The Experience in Personal Social Systems Questionnaire (EXIS.pers): Development and Psychometric Properties

The Experience in Personal Social Systems Questionnaire (EXIS.pers): Development and Psychometric Properties CHRISTINA HUNGER* ANNETTE BORNHAUSER* € LEONI LINK* JULIAN GEIGGES* ANDREAS VOSS† JAN WEINHOLD* JOCHEN SCHWEITZER* This study presents the theoretical background, development, and psychometric properties of the German and English versions of the Experience in Personal Social Systems Questionnaire (EXIS.pers). It assesses how the members of a personal social system experience their situation within that system. It is designed as a research tool for interventions in which only one member of the system participates (e.g., Family Constellation Seminars). The EXIS.pers was created to measure change on the individual level relating to one’s own important personal social system. In Study 1, we used exploratory factor analysis (EFA) for latent variable identification of the original German EXIS.pers (n = 179). In Studies 2 and 3, we used confirmatory factor analysis (CFA) to examine the dimensionality of the German (n = 634) and English (n = 310) EXIS.pers. Internal consistencies and cross-cultural structural equivalence were assessed. EFA indicated that a four-factor model provided best fit for the German EXIS.pers. For both the German and English EXIS.pers, CFA provided the best fit for a five-factor bi-level model that included a general factor (Experience In Personal Social Systems) and four dimensions (Belonging, Autonomy, Accord, Confidence). Good internal consistencies, external associations, and cross-cultural structural equivalence were demonstrated. This study provides first evidence for the *Institute of Medical Psychology, Center for Psychosocial Medicine, University Hospital Heidelberg, Heidelberg, Germany. † Institute of Psychology, Quantitative Research Methods, Ruprecht-Karls-University Heidelberg, Heidelberg, Germany. Correspondence concerning this article should be addressed to Christina Hunger, Institute for Medical Psychology, Center for Psychosocial Medicine, University Hospital Heidelberg, Bergheimer Straße 20, D69115 Heidelberg, Germany. E-mail: christina.hunger@med.uni-heidelberg.de All authors have agreed to authorship in the indicated order. There has been no prior publication, or the nature of any prior publication, and no financial interest in the research. Ethical approval: The research was approved by the Ethics Committee of the Heidelberg Medical Faculty (S-178/2011). Publication of this article has been made possible by generous funding from the DFG (German Research Foundation) within the framework of the Collaborative Research Center (SFB) 619 “Dynamics of Ritual,” University of Heidelberg. We thank Diana Drexler and Gunthard Weber for sharing what they expected to be the positive systemic outcomes of their work and for continuously providing valuable advice during the development of the EXIS. We thank Michael Kerman, the director of child psychotherapy at KIDS Company, for his very generous support. Fletcher Dubois, Vanessa Gordon, Marcel Laqueur, Carol Bornhauser, Mary Beth € Robinson, and Hilary Niehues as our bilinguals translated the EXIS. We are very thankful for their critique during the translation process. Finally, with meticulous corrections and suggestions for revision, Anja Sander critically revised our research methods and statistical analyses, and Mary Beth Robinson contributed greatly to the clarity and readability of our paper. 1 Family Process, Vol. x, No. x, 2016 © 2016 Family Process Institute doi: 10.1111/famp.12205 German and English EXIS.pers as an economical and reliable measure of an individual’s experience within his or her personal social systems. Keywords: Experience in Personal Social Systems (EXIS.pers); Scale Development; Confirmatory Factor Analysis; Structural Equivalence; Cross-Cultural Equivalence; Systems Measure; Systemic Interventions; Systemic Therapy; Systemic Family Constellations Fam Proc x:1–17, 2016 Given the predominating individualistic focus in counseling and psychotherapy research (for an overview see Lambert, 2013), it is still rare to find measures that enable the identification of the effects of interventions on the individual level, as they relate to an important social system. The Experience in Social Systems Questionnaire (EXIS) aims to measure a client’s experience of how well he or she functions within an important social system. It allows each client to decide who the principal members of the social system are, irrespective of whether the focus of that social system is personal (EXIS.pers) or organizational (EXIS.org). WHAT DO WE MEAN BY “SOCIAL SYSTEMS” AND “SOCIAL SYSTEMS MEASURES”? Since family systems terminology is not quite identical between the United States and Europe, we start by clarifying our central terms (see von Schlippe & Schweitzer, 2012). We understand social systems as a collectivity of individuals who are bound together as a unit by multiple interactions (Hall & Fagen, 1956) and differentiated from their environment by a boundary of meaning (Willke, 1993). Verbal and nonverbal communication connects the members of a social system (Watzlawick, Weakland, & Jackson, 1967). Through communication, every member contributes to a social system’s functioning. Membership issues can a priori be defined in a normative way (e.g., both partners in a couple, all members of a nuclear family such as the father, mother, and children). Alternatively, they may also be defined as whomever an individual decides to call a member of this social system. In personal social systems, this may be for example an attachment figure who supported the child in its search for identity if the birth mother and/or father were not emotionally involved in the mother–father–child relationship, or an intimate friend after a student has left her parents’ home but is not yet in a romantic partnership, or a sibling subsystem that was separated from the nuclear family by domestic violence. In organizational social systems, this may not include all team members of, say, a surgical ward but only the chief physician, the intern, and the chief surgery nurse. When we reviewed social systems measures, we found some flexible family systems measures (Hamilton, Carr, Cahill, Cassells, & Hartnett, 2015; Pinsof et al., 2015), but mostly we found measures of systems functioning (see Sprenkle & Piercy, 2005) like the Family Assessment Device (Epstein, Baldwin, & Bishop, 1983; Mansfield, Keitner, & Dealy, 2014; Staccini, Tomba, Grandi, & Keitner, 2014) and the Family Adaptability and Cohesions Scale (Olsen, Protner, & Lavee, 1985). These measures assess the level of functioning in dyadic relationships (“you-me”; e.g., Alonso-Arbiol, Balluerka, Shaver, & Gillath, 2008) or whole families (“we-as-a-whole”; e.g., Green, Harris, Forte, & Robinson, 1991) and encourage judgments from a meta-perspective on the system, for example, “How would you value what is happening in the system that you are part of in terms of we?” Systemic measures that allow the clients to decide whom they consider to be the www.FamilyProcess.org 2 / FAMILY PROCESS principal members of their social system still are rare: Most focus on problems rather than solutions. The individuals’ positive experience within their social systems usually is not addressed, that is “How do you personally experience your personal situation from within your system in terms of you or I?” Evaluating systemic interventions in which only one member of a social system attends the intervention, such as in systemic individual therapy, systemic group therapy with strangers, Family Reconstruction Seminars (Nerin, 1991), and Family Constellation Seminars (Weber, Schmidt, & Simon, 2005), may require a different type of systemic outcome questionnaires. TO WHAT END DID WE DEVELOP THE EXIS? The particular goal of the EXIS is to be useful within process and outcome research on systemic interventions (1) while assessing an individual’s personal experience of his or her level of systemic functioning, especially (2) when the intervention does not necessarily target a couple or nuclear family but, say, a family of origin, a sibling subsystem, a group of close friends, or a residential treatment group, (3) when only the individual but not all members of the focal system attend the intervention, as in systemic individual therapy, systemic group therapy, Family Constellation Seminars, or Family Reconstruction Seminars, and (4) when the same questionnaire is used to measure not only interventions in intimate, personal social systems like couples and families but also the effects of organizational interventions like leadership coaching, team consultation, or large group interventions. The EXIS dimensions and items are formulated broadly in order to measure change in a variety of social systems, sometimes even in different social systems within the same study. We have based EXIS on systems theory concepts and general concepts of psychotherapy change, and we have attempted to make it sensitive enough to capture even small changes after brief interventions, so that it is suitable both for process research in repeated measurement designs and for outcome research after just a few sessions or, indeed, only one. Therefore, EXIS must be economical (short) and easy to use in largescale field research. Sparrer and Varga von Kibed (2008) and Weber (1993) formulated basic assumptions which stimulate an “appropriate order” in human systems, among them being the necessity to belong, the right for autonomy, and the ability to agree (accord). These ideas are in part familiar with Contextual Family Therapy (Boszormenyi-Nagy & Krasner, 1986), which tries to foster change by attending to the multigenerational transmission of family dynamics (Gangamma, Bartle-Haring, & Glebova, 2012). The focus is on relational ethics, a balanced give-and-take (fairness) influenced by the experience of trustworthiness, loyalty, entitlement, and indebtedness (Gangamma et al., 2012). Hargrave and Pfitzer (2003) define trust as the feeling of confidence in interactions (cited by Gangamma et al., 2012, p. 826). Loyalty is associated with a sense of accord with oneself and others. Entitlement refers to being cared for, including the right to belong to the family. Indebtedness is a kind of relational autonomy in which one can negotiate rights and responsibilities within the family (Boszormenyi-Nagy & Krasner, 1986; Gangamma et al., 2012). We suppose that interventions in which only one member of a system participates (e.g., Family Constellation Seminars, Family Reconstruction Seminars) will stimulate changes in the individual’s inner image of the focal social system and provide a new framework for feeling, thinking, and behaving (Becvar & Becvar, 2009). Experiences of trust, entitlement, appropriate loyalty, and appropriate indebtedness (Boszormenyi-Nagy & Krasner, 1986; Gangamma et al., 2012), or of confidence, belonging, accord, and autonomy (Sparrer & Varga von Kibed, 2008), can arise. This may help modify one’s own contribution to the interactional cycles within the social system, which, in turn, may contribute to whole-systems change. Depending on the nature of the conflict, constellations can be used to address Fam. Proc., Vol. x, xxxx, 2016 HUNGER ET AL. / 3 issues within families or organizations. They can integrate the symptom as a part of the system and show the system as important for the emergence and maintenance of the symptom (e.g.,Weber et al., 2005). A detailed description of Family Constellation Seminars can be found elsewhere (Hunger, Bornhauser, Link, Schweitzer, & Weinhold, 2014; € Hunger, Weinhold, Bornhauser, Link, & Schweitzer, 2015; Weinhold et al., 2013). € HOW DID WE DEVELOP THE EXIS? The EXIS is intended to measure concepts of importance in systems constellation theories, and these concepts should fit with corresponding concepts from clinical, personality, social, and organizational psychology. Two parallel versions exist, with identical items and different instructions: One describes the experience within intimate or personal systems (EXIS.pers) while the other describes the experience within task-oriented formal systems or organizations (EXIS.org). They are intended to be easy to use in large-scale field research. In this paper, we first present the theoretical background and development of the EXIS.pers and then an assessment of the psychometric properties of the German and English versions thereof. We used three subsequent strategies. (1) We intensively interviewed two prominent protagonists of the family constellation approach (Diana Drexler, Ph.D., and Gunthard Weber, M.D.) about the expected positive systemic outcomes of their work, reformulating their statements into potential dimensions and specific items. (2) Comparing their views with concepts from the broader family constellation literature (Schneider, 2009; Sparrer & Varga von Kibed, 2008; Weber et al., 2005) led us to three dimensions central to the theory of change in family constellations (Sparrer & Varga von Kibed, 2008): the experiences of belonging, autonomy, and accord. (3) Comparing these three dimensions with concepts on change processes and outcomes used in social, differential, clinical, and organizational psychology yielded two more important dimensions: confidence and clarity. Table 1 gives an overview of these five dimensions and their theoretical roots. Belonging describes the experience of being an acknowledged member of a social system (Sparrer & Varga von Kibed, 2008), of being respected and welcome (Boszormenyi-Nagy & Krasner, 1986), and of forming and maintaining significant interpersonal relationships (Baumeister & Leary, 1995; Gere & MacDonald, 2010). The experience of being cared for by other individuals and being socially integrated (Cyranowski et al., 2013) is an essential part of social support (Cohen, 2004; Woods, Priest, & Roush, 2014), in addition to instrumental, emotional, and informational support. Belonging is essential to protecting the boundaries of the family system (Sparrer & Varga von Kibed, 2008). The “principle of non-exclusion” (ibid., p. 181) describes the right of every member not to be excluded from the system as long as she fulfills the conditions for belonging; in families, this is usually based on biological descent. Autonomy describes standing up for one’s own needs and being assertive within the social system; in families, this means the gradual distinction between one’s family of origin and one’s own conjugal family, as well as the establishment thereof without having to abandon ties (ibid.). Autonomy implies the understanding that rights, responsibilities, appropriate indebtedness (Boszormenyi-Nagy & Krasner, 1986), closeness, and distance can be negotiated (Busby & Gardner, 2008; Christman, 2004; Stierlin, 1971); it is essential for the growth of the family system (Sparrer & Varga von Kibed, 2008). Accord means a basic acceptance of the critical and unpleasant aspects of one’s social system “as it is”, including the acceptance of positive and negative experiences (Sparrer & Varga von Kibed, 2008). It is associated with appropriate loyalties (Boszormenyi-Nagy & Krasner, 1986) and the experience of accepting emotions, cognitions, and behaviors “as they are” (Durm & Glaze, 2001; King & Wynne, 2004). In case reports from Family Constellation Seminars, clients describe an increased sense of belonging, www.FamilyProcess.org 4 / FAMILY PROCESS autonomy and accord after participation, and they attribute this change to the resolution of conflicts with members of their focal social system (Weber et al., 2005). Comparison with empirically supported concepts from clinical, differential, social, and organizational psychology show that belonging and autonomy are regarded as fundamental human motivations (Baumeister & Leary, 1995) and that accord is linked to the acceptance of one’s self and others (Durm & Glaze, 2001). Confidence describes the expectation that one will successfully cope with future challenges in one’s social systems (Sparrer & Varga von Kibed, 2008; Weber et al., 2005); it is associated with interpersonal trust (Boszormenyi-Nagy & Krasner, 1986) and with faith in the future (Krampen & Hank, 2004). Raising confidence is typically the first success experienced in the course of any psychotherapy (Howard, Lueger, Maling, & Martinovich, 1993). Clarity means the experience of clearly understanding relational configurations within one’s social system. Clarity arises when the contexts of an TABLE 1 Belonging, Autonomy, Accord, Confidence, and Clarity in Family Constellations Theory and in Clinical, Personality, and Social Psychology Family Constellation Theories Clinical, Personality, Social Psychology Belonging Background First basic assumption of family constellations (Sparrer & Varga von Kibed, 2008) Belonging (Baumeister & Leary, 1995; Cohen, 2004; Gere & MacDonald, 2010) Construct For example experience of being part of a personal system For example experience of forming and maintaining significant interpersonal relationships Autonomy Background Second basic assumption of family constellations (Sparrer & Varga von Kibed, 2008) Autonomy (Christman, 2004; Stierlin, 1971) Construct For example experience of standing up for one’s needs in one’s social system For example negotiation of rights and responsibilities or interpersonal closeness and distance (related autonomy/individuation) Accord Background Meta-assumption of family constellations (Sparrer & Varga von Kibed, 2008) Self-acceptance and acceptance of others (Durm & Glaze, 2001) Construct For example consent with how one experiences one’s social system “as it is,” including the acceptance of positive and negative experience For example acceptance of emotions, cognitions, and behaviors “as they are” Confidence Background No specific assumption of family constellations but a general dimension when treating clients in systemic constellations (Sparrer, 2006; Weber, Schmidt, & Simon, 2005) Trusting in the future (Krampen & Hank, 2004) Construct For example hope that unchangeable situations in one’s social system can be dealt with positively in the future For example faith that things will develop positively in the future Clarity Background No specific assumption of family constellations but a general dimension when treating clients in systemic constellations (Sparrer, 2006; Weber, Schmidt, & Simon, 2005) Emotional clarity (Salovey et al., 1995) Construct For example experience of understanding the configurations of one’s relationships in one’s social system For example clarity about one’s feelings and not being confused about how one feels Fam. Proc., Vol. x, xxxx, 2016 HUNGER ET AL. / 5 individual’s conflict become visible in their spatial representation (Sparrer & Varga von Kibed, 2008; Weber et al., 2005). Confidence is a construct similar to trusting in the future (Krampen & Hank, 2004), and clarity in relationships is related to “emotional clarity” (Salovey, Mayer, Goldman, Turvey, & Palfai, 1995). EMPIRICAL DEVELOPMENT OF THE EXIS.PERS Our questionnaire was developed through three sequential studies over 5 years. In Study 1 (2010–2011), the German version was pilot-tested on a sample of 179 participants from the German adult population. Exploratory factor analysis (EFA) was used to identify latent variables (Floyd & Widaman, 1995). In Study 2 (2011–2013), 634 German adults in Heidelberg and Landau completed our questionnaire. In Study 3 (2013–2014), 310 English adults in London filled it in. In Studies 2 and 3, we conducted confirmatory factor analysis (CFA), assessing internal consistencies, external correlations (e.g., with age, gender, employment), and the structural equivalence of the German and English EXIS.pers. All studies were approved by the Ethics Committee of the Heidelberg University. STUDY 1: PILOT STUDY Study 1 pilot-tested the German version of our questionnaire to identify latent variables based on a sample of 179 participants from the adult population. Method Participants Recruiting occurred in a snowballing manner: We distributed questionnaires to colleagues at our own Institute of Medical Psychology, who gave them out to their family, friends, and colleagues, who, in turn, further distributed them throughout their networks. In the end, 223 adults responded: 23 participants were excluded because they were under 18 or over 63 years of age, while another 21 questionnaires were disqualified due to missing values (which were found to be completely at random according to Little’s MCAR test: v²(44) = 48.35, p = .302; Little & Rubin, 2002). This left a sample of 179 persons for the EXIS.pers: 69% were women (M = 35 years, SD = 11), 42% lived with a partner, 55% had a college degree, and 93% were employed. Item generation and measures As described above, 10 items for each dimension (belonging, autonomy, accord, confidence, clarity) were created and examined for their appropriateness to capture “experience in personal social systems.” All items were made compatible with colloquial language use. The item sequence was randomized. Items were to be rated from 1 (“not at all”) to 6 (“entirely”). Procedures The pilot measure was handed out as a stand-alone questionnaire. The instructions included the following statements: Please consider the following statements about the experiences people have in personal social systems (e.g., with parents, partners, children, friends). Please especially consider your relationships with people who are important to you. Please refer to the past two weeks when answering the questions. For every statement, mark the answer that comes closest to your experience. If in doubt, follow your first impulse. Please take a moment at the end to make sure you have ticked a box in every line. www.FamilyProcess.org 6 / FAMILY PROCESS The measure also included the following instruction: Concerning my important personal relationships, over the course of the past two weeks I experienced...(followed by the items). Results In EFA, we used principle axis factoring because distribution was not normal for all variables (Shapiro–Wilks Test, p < .01; Finch & West, 1997), and because it is effective for latent variable identification (Floyd & Widaman, 1995). We used direct promax rotation to extract the minimum number of factors while allowing for correlated factors (Field, 2009), which we expected to be correlated as they assessed the fundamental cognitive, emotional, and behavioral motivations of the social system members. The means were higher than the theoretical mean (Memp 4.4, SD 1.00; Mth = 3.5; Table 2). The Kaiser–Meyer–Olkin measure of sampling adequacy was good, at 0.89. To determine the number of relevant factors we used parallel analysis (Horn, 1965; O’Connor, 2000), which concentrates on the extraction and comparison of eigenvalues from random data that parallel the actual data with respect to the number of cases and variables in the focal sample. Serving as a baseline are the eigenvalues corresponding to the 95th percentile of the distribution of the random data. The number of factors to retain is derived by determining the number of eigenvalues in the actual data that are larger than the corresponding 95th-percentile (and mean) random data eigenvalues. Parallel analysis indicated that a four-factor model provided the best fit, accounting for 47.6% of the variance. The first five observed eigenvalues were 15.39, 3.21, 2.90, 2.31, and 1.63. The synthetic eigenvalues (95th percentile) were 2.19, 2.05, 1.99, 1.90, and 1.83. Factor loadings and item-total correlations are shown in Table 2. The four dimensions were (1) the experience of being respected and welcome, (2) the experience of standing up for one’s own needs and being assertive, (3) the experience of interpersonal trust and trust in the future, and (4) the acceptance of positive and negative experiences. Items measuring the experience of clarity within the personal social system did not turn out to be a distinct factor. These five dimensions paralleled the five concepts from the literature section above. Final item selection relied on both statistical and theoretical considerations. To develop a short and economical questionnaire, we followed Buhner’s (2006) conservative procedure € for item selection. First, items had to have high factor loadings, k ≥ .60. We found 19 items that met this criterion. Second, one factor should consist of three items, the absolute minimum requirement for factor composition (ibid.). As for clarity, only two items showed factor loadings of k ≥ .60, so they did not emerge as a distinct factor. Of the remaining 17 items, four authors characterized the three most appropriate content-related items of each factor. The Belonging, Accord, Autonomy, and Confidence items can be found in Table 2. All (sub)scales had good internal consistency with Cronbach’s alphas, ranging from .79 to .83 (Bernstein & Nunnally, 1994). Discussion Our statistical calculations supported four dimensions: Belonging, Autonomy, Accord, and Confidence. The explained variance was on the lower end (47.6%), possibly due to the uni-level structure of EFA. In contrast to bi-level models, otherwise unexplained variance cannot be captured by a separate general factor. In Study 2, we therefore integrated the testing of a bi-level model using CFA. From a methodological point of view, clarity did not appear as a distinct factor. This may be because our item development and wording did not allow for the production of a clarity factor and may have created a blending of other constructs like meaning, for example, “that things became meaningful,” and empathy, for Fam. Proc., Vol. x, xxxx, 2016 HUNGER ET AL. / 7 example, “that I was able to empathize with how the others were feeling.” We do not reject our original five-factor theory but continue to believe that clarity is an important dimension, so future work should include formulating more appropriate clarity items. Participants made clear distinctions between their experiences of belonging, autonomy, accord, and confidence; these four factors demonstrated good internal consistency. The EXIS.pers means and standard deviations presented in this study do not represent normative data. We will come back to this point in the general discussion. STUDY 2 AND 3: PSYCHOMETRIC PROPERTIES OF THE GERMAN AND ENGLISH EXIS.PERS To evaluate the factor structure and internal consistencies found in our first exploratory study, Study 2 and 3 were conducted. We also used external correlations for the assessment of the psychometric quality, and we tested the structural equivalence of the German and English EXIS.pers. Method Participants Study 2 considered 663 German adults from two subsamples: Subsample n1 consisted of 455 participants from Heidelberg and Landau, and subsample n2 consisted of 208 from an TABLE 2 Study 1 Descriptive Statistics, Factor-Item Loadings, and Item-Total Correction M SD Factor 1 Factor 2 Factor 3 Factor 4 rit Belonging ... feeling a sense of belonging 4.83 1.02 .72 .11 .32 .33 .72 ... feeling that I was given attention 4.74 1.03 .69 .26 .19 .26 .68 ... feeling that my presence was welcome 4.89 0.97 .59 .06 .10 .29 .60 Accord ... feeling satisfied 4.45 1.07 .11 .75 .29 .28 .75 ... being in accord with the way things are 4.04 1.10 .12 .74 .19 .29 .72 ... being at peace with others 4.31 0.97 .06 .63 .16 .18 .60 Autonomy ... being able to honor my needs 4.13 1.14 .06 .23 .72 .08 .61 ... being able to maintain adequate boundaries between myself and others 4.14 1.09 .38 .30 .64 .12 .71 ... being able to decide when to offer my contributions 4.40 0.93 .34 .33 .64 .04 .70 Confidence ... being hopeful that things will continue positively 4.54 0.97 .39 .23 .05 .71 .68 ... being optimistic that I am strong enough to cope with upcoming challenges 4.32 1.15 .33 .09 .15 .63 .66 ... being confident that I am able to come to terms with things I cannot change 4.13 1.05 .31 .27 .17 .60 .58 Note. Items per factor are indicated by bold values. www.FamilyProcess.org 8 / FAMILY PROCESS RCT investigating the efficacy of Family Constellation Seminars (Weinhold et al., 2013). Of these 663 questionnaires, 29 were excluded for having at least one missing value, leaving a sample of 634 participants: 66% were women (M = 43 years, SD = 12, range: 18–65), 73% lived with a partner, 82% had a college degree, 98% were employed (6% students), 30% were currently or have previously been in psychotherapy, and 43% had experience with systems constellations. All participants were Europeans, 95% of them Germans. In Study 3, 330 adults were recruited from the 600-person multi-ethnic and multinational staff of KIDS Company, a charity in London, England, that provides support to vulnerable inner-city children www.kidsco.org.uk. We chose KIDS Company as our place of recruitment because it works with families, it has a large staff, and it was receptive to our endeavor due to its shared interest in family-, adolescent-, and child-focused work. Of the 330 questionnaires received, 20 were excluded for having at least one missing value, reducing the sample to 310 participants: 76% were women (M = 36 years, SD = 11, range: 20–70), 49% lived with a partner, 93% had a college degree, 100% were employed at KIDS Company, 42% had experience with systems constellations, and 69% were British. Translation We translated our questionnaire in accordance with the standards formulated by Schmitt and Eid (2007). Three bilingual colleagues from the Institute of Medical Psychology independently translated items and instructions. Three authors (C. H., A. B., J. S.) checked this first draft for theoretical consistency and discussed variations with the bilinguals. This second draft was re-translated by three other bilinguals. At this point, no further revisions proved necessary. We sent this English version to four British acquaintances of the first author (C. H.). They were neither psychologists nor experienced in family constellations. These native speakers filled in the questionnaire and discussed the comprehensibility of item formulations and instructions with the second author (A. B.). Again, no further revisions proved necessary. Procedure In the German Study 2, to recruit the participants for the subsample n1, we distributed 564 questionnaires (response rate: 81%; n1 = 455) in undergraduate psychology classes at the Universities of Heidelberg and Landau, asking students to distribute the questionnaire booklets to friends and family. After having filled in the EXIS.pers, 387 participants also indicated whom they had had in mind while answering our questionnaire, that is their relationship with parents, children, a partner, or friends. The undergraduates were rewarded with course credits and participation in a raffle for an iPod. Descriptions and instructions for Study 2 were identical to Study 1. The recruitment of the 208 participants of the RCT subsample n2 is described elsewhere (Weinhold et al., 2013). There, 154 questionnaire booklets were returned (response rate: 74%, as recruitment for the RCT was still underway). We excluded one questionnaire for more than 20% missing data, which left a sample of 153 participants. In Study 3, we distributed 344 questionnaire booklets to KIDS Company staff members during a staff meeting, explicitly stating that participation was completely optional. After having filled in the EXIS.pers, all participants indicated whom they had had in mind while answering our questionnaire, that is parents, children, a partner, or friends. The response rate was 96%. Participants were rewarded by the chance to win a raffle for an iPod. Descriptions and instructions for Study 3 were identical to Studies 1 and 2. Measures In both Studies 2 and 3, the demographic data were assessed as in Study 1. Fam. Proc., Vol. x, xxxx, 2016 HUNGER ET AL. / 9 Analysis plan With respect to the 208 RCT participants of subsample n2 in Study 2, we conducted a two-level (participant, family constellation intervention) linear regression analysis to account for any positive intraclass correlations (ICC) indicating variance attributable to the family constellation intervention above and beyond the variance attributable to its participants (Kreft & de Leeuw, 1998; Murray, Varnell, & Blitstein, 2004). The ICC coefficient for the EXIS.pers total score was zero, indicating no ICC among participants of the same seminar and thus indicating no data dependency (Hunger et al., 2014; Weinhold et al., 2013). With respect to n1 in Study 2, we distributed one questionnaire to each student but did not control for data dependency. Although there may be siblings in our sample, since each student was only awarded course credits for one returned questionnaire booklet, we assumed no data dependency. In Study 3, we distributed one questionnaire booklet to each participating staff member, who completed it on the spot at the staff meeting. Again, we assumed no data dependency. Using CFA (n = 634 in the German Study 2, n = 310 in the English Study 3), the hypothesized four-factor model from Study 1 was tested against a number of factor models. Of these, the one-factor model and a five-factor bi-level model, including a general factor and four specific factors, will be reported. The last two models were indicated by the strong decrease in eigenvalues in Study 1.1 We used a bi-level model because we supposed each item of the EXIS.pers to be indicative of two factors at different logical levels: the general factor and a specific factor. In contrast to a second-order model in which the general factor is simply indicative of the sum of the four underlying factors, we propose that each EXIS.pers item contains both a unique variance in one of the four content-specific factors and a common variance as represented by the general factor. Otherwise unexplained variances can then be explained by the separate general factor. Factor reliabilities and item discriminability were also calculated. External associations were tested (Pearson product-moment correlation coefficient; chi-squared test) using age, gender, civil status, education, employment, nationality, years lived in Germany/England, and previous experiences with systems constellations. In the German sample, we also asked about current and previous outpatient psychotherapy. Structural equivalence measured the item-intercorrelations and factor structures of the German and English data. Invariance of factor loadings across cultural groups was required for any between-group comparison (Welkenhuysen-Gybels & van de Vijver, 2007). We used congruence indices to test structural equivalence (van de Vijver & Leung, 1997) with the most commonly used proportional coefficient, Tucker’s Φ (Tucker, 1951), which is not affected by multiplications of the factor loadings (Fischer & Fontaine, 2011). Φ < 0.90 indicates non-negligible incongruities, while Φ > 0.95 demonstrates factorial invariance (van de Vijver & Leung, 1997). Results Confirmatory factor analysis First, using SPSS.AMOS 19 (Byrne, 2010), a CFA with maximum-likelihood estimation was conducted to test the one-factor model (Model 1). Second, a four-factor solution (Model 2) was tested, which was developed from theoretical considerations on Belonging, Autonomy, Accord, and Confidence. Third, the four-factor model was extended to a five-factor 1 We also tested the two- and three-factor models, and the five-factor hierarchical model. All models showed a poor fit to the observed data; that is, they fit better than the one-factor solution but not as well as the five-factor bi-level solutions. www.FamilyProcess.org 10 / FAMILY PROCESS bi-level model that included a general factor and four specific factors (Model 3). All three models showed significance (p < .001), which was not surprising and of minor importance because of our large samples (Marsh, Wen, & Hau, 2004). The following fit-indices and cut-off values were used to assess model fit (Hu & Bentler, 1999): comparative fit index (CFI; ≥0.95), normed fit index (NFI; ≥0.90), root-mean-square error of approximation (RMSEA; ≤0.06). Akaike’s information criterion (AIC; Akaike, 1974) was used to detect the best model as indicated by the lowest AIC. The German Model 1 showed a poor fit to the observed data: v²(54) = 916.69, p < .001; v²/df = 16.98; CFI = 0.79; NFI = 0.78, RMSEA = 0.16, 90% confidence interval (CI) [0.15, 0.17]; AIC = 964.69. Model 2 showed a good fit: v²(48) = 176.21, p < .001; v²/df = 3.67; CFI = 0.97; NFI = 0.96, RMSEA = 0.07, 90% CI [0.05, 0.08]; AIC = 236.22. Model 3 also showed a good fit: v²(38) = 135.23, p < .001; v²/df = 3.56; CFI = 0.98; NFI = 0.97, RMSEA = 0.06, 90% CI [0.05, 0.08]; AIC = 208.28 (Figure 1). Similar to the German study, the English Model 1 showed a poor fit: v²(54) = 446.56, p < .001; v²/df = 8.27; CFI = 0.83; NFI = 0.81, RMSEA = 0.15, 90% confidence interval (CI) [0.14, 0.17]; AIC = 494.56. Model 2 showed a better fit: v²(54) = 124.17, p < .001; v²/ df = 2.48; CFI = 0.97; NFI = 0.95, RMSEA = 0.07, 90% CI [0.05, 0.09]; AIC = 180.17. Model 3 also showed a good fit: v²(38) = 90.78, p < .001; v²/df = 2.39; CFI = 0.98; NFI = 0.96, RMSEA = 0.07, 90% CI [0.05, 0.09]; AIC = 170.78 (Figure 1). Descriptive statistics, reliability In the German study, half of the participants (53%) who indicated whom they were thinking of when answering our questionnaire named a combination of partners and children (21%), a combination of partners and friends (11%), just a partner (11%), or a combination of parents, children, and other social relationships (9%; e.g., classmates, housemates, club members). In the English study, half of the participants (52%) named a combination of parents, partners, and friends (16%), a partner (10%), a combination of partners and friends (9%), a combination of parents and friends (9%) or just friends (8%). Although the questionnaire explicitly asked for personal social systems, 5% of the German participants and 11% of the English participants also referred to their work colleagues. Mean scores for each subscale in the German study ranged from 4.24 to 4.66 and in the English version from 4.45 to 4.79. In both studies, Cronbach’s alphas were excellent to good, except for the Autonomy subscale, which did show satisfactory internal consistency (Table 3). For each of the 12 items, skewness ranged from 0.24 to 0.68 in the German study and from 0.50 to 1.08 in the English study. Kurtosis ranged from 0.36 to 0.31 in the Germany study and from 0.28 to 1.53 in the English study. Validity External associations are shown in Table 4. In the German sample, older people significantly seemed to have less positive experiences in personal social systems, especially in terms of belonging and accord. In contrast, in the English sample, advanced age was significantly associated with more confidence. However, in both countries, being male significantly correlated with less positive experiences in confidence. Employment emerged as producing better experiences in personal social systems; although for the total score and all subscales, only in Germany did these correlations reveal significance. In the German sample, current and previous psychotherapy was associated with better experiences in personal social systems, in both the total score and the experiences of belonging, accord, and confidence—but not for autonomy. Neither in Germany nor in England did civil status, education, nationality, years lived in one of these countries, or previous experience with systems constellations reveal significance. Fam. Proc., Vol. x, xxxx, 2016 HUNGER ET AL. / 11 German EXIS.pers English EXIS.pers FIGURE 1. Study 2 and 3 CFA Model 3: five-factor bi-level-model with Experience in Personal Social Systems as General Factor (EXIS.pers, n = 634 German adults, n = 310 English adults) www.FamilyProcess.org 12 / FAMILY PROCESS Structural equivalence The structural equivalence between the German and English EXIS.pers was demonstrated well using Tucker’s Φ = 0.996 for the total score, Φ = 0.998 for the Belonging subscale, Φ = 0.959 for the Autonomy subscale, Φ = 0.965 for the Accord subscale, and Φ = 0.998 for the Confidence subscale. Discussion In both the German and English psychometric studies, we provided significant information about the factor structure of the EXIS.pers, the membership composition of the social systems in focus when answering the EXIS.pers, internal consistencies, and structural equivalence. In both the German Study 2 and the English Study 3, the second and third model showed good fit. We decided to use the five-factor bi-level model, including a general factor measuring one’s Experience In Personal Social Systems (EXIS.pers) and four dimensions: Belonging, Autonomy, Accord, and Confidence (Figure 1). As described in the methods section, we tested the bi-level model, as we supposed each item of the EXIS.pers to be indicative of two factors at different logical levels: the general factor and the more specific TABLE 3 Study 2 and 3 Descriptive Statistics of the German and English EXIS.pers Domain/Subscale German Sample English Sample M SD SE a M SD SE a EXIS.pers Belonging 4.66 0.96 .04 .85 4.79 0.95 .05 .88 Autonomy 4.35 0.86 .04 .74 4.57 0.84 .05 .79 Accord 4.24 0.96 .04 .87 4.45 0.94 .05 .84 Confidence 4.27 0.90 .04 .83 4.53 0.97 .06 .85 Total 4.38 0.75 .03 .91 4.59 0.78 .04 .92 Note. EXIS.pers = Experience in Personal Social Systems. TABLE 4 Study 2 and 3 External Correlations of the German and English EXIS.pers Total Score and Subscales with External Variables Scale Belonging Autonomy Accord Confidence EXIS.persTS German sample Age .14** .06 .10* .10 .10* Gender (female) .03 .04 .05 .14** .08 Employment .15** .09* .12** .10* .14** Outpatient psychotherapy Current .11* .10 .12* .11* .13* Previous .16** .05 .12* .25** .18** English sample Age .03 .01 .02 .13** .04 Gender (female) .01 .04 .03 .01** .05 Employment .03 .02 .05 .01 .01 Note. EXIS.pers = Experience in Personal Social Systems. The following demographic variables did not reveal significance neither in the German Study 2 nor in the English Study 3: civil status (living with a partner), education (having a college degree), nationality, years lived in Germany or England, respectively, previous experience with systems constellations. Significant correlations are indicated by bold values. **p ≤ .01. *p ≤ .05. Fam. Proc., Vol. x, xxxx, 2016 HUNGER ET AL. / 13 factors. We supposed that each EXIS.pers item contains both a unique variance in one of the four content-specific factors as well as a common variance as represented by the general factor. The bi-level model did a good job of depicting these characteristics of the EXIS.pers. Mean values can be calculated for the general factor (12 items) and for each of the subscales (three items each; Table 2). The subscales are very reliable, except for Autonomy, which showed satisfactory internal consistency. Autonomy may be the broadest of the four psychological constructs, and we know from other measures that complex constructs are less reliable. The EXIS.pers means and standard deviations presented in these studies do not represent normative data. We will come back to this point in the general discussion. Interestingly, when we asked the participants whom they were thinking of while answering our questionnaire, they chose diverse combinations of relationships from their focal social systems. This may broaden our understanding of the composition of social systems membership. A person’s ideas about whom they count as members of their important social system may not always follow predetermined normative definitions like “my partner and I” or “parents and their children.” GENERAL DISCUSSION Our three studies describe the development of the Experience In Personal Social Systems (EXIS.pers), which is based on concepts stemming from the family constellation approach. It is probable that the EXIS.pers can also be applied to many other systemsoriented interventions. It measures clients’ positive experience of their level of systemic functioning on four dimensions—Belonging, Autonomy, Accord, and Confidence—that we have found compatible with concepts from clinical psychology and psychotherapy. Psychometric investigations in German and English adult populations provide first evidence of the EXIS.pers being an economical and reliable measure in both countries. Implications for Research, Clinical Use, Counseling We expect the EXIS.pers to be most useful for researchers who are looking for (1) a short and easy-to-administer questionnaire, which is (2) applicable especially when an individual in therapy knows whom he considers important members of his social personal (EXIS.pers) or organizational (EXIS.org) social system, which (3) captures the individual’s experience of his level of systemic functioning with respect to his experiences of belonging, autonomy, accord, and confidence, and which (4) can be used to measure systemic effects although only one member of the social system participates in the intervention. The economical design makes the EXIS.pers promising for practice-oriented research (Castonguay, Barkham, Lutz, & McAleavey, 2013) as well as for studies under time constraints. Strengths, Limitations, Future Directions The EXIS.pers’ strengths are (1) its brevity, (2) that it has been evaluated in a large German and English adult population, (3) that both the German and English versions of EXIS.pers provide psychometrical soundness grounded in conservative analyses, (4) that it has been tested on employed adults, not just students, and (5) that its psychometric property in clinical samples is currently being tested on psychotherapy patients, patients with eating disorders and patients with social anxiety disorders, and with the relatives of psychotherapeutic patients. Finally, (6) the construct validity and sensitivity to change of the EXIS.pers is currently being tested. www.FamilyProcess.org 14 / FAMILY PROCESS EXIS.pers also has limitations. (1) This is the first time we have tested it, which makes more psychometric and validity research necessary, especially with regard to the experience of autonomy and accord subscales (e.g.,Barbaranelli, Caprara, & Maslach, 1997). (2) We only considered data dependency with respect to subsample n2 in Study 2. Future studies therefore should gather data on dependencies and consider them in the analysis. (3) As the EXIS.pers is tested on individuals, we cannot provide EXIS.pers data on an entire personal social system (e.g., nuclear family, group of intimate friends). It would be worthwhile to broaden the focus to include group data in future studies: We currently collect data of patients plus one significant reference person of their choosing. (4) In all three of our studies, we have an oversampling of women. Future studies should concentrate more on men. (5) The psychometric property of the EXIS.pers in clinical samples has not yet been published. As Family Constellation Seminars are commonly conducted outside the formal healthcare system—and EXIS.pers was originally created to measure change after Family Constellation Seminars—our initial recruitment did not require the participants to have been diagnosed clinically. (6) The EXIS.pers means and standard deviations presented in these studies do not represent normative data. There are as of yet no norms for a general population or for clinical groups. Future research should investigate the EXIS.pers (1) in more varied and clinical groups and in various types of psychotherapy and counseling to develop benchmark data, as well as (2) test its psychometric properties for children and adolescents. Copies of the German and English versions of EXIS.pers are available from the first author. REFERENCES Akaike, H. (1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control, 19 (6), 716–723. doi:10.1109/TAC.1974.1100705. Alonso-Arbiol, I., Balluerka, N., Shaver, P. R., & Gillath, O. (2008). Psychometric properties of the Spanish and American versions of the ECR Adult Attachment Questionnaire: A comparative study. European Journal of Psychological Assessment, 24(1), 9–13. doi:org/10.1027/1192-5604/a000046. Barbaranelli, C., Caprara, G. V., & Maslach, C. (1997). Individuation and the five factor model of personality traits. European Journal of Psychological Assessment, 13(2), 75–84. doi:10.1027/1015-5759.13.2.75. Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497–529. doi:10.1037/0033-2909.117.3.497. Becvar, D. S., & Becvar, R. J. (2009). Family therapy: A systemic integration. Boston, MA: Pearson. Bernstein, I. H., & Nunnally, J. C. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill. Boszormenyi-Nagy, I., & Krasner, B. R. (1986). Between give and take: A clinical guide to contextual therapy. New York: Routledge. Bu¨hner, M. (2006). Einfuhrung in die Test- und Fragebogenkonstruktion € . Munchen: Pearson. € Busby, D. M., & Gardner, B. C. (2008). How do I analyze thee? Let me count the ways: Considering empathy in couple relationships using self and partner ratings. Family Process, 47(2), 229–242. doi:10.1111/j.1545- 5300.2008.00250.x. Byrne, B. M. (2010). Structural equation modeling with AMOS: Basic concepts, applications, and programming (2nd ed.). New York: Routledge. Castonguay, L., Barkham, M., Lutz, W., & McAleavey, A. (2013). Practice-oriented research: Approaches and applications. In M. J. Lambert (Ed.), Bergin and Garfield’s handbook of psychotherapy and behavior change (pp. 85–133). Hoboken, NJ: Wiley. Christman, J. (2004). Relational autonomy, liberal individualism, and the social constitution of selves. Philosophical Studies, 117, 143–164. doi:10.1023/B:PHIL.0000014532.56866.5c. Cohen, S. (2004). Social relationships and health. American Psychologist, 59, 676–684. doi:10.1037/0003- 066X.59.8.676. Cyranowski, J. M., Zill, N., Bode, R., Butt, Z., Kelly, M. A., & Pilkonis, P. A. (2013). Assessing social support, companionship, and distress: National Institute of Health (NIH) Toolbox Adult Social Relationship Scales. Health Psychology, 32(2), 293–301. doi:10.1037/a0028586. Durm, M. W., & Glaze, P. E. (2001). Relation of self-acceptance and acceptance of others. Psychological Reports, 88(2), 410. doi:10.2466/PR0.89.6.386-386. Fam. Proc., Vol. x, xxxx, 2016 HUNGER ET AL. / 15 Epstein, N. B., Baldwin, L. M., & Bishop, D. S. (1983). The McMaster family assessment device. Journal of Marital and Family Therapy, 9(2), 171–180. doi:10.1111/j.1752-0606.1983.tb01497.x. Field, A. P. (2009). Discovering statistics using SPSS (3rd ed.). Los Angeles, CA: Sage. Finch, J. F., & West, S. G. (1997). The Investigation of Personality Structure: Statistical Models. Journal of Research in Personality, 31, 439–485. Fischer, R., & Fontaine, J. R. J. (2011). Methods for investigating structural equivalence. In D. Matsumoto & F. J. R. van de Vijver (Eds.), Cross-cultural research methods in psychology (pp. 179–215). Cambridge: Cambridge University Press. Floyd, F. J., & Widaman, K. F. (1995). Factor analysis in the development and refinement of clinical assessment instruments. Psychological Assessment, 7(3), 286–299. doi:10.1037/1040-3590.7.3.286. Gangamma, R., Bartle-Haring, S., & Glebova, T. (2012). A study of contextual therapy theory’s relational ethics in couples in therapy. Family Relations, 61(5), 825–835. doi:10.1111/j.1741-3729.2012.00732.x. Gere, J., & MacDonald, G. (2010). An update of the empirical case for the need to belong. Journal of Individual Psychology, 66(1), 93–115. Green, R. G., Harris, R. N., Forte, J. A., & Robinson, M. (1991). Evaluating FACES—III and the Circumplex Model: 2,440 families. Family Process, 30(1), 55–73. doi:10.1111/j.1545-5300.1991.00055.x. Hall, A. D., & Fagen, R. E. (1956). Definition of system. General Systems, 1(1), 18–28. Hamilton, E., Carr, A., Cahill, P., Cassells, C., & Hartnett, D. (2015). Psychometric properties and responsiveness to change of 15- and 28-item versions of the SCORE: A Family Assessment Questionnaire. Family Process, 54, 454–463. doi:10.1111/famp.12117. Hargrave, T. D., & Pfitzer, F. (2003). The new contextual therapy: Guiding the power of give and take. New York: Brunner-Routledge. Horn, J. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30(2), 179–185. Howard, K. I., Lueger, R. J., Maling, M. S., & Martinovich, Z. (1993). A phase model of psychotherapy outcome: Causal mediation of change. Journal of Consulting and Clinical Psychology, 61(4), 678–685. doi:10.1037/0022- 006X.61.4.678. Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. doi:10.1080/10705519909540118. Hunger, C., Bornhauser, A., Link, L., Schweitzer, J., & Weinhold, J. (2014). Improving experience in personal € social systems through family constellation seminars: Results of a randomized controlled trial. Family Process, 53(2), 288–306. doi:10.1111/famp.12051. Hunger, C., Weinhold, J., Bornhauser, A., Link, L., & Schweitzer, J. (2015). Mid- and long-term effects of family € constellation seminars in a general population sample: 8- and 12-month follow-up. Family Process, 54(2), 344– 358. doi:10.1111/famp.12102. King, D. A., & Wynne, L. C. (2004). The emergence of “family integrity” in later life. Family Process, 43(1), 7–21. doi:10.1111/j.1545-5300.2004.04301003.x. Krampen, G., & Hank, P. (2004). The confidence trias. Report Psychologie, 29, 666–777. Kreft, I., & de Leeuw, J. (1998). Introducing multilevel modeling. Thousand Oaks, CA: Sage. Lambert, M. J. (Ed.) (2013). Bergin and Garfield’s handbook of psychotherapy and behavior change (6th ed.). New York: Wiley. Little, R. J., & Rubin, D. B. (2002). Statistical analysis with missing data. Hoboken, NJ: Wiley. Mansfield, A. K., Keitner, G. I., & Dealy, J. (2014). The family assessment device: An update. Family Process, 54, 82–93. doi:10.1111/famp.12080. Marsh, H. W., Wen, Z., & Hau, K.-T. (2004). Structural equation models of latent interactions: Evaluation of alternative estimation strategies and indicator construction. Psychological Methods, 9(3), 275. doi:10.1037/ 1082-989X.9.3.275. Murray, D. M., Varnell, S. P., & Blitstein, J. L. (2004). Design and analysis of group-randomized trials: A review of recent methodological developments. American Journal of Public Health, 94(3), 423–432. doi:10.2105/ AJPH.94.3.423. Nerin, W. (1991). Family reconstruction: The masterpiece of Virginia Satir. Journal of Couples Therapy, 2(1–2), 103–118. doi:10.1300/J036v02n01_08. O’Connor, B. P. (2000). SPSS and SAS programs for determining the number of components using parallel analysis and Velicer’s MAP test. Behavior Research Methods, Instruments & Computers, 32(3), 396–402. doi:10.3758/BF03200807. Olsen, D. H., Protner, J., & Lavee, Y. (1985). FACES III—Family Adaptability and Cohesion Evaluation Scales. St. Paul, MN: University of Minnesota. Pinsof, W., Zinbarg, R., Shimokawa, K., Latta, T., Goldsmith, J., Knobloch-Fedders, L. et al. (2015). Confirming, validating and norming the factor structure of STIC initial and intersession. Family Process, 54, 464–484. doi:10.1111/famp.12159. www.FamilyProcess.org 16 / FAMILY PROCESS Salovey, P., Mayer, J. D., Goldman, S. L., Turvey, C., & Palfai, T. P. (1995). Emotional attention, clarity, and repair: Exploring emotional intelligence using the Trait Meta-Mood Scale. In J. W. Pennebaker (Ed.), Emotion, disclosure, and health (pp. 125–154). Washington, DC: American Psychological Association. Schmitt, M., & Eid, M. (2007). Richtlinien fur die € Ubersetzung fremdsprachlicher Messinstrumente. € Diagnostica, 53, 1–2. doi:10.1026/0012-1924.53.1.1. Schneider, J. R. (2009). Family constellations. Basic principles and procedures. Heidelberg: Carl-Auer. Sparrer, I. (2006). Systemic structural constellations. Theory and practice. Heidelberg: Carl-Auer. Sparrer, I., & Varga von Kibed, M. (2008). The systemic principles as a basis for a Systemic Structural Constellation. In R. Daimler (Ed.), Basics of systemic structural constellations. A manual for beginners and advanced learners (pp. 39–62). Munchen: K € €osel-Verlag. Sprenkle, D. H., & Piercy, F. P. (2005). Research methods in family therapy. New York: Guilford. Staccini, L., Tomba, E., Grandi, S., & Keitner, G. (2014). The evaluation of family functioning by the family assessment device: A systematic review of studies in adult clinical populations. Family Process, 54, 94–115. doi:10.1111/famp.12098. Stierlin, H. (1971). The conduct of one is the conduct of the other [Das Tun des Einen ist das Tun des Anderen]. Berlin: Suhrkamp Verlag. Tucker, L. R. (1951). A method for synthesis of factor analysis studies (personnel research section report no. 984). Washington, DC: Department of the Army. van de Vijver, F. J. R., & Leung, K. (1997). Methods and data analysis for cross-cultural research (Vol. 1). New York: Sage. von Schlippe, A., & Schweitzer, J. (2012). Textbook of systemic therapy and counseling (Vol. 1). G€ottingen: Vandenhoek & Ruprecht. Watzlawick, P., Weakland, J. A., & Jackson, D. L. (1967). Pragmatics of human communication. New York: W.W. Norton. Weber, G. (Ed.) (1993). Two kinds of luck: The systemic psychotherapy of Bert Hellinger. Heidelberg: Carl-Auer. Weber, G., Schmidt, G., & Simon, F. B. (2005). Constellation work revisted...according to Hellinger? With a metacomment by Matthias Varga von Kibed. Heidelberg: Carl-Auer. Weinhold, J.*., Hunger, C.*., Link, L., Rochon, J., Wild, B., Bornhauser, A. et al. (2013). Family constellation € seminars increase psychological functioning in a general population sample: Results of a randomized controlled trial. Journal of Counseling Psychology, 60(4), 601–609. doi:10.1037/a0033539 *shared first authorship Welkenhuysen-Gybels, J., & van de Vijver, F. J. R. (2007). Methods for the evaluation of construct equivalence in studies involving many groups. Retrieved May 15, 2015, from http://soc.kuleuven.be/web/files/6/35/webversion.pdf Willke, H. (1993). Systemtheorie: Eine Einfuhrung in die Grundprobleme der Theorie sozialer Systeme € . Stuttgart: UTB. Woods, S. B., Priest, J. B., & Roush, T. (2014). The biobehavioral family model: Testing social support as an additional exogenous variable. Family Process, 53, 672–685. doi:10.1111/famp.12086