at a satisfactory level (Cronbach’s α = 0.75). However, internal consistency of the user involvement scale items (items 5-8) was in the lower range (Cronbach’s α = 0.62).

As the initial step in the PCA, the KMO value was 0.77, and Bartlett’s test of sphericity was statistically significant (p < 0.001), so we proceeded with the analysis. The communalities of the items were between 0.23 (item 8) and 0.79 (item 6). Two factors with Eigenvalues > 1 were extracted: Factor 1, λ = 3.54, explaining 44.2 % of the data variance, and Factor 2:  λ = 1.26, explaining 15.8 % of the data variance. However, the pattern matrix revealed that the items only partly followed the pattern of the previous study. Items 1, 2, and 3 clearly loaded on the first factor, whereas item 4 loaded on the second factor, in contrast to the established solution. Items 6 and 8 clearly loaded on the second factor, as expected, but items 5 and 7 cross-loaded (loadings > 0.40 on both factors). The results from the first PCA are displayed in Table 2, along with the results of the 2015 survey (for comparison).

Table 2  Factor structure of the Clubhouse User Satisfaction Survey established in 2015

2015 member survey data

2016 member survey data

 

Item #

Factor 1

Factor 2

Communalities

Item #

Factor 1

Factor 2

Communalities

1

Meaningful work

0.75

0.07

0.60

1

0.91

-0.16

0.78

2

Can use resources

0.72

-0.07

0.49

2

0.85

-0.06

0.73

3

Feel useful

0.69

0.14

0.57

3

0.83

0.09

0.70

4

Belief in return to work

0.65

-0.04

0.40

4

0.18

0.55

0.38

5

Influence on my workday

-0.17

0.80

0.57

5

0.46

0.45

0.52

6

Participate in meetings

0.13

0.75

0.65

6

0.08

0.87

0.79

7

Take part in decisions

0.01

0.74

0.56

7

0.57

0.46

0.68

8

Want more support

0.18

0.48

0.33

8

-0.18

0.49

0.23

Eigenvalue (λ)

2.99

1.18

 

 

3.54

1.26

 

Explained variance

37.4 %

14.8 %

 

 

44.2 %

15.8 %

 

Total explained variance

52.2 %

 

 

60.0 %

 

Note. Factor structure derived from explorative PCA using the Direct Oblimin rotation method with Kaiser normalization. Factor loadings are from the pattern matrix.

 


In view of these results, we decided to discard items 4 and 8 from their respective scales. The reasons were that item 4 had low communality (0.38) and did not load as expected. In addition, we found that it did not theoretically belong to the other “personal outcome” items – it rather reflected a belief in a future outcome. Item 8 was also discarded due to its low commonality (0.23), and because it did not theoretically belong to the three other “user involvement” items – it rather reflected a wish that staff would be more involved in assisting the members in getting a job. For the same theoretical reasons, wanting the items to reflect the conceptual content “personal outcome” or “user involvement” as clearly as possible, items 5 and 7 were retained, in spite of their high loadings on both factors.

Internal consistency and factor structure of the revised scales
The revised scales, omitting items 4 and 8, were then analyzed for internal consistency. The personal outcome scale, consisting of items 1, 2, and 3, showed Cronbach’s α = 0.84. The user involvement scale, consisting of items 5, 6, and 7, had Cronbach’s α = 0.75. In the eventual case of a one-factor solution being fitted to the data, Cronbach’s α of all six items together was 0.84.

As the first step in the second PCA, we found that the KMO value was 0.78, and Bartlett’s test of sphericity was statistically significant (p < 0.001). The items’ communalities ranged between 0.64 (item 5) and 0.81 (item 1). Two factors with Eigenvalues > 1 were extracted: Factor 1, λ = 3.35, explaining 55.8 % of the data variance, and Factor 2:  λ = 1.11, explaining 18.6 % of the data variance. The pattern matrix showed no cross-loadings, and all items loaded substantially on the proposed factor, with loadings ranging between 0.67 (item 7) and 0.94 (item 6).

However, when controlling the factor extraction with the Parallel Analysis, we found a randomly generated λ = 1.18, which was higher than the λ found for the second extracted factor in the PCA. Thus, the Parallel Analysis suggested that only one factor should be extracted. As a result, we conducted a third PCA using a confirmatory approach where we specified that only one factor should be extracted. This analysis yielded commonalities ranging between 0.32 (item 6) and 0.68 (items 2 and 7). All factor loadings were substantial, ranging from 0.56 (item 6) and 0.83 (item 2). The results from the PCAs after the removal of two items are displayed in Table 3. The one-factor solution and the two-factor solution are both shown in Table 3, and the revised Clubhouse User Satisfaction Survey is displayed in Table 4.

Table 3

Factor structure of the Clubhouse User Satisfaction Survey as revised based on the 2016 survey data

Two-factor structure

One-factor structure

 

Item #

Factor 1

Factor 2

Communalities

Item #

Factor 1

Communalities

1

Meaningful work

0.93

-0.08

0.81

1

0.77

0.59

3

Feel useful

0.88

-0.02

0.76

3

0.76

0.58

2

Can use resources

0.75

0.20

0.73

2

0.83

0.68

6

Participate in meetings

-0.19

0.94

0.78

6

0.56

0.32

5

Influence on my workday

0.18

0.71

0.64

5

0.71

0.51

7

Take part in decisions

0.34

0.67

0.75

7

0.82

0.68

Eigenvalue (λ)

3.35

1.11

 

 

3.35

 

Cronbach’s α

0.84

0.75

 

 

0.84

 

Explained variance

55.8 %

18.6 %

 

 

55.8 %

 

Total explained variance

74.4 %

 

 

55.8 %

 

Note. Two-factor structure derived from explorative PCA using the Direct Oblimin rotation method with Kaiser normalization. One-factor structure derived from confirmative PCA with the fixed extraction of one factor only. Factor loadings are from the pattern matrix.

 


Table 4

The revised Clubhouse User Satisfaction Survey

To what extent do you agree with the following statements?

 

Personal outcome

1

I do meaningful work at the clubhouse

2

I can use my resources at the clubhouse

3

I feel useful at the clubhouse

 

User involvement

4

I feel that I have an influence on my workday at the clubhouse

5

At the clubhouse, I can participate in all meetings where important matters are discussed

6

I can take part in making decisions in matters that concern the clubhouse

Note. All statements are rated: 1 = totally disagree, 2 = partly disagree, 3 = not sure, 4 = partly agree, 5 = totally agree. The survey can be used with a one-factor structure: User satisfaction = sum of all item scores. Alternatively, it can be used with a two-factor structure: Personal outcome = sum score (items 1, 2, 3), and User involvement = sum score (items 4, 5, 6).


Discussion

This is the first study to investigate the psychometric properties of the revised Clubhouse User Satisfaction Survey tool. The aim of the study was to validate and potentially improve the instrument. The results achieved from the revision of the questionnaire (after discarding items 4 and 8) showed that the new version had good psychometric properties, and functioned better than the previous one.

When replicating the psychometric investigation of the original instrument, item 4 (belief in return to work) did not load on the “personal outcome” factor, as it did in the analysis of the 2015 data (see Table 2). Instead, the item loaded on the “user involvement” factor, but this seems not to be coherent with the content of the remaining three items belonging to this factor. Similarly, item 8 asks the participants to comment on their need for more support. This statement does not reflect user involvement as such, but rather refers to staff involvement. As a result, items 4 and 8 were considered not to contribute properly to their respective factors, and were therefore removed from the instrument.

Although the two items were excluded from this instrument they can still be important for the purpose of securing good quality services for members at clubhouses, as previously suggested (Thørrisen et al., 2018). The removed items might help to provide important information if they are used as separate questions in another context. However, the outcome of this analysis indicates that items 4 and 8 should be kept aside from the Clubhouse User Satisfaction Survey tool, as they do not reflect well the latent factors in this tool.

The results of the initial factor analysis of the revised instrument showed a two-factor structure. The pattern matrix was clean with no cross-loading items; all items loaded on the theoretically proposed factor, and the measures of internal consistency were good. The multidimensional construct matches the findings of international investigations into patient satisfaction tools (Almeida et al., 2015, Gonzales et al., 2005), indicating that patient satisfaction may be one too broad concept. Rather, patient satisfaction appears to be specific to a relevant domain. In this study, items 1 (meaningful work), 2 (feel useful) and 3 (can use resources) loaded on factor 1. In line with Thørrisen and colleagues (2018), we label this factor “personal outcome”. Items 5 (influence on my workday), 6 (participate in meetings) and 7 (take part in decisions) loaded on factor 2, representing “user involvement” more broadly.

However, the parallel analysis suggested that a one-factor solution might be preferred. This implies that all items would load on the same latent factor, such that a general measure of patient satisfaction derived from these items would in fact be useful. The subsequent analysis, where one factor was the pre-determined solution, confirmed that all items had substantial loadings (i.e., < 0.40) on only one factor. All items were also intrinsically related, as shown from the internal consistency analysis (see Table 3). This adds to the existing literature concerned with the Clubhouse User Satisfaction Survey tool (Thørrisen et al., 2018), demonstrating that the tool may be used in different ways, depending on the purpose of the inquiry. The survey tool may be used to assess clubhouse members’ perceptions of personal outcome from being a member, and their perceptions of user involvement as members of the clubhouse. This would require using the two-factor structure. Alternatively, one may employ the one-factor structure to assess clubhouse members’ general perceptions of satisfaction with being a member.

Study limitations
The study sample was small and one of convenience. All data were collected from members of one particular clubhouse in Norway. The questionnaire was available only to those who visited the clubhouse during the time of the data collection. It is therefore possible that members from other clubhouses, or members who did not attend the clubhouse during the data collection period, would rate the survey items differently. As a result, generalizations based on the study results should be made with caution.
This study demonstrated good psychometric properties of the survey tool. However, the study is limited in its employment of only two statistical procedures. Analysis of the tool’s internal consistency and factor structure do not provide a satisfactory validation of the instrument (Thørrisen et al., 2017). As the study is based on a cross-sectional design, test- retest reliability cannot be measured. Further research should include data collected from members of different clubhouses.

Conclusion
In this study, we found that the Clubhouse User Satisfaction Survey tool improved its measurement properties if two items were removed. The improvements were no cross-loadings, higher factor loadings, increased explained variance, and the explicitly stated possibility of using the survey tool in two different ways. Those using the tool may employ a two-factor structure to assess clubhouse members’ perceptions of personal outcome and user involvement, or they may employ a one-factor structure to assess user satisfaction more broadly.

Funding
The study received no funding.

Conflicts of interest
There are no conflicts of interest related to this article.



References
Al-Abri, R., & Al-Balushi, A. (2013). Patient satisfaction survey as a tool towards quality improvement. Oman Medical Journal, 19(2), 3-7.

Almeida, R. S., Bourliataux- Lajoinie, S., & Martins, M. (2015). Satisfaction measurement instruments for healthcare service users: A systematic review. Cadernos de Saúde Pública, 31(1), 11-25.
 doi:10.1590/0102-311X00027014 
Bartlett, M. S. (1954). A note on multiplying factors for various chi square approximations. Journal of the Royal Statistical Society, 16(2), 296-298.

Bonsaksen, T., Fouad, M., Skarpaas, L. S., Nordli, H., Fekete, O., & Stimo, T. (2016). Characteristics of Norwegian clubhouse members and factors associated with their participation in work and education. British Journal of Occupational Therapy, 79(11), 669-676. doi:10.1177/0308022616639977

Boyer, L., Baumstarck- Barrau, K., Cano, N., Zendjidjian, X., Belzeaux, R., Limousin, S., . . . Auquier, P. (2009). Assessment of psychiatric inpatient satisfaction: A systematic review of self-reported instruments. European Psychiatry, 24(8), 540-549. doi:10.1016/j.eurpsy.2009.05.011

Cerny, B. A., & Kaiser, H. F. (1977). A study of a measure of sampling adequacy for factor-analytic correlation matrices. Multivariate Behavioral Research, 12(1), 43-47. doi:10.1207/s15237906mbr1201_3
Clubhouse International. (2016). International standards for clubhouse programs. Retrieved from https://www.clubhouse-intl.org/documents/standards_2016_eng.pdf

Clubhouse International. (2017). Creating community: Changing the world of mental health.   Retrieved from http://clubhouse-intl.org/what-we-do/what-clubhouses-do/

Crow, R., Gage, H., Hampson, S., Hart, J., Kimber, A., Storey, L., & Thomas, H. (2002). The measurement of satisfaction with healthcare: implications for practice from a systematic review of the literature. Health Technology Assessment, 6(32), 1-244.

Doyle, A., Lanoil, J., & Dudek, K. J. (2013). Fountain House: Creating community in mental health practice. New York: Columbia University Press.

Fekete, O., Kinn, L. G., Larsen, T. M. B., & Langeland, E. (2018). Exploring the clubhouse model in a salutogenic perspective (submitted for publication).

Fontenehus Norge (2017). http://www.fontenehus.no

Garbo, G. L., & Jackbo, A. (2012). A source of growth. The Clubhouse - a work community for persons with mental health challenges [En kilde til vekst. Fontenehuset - et arbeidsfellesskap for mennesker med psykiske utfordringer]. Oslo: Oslo Clubhouse [Fontenehuset i Oslo].

Gonzalez, N, Quintana, J. M., Bilbao, A., Escobar, A., Aizpuru, F., Thompson, A., Esteban, C.,     San Sebastian, J. A. & de la Sierra, E. (2005). Development and validation of an in-    patient satisfaction questionnaire. International Journal for Quality in Health Care, 17     (6), 465-472

Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30(2), 179-185. doi:10.1007/BF02289447

Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39(1), 31-36. doi:10.1007/BF02291817

McKay, C., Nugent, K. L., Johnsen, M., Eaton, W. W., & Lidz, C. W. (2016). A systematic review of evidence for the clubhouse model of psychosocial rehabilitation. Administration and Policy in Mental Health and Mental Health Services Research, 1-20. doi:10.1007/s10488-016-0760-3

Propst, R. N. (1997). Stages in realizing the international diffusion of a single way of working: The clubhouse model. New Directions for Mental Health Services (74), 53-66.

Raeburn, T., Halcomb, E., Walter, G., & Cleary, M. (2013). An overview of the clubhouse model of psychiatric rehabilitation. Australasian Psychiatry, 21(4), 376-378. doi:10.1177/1039856213492235

Sitzia, J. (1999). How valid and reliable are patient satisfaction data? An analysis of 195 studies. International Journal for Quality in Health Care, 11(4), 319-328. doi:10.1093/intqhc/11.4.319

Stimo, T., Jarål, G. B., Ellestad, A. K., Ellingham, B., Skarpaas, L. S., & Bonsaksen, T. (2015). The Clubhouse model in Norway: A method in harmony with occupational therapy theory [in Norwegian: Fontenehusmodellen i Norge: En metode i harmoni med tenkning i ergoterapifaget]. Ergoterapeuten, 58(4), 22-30.

Sverdrup, S., Myrvold, T. M., & Kristofersen, L. B. (2005). User participation and mental health. Oslo: NIBR-report 2005:6.

Thørrisen, M. M., Nordli, H., Fekete, O., & Bonsaksen, T. (2018). Systematically measuring user experiences among members of a clubhouse for psychosocial rehabilitation: A psychometric evaluation [in Norwegian: Systematisk måling av brukererfaringer blant medlemmer på Klubbhus for psykososial rehabilitering: En psykometrisk vurdering] (in press). Tidsskrift for omsorgsforskning.

Zwick, W. R., & Velicer, W. F. (1986). Comparison of five rules for determining the number of components to retain. Psychological Bulletin, 99(3), 432-442. doi:10.1037/0033-2909.99.3.432




Copyright © 2017 ADG, SA. All Rights Reserved.  
A Private Non-Profit Agency for the good of all, 
published in the UK & Honduras