Development and validation of a scale to measure subjective liminality: individual differences in the perception of in-betweenness

Abstract

Purpose

Modern careers are marked by periods of feeling betwixt, or “in-between,” – yet, there is no validated measure of this experience, recognized as subjective liminality. The present research aims to (1) operationalize subjective liminality and (2) develop and validate a scale to measure it.

Design/methodology/approach

A literature review was used to operationalize subjective liminality, and the scale validation was performed using four separate samples: 150 workers on M-Turk, 151 graduate and professional students at a large Midwestern University, 252 unemployed individuals in the US and Canada, and 416 full-time employed individuals in the US.

Findings

Subjective liminality was conceptualized as a second-order latent construct reflected by three dimensions: feelings of anxiety, ambiguity and reduced group identification. A 9-item scale was developed and validated to measure it.

Originality/value

This study clarifies and measures an emergent construct in the career transition and organizational change literature.

Citation

Dhar, U. and Boyatzis, R. (2023), "Development and validation of a scale to measure subjective liminality: individual differences in the perception of in-betweenness", Journal of Organizational Change Management, Vol. 36 No. 8, pp. 129-140. https://doi.org/10.1108/JOCM-07-2023-0279

Introduction

The idea that people feel betwixt or “in limbo” when they encounter individual or organizational-level change is not new. Such feelings have been documented among people encountering change in their professional lives, that is, a career transition that implies movement from one position to another which is directly observable, measurable and verifiable (Chudzikowski, 2012), like returning to work after maternity leave (Millward, 2006) or an organization-level change, such as during a merger and acquisition (Choi, 2011). Similar feelings have also been reported in workers who are in “in-between” spaces – such as expatriate workers (Au and Fukuda, 2002). The term used in the literature to refer to this phenomenon is “subjective liminality” (Ibarra and Obodaru, 2016), drawing upon ideas introduced by Ibarra and Petriglieri (2010) in a highly cited Journal of Organizational Change Management conceptual piece.

The liminal experience is a high-intensity event from which new perspectives can emerge, which leads to a broader and improved understanding of prevailing conventions. The degree of subjective liminality predicts the extent to which people will display more agentic playfulness in trying out new provisional selves (Ibarra and Obodaru, 2016). Given the right conditions, subjective liminality might offer a more fertile ground for identity development compared to traditional institutionalized liminal phases such as participating in a formal training program. People who can manage the subjective liminal experience effectively can be more successful in the context of modern careers that are increasingly “boundaryless” and “protean” or self-driven (Briscoe and Hall, 2006). Given the distinction between objective and subjective liminality, and the implications of the latter in the modern workplace, studying within population variance in the subjective experience will also be important for future theory building.

Despite the promise this construct holds for research on individual and organizational level change, there has been no systematic review yet and no validated measurement exists. Moreover, while theorists argue that subjective liminality is closely tied with personality and can be conceptualized as both a state and a trait, empirical evidence is lacking that can throw light on the precise nature of the relationship (Ibarra and Obodaru, 2016). The objective of this research is to create an operational definition of subjective liminality through a literature review; develop a scale to measure subjective liminality and demonstrate its validity and reliability. We followed a scale development and validation process as outlined by Hinkin (1998). Studies 1 through 3 were approved by the IRB of the second author's university. Study 4 was approved by the IRB of the first author's university.

Operational definition

Subjective Liminality refers to the feeling of “in-betweenness” associated with doubt, uncertainty, confusion and anxiety (Ibarra and Obodaru, 2016; Ibarra and Petriglieri, 2010). Consistent with that framing, we define it as the internal, subjective experience of feeling suspended or in a betwixt state, which may or may not be related to any specific objective change. That is, subjective liminality is conceptually distinct from objective liminality, originally conceptualized to describe institutional rites of passage (Turner, 2008). Subjective Liminality is closely associated with various forms of identity transitions, including changes in how people see themselves in their professional roles (Ibarra and Petriglieri, 2010). Over such a process of change, the person undergoes a shift in their internalized integrative narrations of their personal past, present and future self (McAdams, 1996). Such a narrative consists of, among other dimensions, their personality traits, like those included within the Big Five taxonomy, as well as tasks, goals, projects, tactics, defenses, values and other developmental, motivational, or strategic concerns that contextualize a person's life in time, place and role (McAdams, 1996).

In order to identify the dimensions of subjective liminality, we conducted a literature review of the last 20 years of published studies using keywords such as liminality, liminal periods, betwixt, boundary spanning and transition. The research databases and search tools J-stor, Scopus, Web of Science and Google Scholar were used for this purpose. We selected papers that addressed the subjective element of liminality by focusing on the feelings associated during change. In this manner, we identified 53 papers for our review (45 empirical and 8 conceptual papers). The first author reviewed these papers, taking a deductive approach to identify the dimensions. A deductive approach is more appropriate during the scale development process when the theoretical foundation provides enough information to identify the dimensions (Hinkin, 1998). In this case, building of the theoretical framing by Ibarra and Obodaru (2016) as well as the definition mentioned earlier; and by undertaking a careful reading of the results and conclusions of these 53 papers, the researchers were able to gather sufficient information to be able to propose possible dimensions. Following the initial review, the first author discussed the emerging dimensions with the second author, who also reviewed some of the articles independently, based on which three dimensions were identified. In the following sections, we provide the theoretical rationale for them.

Dimension 1: anxiety

Uncertainty, confusion and disorientation mark the subjective experiences of people undergoing career transitions, organizational change, as well as people in non-traditional work arrangements (Ibarra and Petriglieri, 2010). For example, Byrnes and Taylor's (2015) study of voluntary transition of CEOs demonstrated that their subjects experienced a diminished sense of self as part of their subjective experience of the liminal phase in their lives. Hennekam and Bennett's (2016) study of involuntary career transition among artists population showed how the loss of their creative identity is acutely felt and it leads to psychological stress and grief, making the professional transition problematic. Consider a person who has been recently promoted to a leadership position. While they may already be formally designated as a leader, anxiety about issues such as seeming bossy, seeming unqualified and seeming different from one's peers (Cunningham et al., 2023) can prevent them from fully embracing their new leader identity, leaving them in a state of high subjective liminality.

Dimension 2: ambiguity

Identity reconstruction during a liminal phase involves a disruption of certain elements of the old identity (Beech, 2011); and the transition involves cognitive and emotional processing in two domains: loss orientation and restoration orientation (Conroy and O'Leary-Kelly, 2014). This simultaneous loss and gain can create a sense of not knowing where one stands in the change process. For example, Millward's (2006) interviews with women returning to work after a maternity leave found that they struggled subjectively during the process of simultaneously maintaining their identity as a new mother and as a valued and functional member of the organization. Specifically, the acquisition of a “mother” identity during maternity leave creating return dilemmas. Bamber et al.'s (2017) research on occupational limbo using teaching-only staff (i.e. faculty members who focus on teaching over conducting research) as their subjects showed that they felt “locked-in” to an uncomfortable state by a set of structural and social barriers often perceived as insurmountable. They found themselves unable to cross the in-between space to the elevated status of “proper academics.” Similarly, Bettencourt and Brown (2003) had found that among employees of service organizations, even behaviors that can be considered liminal or boundary-spanning, such as others that their company is a good place to work, was associated with role ambiguity.

Dimension 3: reduced group identification

Transitions involve uncertainty in the social and cultural worlds of the individual (Crafter and Maunder, 2012). Infact, social exclusion or marginalization is common across a range of liminal experiences (Ibarra and Petriglieri, 2010). Consistent with that, we found evidence in our literature review that subjective liminality involves feeling entirely committed to any one particular social or work group. For example, George and Chattopadhyay's (2005) study of contract workers found that they identify with both the employing and client organizations based on perceived characteristics of the organization as well as social relations within the organization. Borg and Söderlund’s (2015) study of engineers working for one of Scandinavia's leading technical consultancies had also found that one of the ways in which they experience liminality is social or group related, which involves aspects of inter-personal relationships, the organization and the role specification. Consider people working in the liminal space of a multinational firm's subsidiary in a host country. Such employees can find themselves in limbo between embracing the organizational identity, which its foreign associations, and their national identity (Lee et al., 2022).

A summary of the three dimensions is provided in Table 1.

Study 1: item development and refinement

As the first step, based on the literature review and several rounds of discussions among the co-authors, a total of 23 items were generated across the three dimensions. The definitions of the items were used as a guide for the development of the items (Schwab, 1980). In doing so, the authors were mindful to capture all possible manifestations of the dimensions in the first round of item generation to create an exhaustive list. For example, ambiguity could manifest itself in terms of not knowing what one does as well as not knowing who one is. Every identified aspect of the dimensions was represented by at least two items in the initial list. Based on Hinkin's (1998) recommendations, we were sure to word the items consistently in terms of describing a single affective response. Next, we conducted a review of the items by a panel of experts in the field of career transitions and change management to determine content validity, clarity and relevance. The panel consisted of 5 faculty members from three different management colleges in the USA and 4 advanced doctoral students of Organizational Behavior at a large Midwestern American University. All 9 experts had several years of either full-time or part-time practitioner experience as organizational managers or consultants. Feedback was gathered through individual discussions with each of the experts. Some of the items were reworded based on the feedback and none were dropped. A draft questionnaire was developed based on this step. The questionnaire stem was as follows: “Lately, in my professional life, I have been feeling …”. The responses were on a 6-point scale ranging from strongly disagree to strongly agree. This was likely to reduce midpoint response bias (Baumgartner, 2006) and make the response range consistent with other instruments used in the validation process. The Anxiety dimension had 5 items, the Ambiguity dimension had 10 items and the Reduced Group Identification dimension had 8 items. It was considered acceptable to have varying numbers of items in each of the dimensions to begin with, since we wished to retain options to drop off at the factor analyses stage.

The next step involved Content Validity Assessment. The objective was to edit or delete items that are found to be conceptually inconsistent. We used the approach recommended by Schriesheim et al. (1993) to assess content adequacy with a recommended sample size of 50 for each dimension. 150 respondents on Amazon's M-Turk platform, an online participant pool in which individuals voluntarily participate in advertised studies in return for a small payment, were recruited as participants. Participants were divided into three groups across the dimensions. They were given the respective construct definition based on their group, along with the full item pool and asked the extent to which each item corresponds to the construct definition provided. For instance, respondents in the anxiety group were provided with the definition of the Anxiety dimension and asked to rate how much each item corresponded to that definition. Items were rated on a 5-point Likert scale ranging from 1 = very weak fit to 5 = very strong fit. Using a Q-correlation matrix and Principal Components Analysis, 18 items with factor loadings >0.4 and no major cross loadings were retained. The sorting analyses are available from the first author upon request.

Study 2: psychometric properties of the scale

The next phase involved assessing Internal Consistency of the Scale. Specifically, the study examined the underlying structure of the latent variable.

Sample and procedure

Participants were recruited from among the graduate and professional students at a large Midwestern University through a post in the University newsletter. Of the 151 students who completed the survey, 116 (76.8%) identified as female, 32 (21.2%) as male and 3 (2.0%) as “other.’ 64 (42.4%) were employed full-time, 20 (13.2%) were employed part-time and 67 (44.4%) were full-time students. Their age ranged from 20 to 56 years with a median age of 27 and a mean of 28.9 years. Items were edited or removed based on the following criteria: (1) A correlation above 0.8 with another item; or (2) multiple correlations above 0.75 with other items and (3) conceptual considerations.

Analyses and results

Based on these criteria, two items were dropped at this step. The two items were, “Suspended in-between who I am and who I want to be” from the Ambiguity dimension; and “Difficulty identifying with one group or another” from the Reduced Group Identity dimension. Next, an exploratory factor analysis was performed using principal components, which revealed that the item, “Like I am a stranger with everyone” from the Reduced Group Identity dimension did not belong to the dominant component and was dropped. Next, with the 15 retained items, a factor analysis was performed using the varimax rotation method with three factors, which confirmed the presence of three clear factors in the data. The three items with high cross-loadings (“like I am caught between two parts of myself”, “like I am somewhere in-between who I am and who I want to be”, “stuck in-between who I am and who I want to be”) were dropped and 12 items retained for the next step. Results of the factor analyses are available from the first author upon request.

Study 3: construct validity of the scale

The objective of this step was to assess the construct validity of the scale by demonstrating reliability, convergent validity and discriminant validity with conceptually related scales. The conceptually related constructs that were considered for this part of the analysis were as follows. A positively correlated construct is Ideal-Real Self Discrepancy (Watson et al., 2010) which refers to the perceived difference between the actual/own self-state and ideal self-states (i.e. representations of an individual's beliefs about his or her own or a significant other's hopes, wishes, or aspirations for the individual). Intentional change occurs when a person perceives a difference between their Ideal and Actual Selves (Boyatzis and Dhar, 2022). The negatively correlated constructs considered for the analyses were career satisfaction and work engagement. Career satisfaction is considered a measure for subjective career success (Hofmans et al., 2008). When people feel successful in their careers based on their internal standards, they do not see the need for a change and therefore subjective career success has been demonstrated to be negatively related to turnover intentions (Laschinger, 2012), reducing the likelihood of a person feeling liminal. According to Kahn (1990), engagement is the harnessing of people’ selves to their work roles; it means being psychologically present when occupying and performing a particular role. During a transition phase, especially one marked by strong feelings of being in limbo, mental and/or psychological attention is often divided between where one is/was and where they are headed, and therefore one is likely to be less engaged at work.

Sample and procedure

The questionnaire consisted of the subjective liminality items; and measures of constructs theoretically related to it. The sample consisted of people in the USA and Canada, who were unemployed at the time of the study – a group that is expected to demonstrate high levels of subjective liminality as they are presently between two jobs. Participants were recruited on prolific.co, an online platform that vets participants for social science research. Participants were paid a modest amount for participation. Of the 251 participants who returned the completed survey, 117 (46.6%) identified as female and 135 (46.4%) as male. Their age ranged from 25 to 55 years with a median age of 31 and a mean age of 34.1 years. A confirmatory factor analysis was performed using the varimax rotation method with three factors, which showed three clear factors in the data (Table 2). Items 4, 6 and 9 were removed because of high cross-loadings.

Measures

Subjective Liminality was measured by the 9-item scale developed after the CFA. Ideal-Real Self Discrepancy was measured using the Real-Ideal Self Discrepancy Test (Watson et al., 2010). Respondents were asked to think about the person they are at work and the person they would ideally like to be. This was followed by eight figures with two squares each, in which one square represented their actual self and the other square represented their ideal self. The level of overlap in the figures was a measure of the perceived discrepancy. Work Engagement was measured using the Utrecht Work Engagement Scale (Schaufeli et al., 2006). A sample item was “At my job, I feel strong and vigorous.” Career Satisfaction was measured using the Career Satisfaction Scale (Greenhaus et al., 1990). A sample item was “I am satisfied with the progress I have made toward meeting my overall career goals.” Since the sample consisted of unemployed individuals, the stem of the questions other than Subjective Liminality were modified to a more generic “With respect to your work-life, respond to the following statements” to ensure that people were responding to the questions with respect to how they generally felt when they work, instead of with reference to a specific job.

Analyses and results

This step of the data analysis was performed using MPlus, using the 9 retained items of the Subjective Liminality scale. The three-factor model showed good model fit (RMSEA = 0.095, at 90% C.I. = 0.072 to 0.118; CFI = 0.974; TLI = 0.962). Internal consistency of the scale was assessed based on the criteria of coefficient alpha >0.7. The coefficient alpha was found to be 0.946, which was considered acceptable. Next, we looked at the scale's relationships with the conceptually related and unrelated constructs. The means, standard deviations and correlation matrix of subjective liminality with the related constructs in shown in Table 3. As predicted, subjective liminality as measured by the scale was positively correlated with Self Discrepancy; negatively correlated with Work Engagement and Career Satisfaction.

Study 4: predictive validity of the scale

The objective of this step was to test the predictive validity of the scale by testing whether a perception of subjective liminality as measured by the scale is predicted by a transition or change in a person's professional sphere.

Sample and procedure

A quasi-longitudinal survey study design was used for this step of the scale validation. The first author collaborated with Qualtrics on a larger study about career transitions to recruit and collect data from workers in the USA satisfying the following criteria: (1) between ages 25 and 65, (2) currently employed full-time, (3) having at least 1 year of full-time work experience. Participants were recruited by Qualtrics. Of the remaining 416 valid responses, 159 (38.2%) identified as female and 257 (61.8%) as male. Their age ranged from 25 to 65 years with an average age of 44.5 years.

Measures

Subjective Liminality was measured by the 9-item scale used in the previous step. Objective change was measured by asking for a yes/no response to the question: “In the past 6 months, have you experienced, or are currently experiencing any transition or change in your professional life (for example – new job or role, company restructuring, etc.) or a change in your personal life with a direct effect on your professional life (for example – leaving for or returning from parental break, etc.).” Personality was measured by the 10-item short version of the Big Five Inventory (Rammstedt and John, 2007).

Analyses and results

The means, standard deviations and correlations among the variables are shown in Table 4. A t-test difference of means was performed to check if there was a statistically significant difference in the mean subjective liminality of the two groups, i.e. between those who reported no change (N = 309, M = 25.5, SD = 11.7) and those who did (N = 107, M = 28.3, SD = 11.1). The one-sided p-value was 0.015 (t = −2.17, equal variances assumed), showing that the group that experienced change reported a significantly higher level of subjective liminality. A multiple regression was performed to check if objective change predicts subjective liminality, while controlling for age, gender and the personality variables. The regression results show that objective change is a statistically significant predictor of subjective liminality (b = 0.067, p = 0.046). Age and personality dimensions (except openness) were also significantly related.

Discussion

In this study, we clarified the construct of subjective liminality and developed a measure for it. The validation process led to the novel insight that what has been considered so far as a single element, is actually an assorted heterogeneous phenomenon. The scale now allows for future quantitative research on topics around career transitions and individual-level responses to organizational change on large samples. For example, future studies using samples in several organizations, or across several teams, can examine the organizational and team-related factors that can help administer change with minimal psychological disruption for employees. In the area of career transitions, while we know that people experience subjective liminality, we do not have a clear picture of the dynamics of the phenomenon over a period and its conditional factors (Hennekam and Bennett, 2016). Researchers can measure subjective liminality over time to determine the individual and environmental factors that affect it at various stages of the change process. Having a scale that measures subjective liminality can help researchers determine which attributes of the transitional time, space and relationships are most conducive to identity growth (Ibarra and Obodaru, 2016).

Despite its strengths, the study has some limitations. The dependence on literature review to generate initial items might have prevented us from capturing other understudied dimensions of subjective liminality. The lack of multisource data might have introduced some inaccuracy in the strengths of the relationships. The cross-sectional design of the study limits our ability to make conclusions about causality. Future research can overcome these limitations and further validate the scale by expanding the nomological network of subjective liminality, using longitudinal data collection methods and testing the scale's cross-cultural validity. Finally, the finding in the regression analysis that all aspects of personality (except openness) have a significant relationship with subjective liminality needs to be investigated further. While we conceptualize subjective liminality as a state rather than a trait, if indeed personality has a strong effect on how much people feel a sense of being in limbo regardless of whether they are experiencing any transition, it could have significant implications for how career counsellors work with people having different personalities.

The study has implications for how organizations can support employees during times of transition or change. By understanding that subjective liminality can be “broken down” into feelings of anxiety, ambiguity and reduced social identification, organizations can tailor their response to each of those needs. Consider again the example of women returning to the workplace following maternity leave (Millward, 2006). In this case, the employees' feelings of anxiety can be addressed through supportive management that assures them of job security; feelings of ambiguity can be addressed by the presence of role models who have successfully straddled the twin identities of motherhood and employment; and feelings of reduced social identification can be addressed by building an inclusive team culture. For career coaches and counsellors, being aware of the three dimensions of subjective liminality can help them be more effective in helping their clients navigate difficult transitions at work (Dhar et al., 2021).

Conclusion

In the constantly evolving world of modern workplaces and careers, subjective liminality is ubiquitous. It is therefore critical for researchers to have access to an instrument that measures this phenomenon and include it as a variable in change-related research. This paper took the first steps towards developing and validating a measure of this phenomenon. Specifically, we established that the construct of subjective liminality has three dimensions: anxiety, ambiguity and reduced group identification. We also found that it is positively associated with self-discrepancy and negatively with career satisfaction and work engagement. We also showed that objective change predicts subjective liminality even after controlling for personality. We hope that this measure will add a new and important dimension to change management research in the future.

Table 1

Summary of the dimensions of subjective liminality

Dimension	Explanation	Theoretical rationale	Exemplary references
Anxiety	Anxiety associated with losing a hitherto valued part of oneself while simultaneously desiring something new	Identity reconstruction during a liminal phase involves a disruption of certain elements of the old identity and the transition involves cognitive and emotional processing about loss and restoration	Byrnes and Taylor (2015), Hennekam and Bennett's (2016), Cunningham et al. (2023)
Ambiguity	A sense of not knowing where one stands at the moment in their career journey	People experience feelings of uncertainty, confusion, and disorientation while undergoing a major change process	Beech (2011), Conroy and O'Leary-Kelly (2014), Millward (2006) Bamber et al. (2017), Bettencourt and Brown (2003)
Reduced group identification	Not feeling comfortable identifying oneself with any specific social group, such as an organization or a professional community	During change, the individual may feel less committed to any one particular work group, or sharing their commitment among one of more groups	Crafter and Maunder (2012), Borg and Söderlund (2015), George and Chattopadhyay (2005), Lee et al. (2022)

Source(s): Authors’ work

Table 2

Confirmatory factor analysis (N = 252) with 12 items

Item	Factor 1	Factor 2	Factor 3
More restless than usual about who/what I am	0.814	0.335	0.151
More dissatisfied than usual about who/what I am	0.799	0.238	0.407
As if I am no longer contented about who/what I am	0.779	0.257	0.407
UNHAPPIER than usual about myself	0.693	0.156	0.579
Like I do not know who I am anymore	0.528	0.260	0.682
Like I am in limbo	0.718	0.332	0.282
At a loss to describe myself to others	0.417	0.441	0.668
At a loss to identify the most important aspects about myself	0.359	0.427	0.731
As if there is no clear category to describe who I am	0.303	0.561	0.656
Uncomfortable identifying myself with any one specific group	0.181	0.802	0.439
Difficulty belonging to one group or another	0.271	0.867	0.252
Somewhat disconnected with the groups that matter	0.414	0.809	0.200

Source(s): Authors’ work

Table 3

Construct validity testing

	Mean	SD	Age	Gender	SL	SDSC	CSAT	ENG
Age	34.09	8.98	–
Gender	0.46	0.50	0.06	–
Subjective liminality (SL)	32.69	12.61	−0.14*	0.10	(0.95)
Self-discrepancy (SDSC)	61.56	24.91	−0.15*	0.04	0.40**	–
Career satisfaction (CSAT)	9.56	4.98	0.25**	0.06	−0.30**	−0.37**	(0.95)
Work engagement (ENG)	27.53	13.37	0.22**	0.05	−0.45**	−0.47**	0.58**	(0.96)

Note(s): Correlations with Subjective Liminality in italic; ** p-value>0.001, *p-value>0.05; Numbers in parameters are Cronbach's alphas

Pearson's Correlations, N = 252

Source(s): Authors’ work

Table 4

Pearson's correlations (N = 416)

	Mean	SD	Age	Gender	OC	SL	P(E)	P(A)	P(C)	P(N)	P(O)
Age	44.48	11.33	–
Gender	0.62	1.00	0.023	–
Objective change	0.26	0.44	−0.086	0.021	–
Subjective liminality	26.20	11.59	−0.209**	−0.004	0.106*	(0.94)
Extraversion P(E)	5.99	1.98	0.003	0.018	0.029	−0.261**	–
Agreeableness P(A)	7.26	1.80	0.072	−0.028	0.001	−0.351**	0.275**	–
Conscientiousness P(C)	8.28	1.70	0.270**	0.035	−0.105*	−0.444**	0.202**	0.392**	–
Neuroticism P(N)	5.25	2.07	−0.102*	0.160**	0.035	0.507**	−0.313**	−0.415**	−0.426	–
Openness	6.79	1.75	−0.080	0.005	−0.015	−0.043	0.040	0.071	0.048	−0.042	–
P(O)

Note(s): ** p-value> 0.001, *p-value>0.05; Numbers in parameters are Cronbach's alphas

Source(s): Authors’ work

Appendix Validated scale for measuring subjective liminality

On a scale of 1–6 indicate your agreement with the following statements (1 = strongly disagree, 6 = strongly agree)

“Lately, at work (or about your professional life in general), I am feeling …”

1. More restless than usual about who/what I am

2. More dissatisfied than usual about who/what I am

3. As if I am no longer contented about who/what I am

4. Like I do not know who I am anymore

5. At a loss to describe myself to others

6. At a loss to identify the most important aspects about myself

7. Uncomfortable identifying myself with any one specific group

8. Difficulty belonging to one group or another

9. Somewhat disconnected with the groups that matter

Measurement Key:

Dimension 1 (Anxiety): Items 1–3.

Dimension 2 (Ambiguity): Items 4–6.

Dimension 3 (Reduced Group Identification): Items 7–9.

Source(s): Authors work