Measurement Equivalence Test on Family Cohesion Scale: Comparison Models for White,Black,Hispanic,and Asian College or University Students in the U.S.

Abstract

Cohesion measure launched from the Family Adaptability and Cohesion Evaluation Scale (FACES) is a well-known instrument to assess family functioning. It is essential to perform a measurement equivalence test of cohesion targeting racially diverse families. The web-based self-administered survey was performed and 291 college/university students’ responses in Ohio were collected (Whites = 37.5%, Hispanics = 23%, Asians = 23%, and Blacks = 16.5%). This study explicated measurement equivalence for family cohesion structure (balanced cohesion, disengagement, enmeshed). The multiple group analysis was performed to investigate whether parameters in the measurement models of the “cohesion sub-structures (balanced cohesion, disengagement, enmeshed; 7 items of each)” were equivalent across the four racial groups. While “balanced cohesion” and “disengagement” structures showed measurement invariance across the groups, the “enmeshed” structure showed significant measurement variance across the groups. Two-items, “Family members feel pressured to spend most free time together” and “We feel too connected to each other” were inadequate for Hispanics’ enmeshment. One-item, “Family members have little need for friends outside the family” was inadequate to explain enmeshment of Blacks and Asians. Professionals should be aware of possible misinterpretations of results from the cohesion measure of FACES IV when they particularly assess the enmeshment status of racially diverse families with closer attention to cross-cultural comparability.

Plain Language Summary

Measurement Equivalence Test on the Family Cohesion Scale to Understand Its Applicability to Diverse Racial Groups

The goal of this study was to investigate whether the Cohesion Scale, which was derived from the FACES IV scale, could be used to measure the sub-structures of cohesion (i.e., balanced cohesion, disengaged, and enmeshed) among different ethnoracial groups (Whites, Blacks, Hispanics, and Asians). The study was carried out on undergraduate students who were currently attending universities or colleges. The study collected responses from 291 college/university students in Ohio through a web-based, self-administered survey, with Whites accounting for 37.5% of the participants, Hispanics accounting for 23%, Asians accounting for 23%, and Blacks accounting for 16.5%. The study employed the Amos program to conduct a multiple group analysis. The findings of the study revealed that the measurement invariance of the ‘balanced cohesion’ and ‘disengagement’ structures were consistent across the different ethnoracial groups. However, the ‘enmeshed’ structure showed significant differences in measurement among the groups. Specifically, the items “Family members feel pressured to spend most free time together” and “We feel too connected to each other” were not effective in reflecting enmeshment among Hispanics. Similarly, the item “Family members have little need for friends outside the family” was not a reliable measure of enmeshment among Blacks and Asians. Professionals should be aware of possible misinterpretations of results from the cohesion measure of FACES IV when they particularly assess the enmeshment status of racially diverse families with closer attention to cross-cultural comparability.

Keywords

family functioning family cohesion disengagement enmeshment diverse racial groups family adaptability and cohesion evaluation scale

Introduction

The family system is an important part of the personal environment because it plays a central role in the outcome of individual behaviors or functioning (Altiere & Kluge, 2008). One of the representative determinants of family functioning is a family cohesion (Altiere & Kluge, 2008; Craddock, 2001; Olson, 2011). Family cohesion is defined, “the emotional bonding that family members have toward one another” (Olson, 2011). The Circumplex Model (CM), which was initially developed by Olson, Sprenkle and Russell (1979), is the theoretical approach delineating aspects of cohesion, adaptability, and communication. Cohesion focuses on how family systems balance separateness versus togetherness (Olson, Waldvogel, & Schlieff, 2019). Adaptability was described as the capacity of a family or marital unit to modify its power dynamics and relational guidelines in reaction to challenges and changes in life circumstances. Communication serves to facilitate the relative levels of cohesion and adaptability within families. Communication within families could have an effect on their adaptability and cohesion (Kaewkerd et al., 2020; Sequeira et al., 2021). The current study focuses on the cohesion of families, which is one of the key constructs suggested by the Circumplex Model.

Families striking equilibrium between separateness and togetherness are considered to have balanced cohesion. Individuals in a balanced cohesive family unit can simultaneously be independent from and connected to their family members. However, when a cohesion level is too high (i.e., enmeshed status), there is too much consensus or emotional closeness within the family and too little independence. Conversely, if family members have a higher level of independence with limited attachment or commitment to the family, the families are considered to be disengaged (Olson et al., 2019). Being at either extreme end of the spectrum is believed to be maladaptive.

Experts projected that ethnic diverse populations will make up nearly half of the population by the year 2050 (Atkin et al., 2022; Burlew et al., 2009). Concerning diverse families in the U.S., it is very important to understand the potential effects of culture on family functions (i.e., cohesion levels, disengaged levels, or enmeshed levels; National Research Council, 2001). In 2022, there were 75.8% of Whites, 18.9% of Hispanics, 13.6% of Blacks or African Americans, and 6.1% of Asians in the U.S. (U.S. Census Breau, 2022).

Each racial group has distinct differences in family dynamics because of variations in cultural heritage and environmental conditions (Lynch & Hanson, 2004). Racially diverse families tend to live in societies where specific racial or ethnic stratification significantly shape family resources and structures in different ways (Zinn, 1990). Individuals not only share common values and beliefs that their families have but also practice them. Those who are from diverse cultural backgrounds practice the values and beliefs based on nationality, common ancestry, or common cultural experiences (Danes et al., 2008), and therefore, racially diverse families’ functions (i.e., cohesion levels, disengaged levels, or enmeshed levels) could vary. Considering family cohesion, Augustine et al. (2022) applied family development theory to hypothesize that family cohesion would indirectly affect anxiety problems through self-regulation. Their study focused on African American participants at three different time points (ages 17, 19, and 21) as part of the Maryland Adolescent Development in Context Study, finding that family cohesion was linked to reduced anxiety problems indirectly through increased self-regulation. A study by Behnke et al. (2008) found that levels of family cohesion mediated the relationship between stress and parenting behaviors, among Mexican American and European American families. Rivera et al. (2008) examined whether family cohesion moderates the impact of cultural conflict on psychological distress in a representative sample of U.S. Latinos (N = 2,540). Their study found that a significant relationship exists between family cohesion and lower psychological distress in the Latino population. Further, strong family cohesion combined with family cultural conflict was associated with higher psychological distress. However, the study also identified variations among different Latino subgroups, including Puerto Ricans, Cubans, and Mexicans. This suggests that family cohesion did not significantly influence on all Latino populations. Similarly, Reeb et al. (2015) found that family cohesion was not significantly influenced Latin student groups: the author investigated how family cohesion influences alcohol-related issues in adolescents and how this relationship differs across racial and ethnic groups, using nationally representative data. The results indicated that while family cohesion had a significant impact on White adolescents, it was not significantly associated with alcohol-related problems among Latin adolescents. With regard to the Asian Americans and Latino Americans (Park et al., 2014), family cohesion was not significantly associated with psychological distress among older Asian adults. However, among older Latino adults, there was a significant relationship between family cohesion and psychological distress.

With respect to the family enmeshment status, Chun and MacDermid (1997) found that Korean adolescents reporting higher levels of family enmeshment tended to have higher levels of self-esteem. In this case, higher levels of enmeshment positively influenced Korean adolescents. Fuhrman and Holmbeck (1995) said that Black adolescents having higher levels of autonomy (lower levels of family enmeshment) showed more behavioral problems and lower grade point averages. Annunziata et al. (2006) also found that family closeness had a beneficial effect on school engagement among younger Black adolescents, particularly in the context of appropriate parental monitoring practices. These findings reject the claim that higher levels of family enmeshment leads to negative outcomes. However, opposite evidence also exists supporting the proposition that higher enmeshment is linked with negative outcomes. Specifically, one study targeting the majority of white adolescents (approximately 18% were non-white adolescents; Kivisto et al., 2015) reported that adolescents who perceived higher family enmeshment demonstrated higher levels of emotional dysregulation such as negative global appraisals of distress tolerance, a stronger increase in subjective negative mood from baseline to post-challenge, lower baseline vagal tone, and vagal augmentation during the challenging task. Across diverse families, what enmeshment means likely varies.

Regarding these possibility of various functioning of racially diverse families, scholars questioned, “Could the general conceptualization of family functioning (e.g., cohesion, disengagement, and enmeshment) be universally applied to racially diverse families?” (Chun & MacDermid, 1997; Manzi et al., 2006). In conjunction, these prior studies indicate that assessing family functions across cultural groups would not be stable. Even with these inconsistent effects of cohesion on family functions across diverse racial families, our understanding of the meaning of a given family function to a diverse ethnoracial group is still limited. These prior study findings relight the question about the universal application of general conceptualizations of family functions (cohesion, disengagement, enmeshed) to different racial families.

An appropriate approach to examine the feasibility of universal utilization of a conceptualization of family functioning would be to investigate the measurement equivalence of the concept across diverse groups. The issue of measurement equivalence for a specific measure is particularly important given the evidence of the variability in the culturally prescribed roles and practices associated with the family and in family structures (Knight et al., 1992). If the measures of family functioning used in research involving diverse cultures and races cannot consistently and accurately capture the same underlying meaning across all groups, our comprehension of cultural or racial distinctions in family functioning will be restricted or entirely inaccurate. However, not many studies performed cross-cultural equivalence tests for measures assessing family functioning. Hence, assessing cross-cultural comparability is necessary for precise and significant evaluations of family functioning among various groups (Choi et al., 2006).

Accordingly, in the current study, as Byrne (2001) indicates, if scores on a measure of the family functioning structure have similar or same precursors, consequents, and correlates across racial families, this measures’ functional equivalence can become evident. Also, when the structure of the family functioning is operationalized in the same procedure across racial families and given operationalization is equally meaningful across racial families, equivalence of family functioning structure operationalization can become evident. Importantly, if each item of family functioning has the same meaning across racial families, item equivalence will become evident (Byrne, 2001). Broadly speaking, measurement equivalence is achieved when a measure assesses the same underlying concept across groups (Burlew et al., 2009).

The Research on Equivalence Test of the Family Adaptability and Cohesion Evaluation Scale

Of the evaluation measures for family functioning, the Family Adaptability and Cohesion Evaluation Scale (FACES) suggested by the Circumplex Model (CM) is the well-known measure assessing family functions including cohesion, disengagement, and enmeshment (Olson et al., 1979). Olson (2011) mentioned that many prior studies used a version of Family Adaptability and Cohesion Evaluation Scale (FACES) since the first version (FACES I; Russel, 1979) of this scale has been published. The latest version is the FACES IV (Olson et al., 2019). Although several studies (Harachi et al., 2006; Knight et al., 1992; Manzi et al., 2006) explicated a cross-racial/ethnic equivalence of the “cohesion structure” of the FACES series, findings were inconclusive. Vega et al. (1986) used FACES II to investigate descriptively whether cross-ethnic differences exist in familial cohesion and adaptability. Their findings revealed that there were no significant differences between Mexican Americans and Anglos in relation to family functioning dimensions like cohesion and adaptability. Nevertheless, there were slight variations in the distribution of scores across the four adaptability levels, with Mexican Americans exhibiting a stronger inclination towards the highest level of adaptability, indicating an outstanding ability to handle family structure and role-related issues with flexibility. Knight et al. (1992) assessed whether the sub-structures of the “cohesion structure FACES II” were equivalent across the ethnoracial groups (Anglo Americans or Hispanics). The results indicated that items adequately reflected Anglo American mothers’ cohesion, while items poorly explained Hispanic mothers’ cohesion. Manzi et al. (2006) performed the structural multiple group analysis for “cohesion and enmeshed” structures targeting the U.K. and Italian groups in Europe. The findings indicated that factor loadings on two items of the enmeshment structure (“family members find it hard to get away from each other”; “family members feel pressured to spend most free time together”) and one item of the cohesion structure (“our family does not do things together”) significantly differed between groups. In other words, partial measurement nonequivalence existed. Most items similarly reflected the structures of cohesion and enmeshment in both racial groups, but there were slight differences in several items, suggesting different understanding or meaning across the groups. Harachi et al. (2006) also performed measurement invariance tests for the ‘cohesion’ structure of the FACES III, targeting Vietnamese, Cambodian, and European American mothers (reference group). Eight of 10 original items from the “FACES III-cohesion” structure were used: eight items were “1 = family members ask each other for help, 2 = we like to do things with just the family in our household, 3 = family members feel closer to other family members than to people outside the family, 4 = family members like to spend free time with each other, 5 = family members feel close to each other, 6 = when our family gets together for activities everyone is present, 7 = we can easily think of things to do together as a family, and 8 = family members consult other family members on their decisions.” Findings indicated measurement nonequivalent status and the factor loadings of all indicators (except for item 8) were significantly different between European Americans and Cambodians, while there was no significant difference in factor loadings between European Americans and Vietnamese.

In summary, the measurement invariance test findings targeting different racial or ethnic groups indicated that given items for the cohesion structure did not always equally reflect the structures (Harachi et al., 2006; Knight et al., 1992; Manzi et al., 2006; Vega et al., 1986). Evaluating measurement equivalence across diverse racial or ethnic groups is a fundamental issue in cross-cultural research. If the cohesion structure does not function equivalently across different groups, cross-cultural comparisons utilizing this measure could cause misleading interpretations of findings. Variance in item functioning could occur when individuals from different racial or ethnic groups show unequal probabilities of endorsing specific items despite having the equal levels of the underlying structures (Kim et al., 2016). The concept of measurement equivalence or bias relates to the possibility that even individuals with similar cultural backgrounds regarding family cohesion may provide different responses to questions about their experiences, influenced by factors such as their race or ethnicity (Carle et al., 2012). Therefore, future studies should seek to gain in-depth knowledge of whether the latest cohesion structure launched from the FACES IV equitably explains diverse racial or ethnic groups’ family cohesion structures. Also, since few previous studies have examined specific cohesion sub-structures, such as Balanced Cohesion, Disengaged, and Enmeshed, conducting a detailed investigation of these sub-structures is critical. In addition, considering that many studies included a broad range of age groups (both young adults and middle-aged adults), future studies may need to specifically target a specific age-range to keep moderate-higher levels of homogeneity of the target population. The current study focused on undergraduate students in the U.S.

Purpose of this Study and Research Question

Using the latest Cohesion Scale (launched from FACES IV scale), the current study aimed to evaluate whether parameters in the measurement models of the cohesion sub-structures (structure 1 = balanced cohesion, structure 2 = disengaged, and structure 3 = enmeshed; Kline, 2015) were equivalent across diverse ethnoracial groups (Whites, Blacks, Hispanics, and Asians) or not, targeting undergraduate students who are studying in universities or colleges. If the measurement model parameters demonstrate a good fit for the data across all four racially diverse groups, our study would suggest that the sub-structures of the cohesion structure adequately reflect family cohesion functioning in the four groups. Conversely, if the measurement models do not fit well with the data from the four racial groups, our study may recommend modifications to the measurements to better reflect the cohesion structure.

The research question is as follow: “Is there significant measurement invariance among the White, Hispanic, Asian, and Black college or university students in the U.S. for the Family Cohesion sub-structures (balanced cohesion, disengagement, enmeshed) of the Olson Family Functioning Scale (FACES IV)?”

Methods

Study Participants Description

The complete dataset (no missing on the Cohesion measure from the FACES IV; the listwise functioning) included 291 undergraduate students. The majority of undergraduate survey participants were 2nd, 3rd, and 4th year students (n = 35.39% [2nd] and 26.11% [3rd and 4th] respectively). Of 291, ethnic breakdown was Whites (n = 109, 37.5%), Hispanics (n = 67, 23%), Asians (n = 67, 23%), and Blacks (n = 48, 16.5%). There were more females (n = 212, 72.9%) than males. Most of the sample were not married (96.9%) and had insurance (95.5%). More than half (n = 154, 53.7%) reported having mothers who received college or university level education, with many of those having received even higher levels of education (e.g., master level or more; 87, 30.3%). Father education included (n = 132, or 46.8%, received college or university levels education, and n = 101, or 35.8% obtained even higher levels of education. Residential status of participants were campus residence hall (n = 128, 44%), off-campus housing (n = 118, 40.5%), and parent/guardian’s home, other college/university housing, or fraternity/sorority house (n = 21, 7.2%; n = 14, 4.8%; n = 10, 3.4%, respectively). As Table 1 shows, grade, father’s education, parents’ marital status, and residential status showed significant differences across the four racial groups (Whites, Blacks, Hispanics, and Asians).

Table 1.

Sample Characteristics (N = 291).

	Whites (n = 109)		Blacks (n = 48)		Hispanics (n = 67)		Asians (n = 67)		χ ²	p-Value
Categories	N	%	n	%	n	%	n	%	χ ²	p-Value
Grade
1st	4	3.7	4	8.3	3	4.5	14	20.9	26.16	.010
2nd	40	36.7	13	27.1	31	46.3	19	28.4
3rd	31	28.4	13	27.1	14	20.9	18	26.9
4th	31	28.4	14	29.2	16	23.9	15	22.4
5th	3	2.8	4	8.3	3	4.5	1	1.5
Gender
Male	28	25.7	11	22.9	21	31.3	16	23.9	3.55	.738
Female	79	72.5	37	77.1	45	67.2	51	76.1
Other	2	1.8	0	0.0	1	1.5	1	0.0
Mother education
At least high school	13	12.1	7	15.2	16	23.9	10	14.9	5.05	.538
College/university+	60	56.1	24	52.2	35	52.2	35	52.2
Master+	34	31.8	15	32.6	16	23.9	22	32.8
Father education
At least high school	13	11.9	12	28.6	17	25.8	7	10.8	23.93	.001
College/university+	61	56.0	21	50.0	28	42.4	22	33.8
Master+	35	32.1	9	21.4	21	31.8	36	55.4
Marital status
Unmarried	106	97.2	46	100.0	62	92.5	66	98.5	6.31	.098
Married	3	2.8	0	0.0	5	7.5	1	1.5	6.31	.098
Parent marital status
Unmarried	29	26.6	26	54.2	25	37.3	7	10.4	27.90	.000
Married	80	73.4	22	45.8	42	62.7	60	89.6	27.90	.000
Residential status
Campus residence hall	38	34.9	20	41.7	28	41.8	42	62.7	23.18	.026
Fraternity or sorority house	4	3.7	0	0.0	5	7.5	1	1.5
Other college/university housing	3	2.8	3	6.3	4	6.0	4	6.0
Parent/guardian’s home	10	9.2	3	6.3	6	9.0	2	3.0
Other off-campus housing	54	49.5	22	45.8	24	35.8	18	26.9
Insurance status
Uninsured (or not sure)	2	1.8	2	4.2	6	9.0	3	4.5	4.93	.177
Insured	107	98.2	46	95.8	61	91.0	63	95.5	4.93	.177
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	F	p-Value
Cohesion
Balanced cohesion	28.45	5.25	28.02	4.45	28.37	5.54	27.57	6.02	0.42	.738
Disengaged	14.69	5.97	17.08	5.95	15.54	6.05	17.12	6.26	3.05	.029
Enmeshed	13.33	4.23	14.54	5.10	14.34	4.56	15.01	4.16	2.25	.083

Data Collection

Using a convenient sampling method, undergraduate students were recruited from two universities (targeted main campuses, regional campuses, and student associations) and community churches in Ohio, USA. The data collection process has been officially approved by the Institutional Review Board (IRB) from the lead university. From September to November in 2019, the web-based, self-administered survey was performed. Participants were first prompted to read a survey introduction and then the consent form. The agreement on the consent form automatically guided the participants to the main questionnaires. The survey included socio-demographic questionnaires and the Family Adaptability and Cohesion Evaluation Scale (FACES IV) questionnaire. Particularly, the questions for the “cohesion structure (sub-structures are balanced cohesion, disengagement, and enmeshed)” were included. In total, 337 undergraduate students completed the survey. As mentioned earlier, of the 337 participants, 291 were included in the complete data analysis for the measurement equivalence test.

Measures (Cohesion Launched from FACES IV: Balanced Cohesion, Disengaged, and Enmeshed)

FACES IV incorporates sub-structures (Olson, 2011). Originally, the second level sub-structures are “cohesion” and “adaptability (flexibility).” The current study focused on the “cohesion” structure only. The “cohesion structure” is reflected by three sub-structures (balanced cohesion, disengaged, and enmeshed). The first structure, balanced Cohesion, includes seven items: (1) Family members are involved in each other’s lives, (2) Family members feel very close to each other, (3) Family members are supportive of each other during difficult times, (4) Family members consult other family members on important decisions, (5) Family members like to spend some of their free time with each other, (6) Although family members have individual interests, they still participate in family activities, and (7) Our family has a good balance of separateness and closeness. The second sub-structure, disengaged, incorporates seven items: (1) We get along better with people outside our family than inside, (2) Family members seem to avoid contact with each other when at home, (3) Family members know very little about the friends of other family members, (4) Family members are on their own when there is a problem to be solved, (5) Our family seldom does things together, (6) Family members seldom depend on each other, and (7) Family members mainly operate independently. The last structure, enmeshed, includes seven items: (1) We spend too much time together, (2) Family members feel pressured to spend most free time together, (3) Family members are too dependent on each other, (4) Family members have little need for friends outside the family, (5) We feel too connected to each other, (6) We resent family members doing things outside the family, (7) Family members feel guilty if they want to spend time away from the family. All items for the constructs are five level Likert response options (1 = strongly disagree to 5 = strongly agree). More details are presented in the Table 2.

Table 2.

Item Information of the Cohesion Sub-Structures.

Sub-structures	Items	Mean (SD)	Alpha
Balanced cohesion	1. Family members are involved in each other’s lives	4.20 (0.98)	.88
	2. Family members feel very close to each other	3.85 (1.04)
	3. Family members are supportive of each other during difficult times	4.30 (0.89)
	4. Family members consult other family members on important decisions	4.05 (1.04)
	5. Family members like to spend some of their free time with each other	3.77 (1.03)
	6. Although family members have individual interests, they still participate in family activities	4.01 (1.00)
	7. Our family has a good balance of separateness and closeness	3.99 (1.03)
Disengaged	1. We get along better with people outside our family than inside	2.77 (1.20)	.88
	2. Family members seem to avoid contact with each other when at home	1.93 (1.05)
	3. Family members know very little about the friends of other family members	2.20 (1.15)
	4. Family members are on their own when there is a problem to be solved	1.84 (1.05)
	5. Our family seldom does things together	2.11 (1.16)
	6. Family members seldom depend on each other	2.19 (1.17)
	7. Family members mainly operate independently	2.80 (1.28)
Enmeshed	1. We spend too much time together	2.17 (0.94)	.74
	2. Family members feel pressured to spend most free time together	2.15 (1.07)
	3. Family members are too dependent on each other	2.04 (0.88)
	4. Family members have little need for friends outside the family	1.99 (1.03)
	5. We feel too connected to each other	2.27 (1.18)
	6. We resent family members doing things outside the family	1.48 (0.85)
	7. Family members feel guilty if they want to spend time away from the family	2.05 (1.14)

Note. Five level Likert response options are consistently applied to these sub-structures. (Strongly disagree, Generally disagree, Undecided, Generally agree, Strongly agree).

Data Analyses

Preliminary Data Evaluation

Univariate frequencies, descriptive statistics, and histograms as well as bivariate scatterplots were examined for outliers, adequate variability, skewness (<2) and kurtosis (<21; Curran et al., 1996): there were no outlier and no missing values for the scales of “balanced cohesion, disengaged, and enmeshed,” and there was no violation in normality, following guidelines (Curran et al., 1996). Assumptions and conditions for performing structural equation modeling (SEM) were also evaluated (Lomax & Hahs-Vaughn, 2013). Specifically, non-collinearity (Pearson’s r < .08) and non-problematic multivariate outliers (cases with non-extreme discrepancies across the squared Mahalanobis distance scores; Kline, 2015) were expected. Bivariate correlation matrices did not show multicollinearity across the four groups (Whites, Blacks, Hispanics, and Asians). The Mahalanobis distance scores indicated that the four groups did not show extreme outliers.

Configural Model and Covariance-Based Multiple-Group Invariance Test

Model Specification and Identification

As mentioned earlier, the current study focused on the “cohesion” structure reflected by balanced cohesion, disengaged, and enmeshed. The single-factor structure (Olson, 2011) of the three sub-structures was satisfied with the single-factor rule. T-rule was used for identification (if n is the number of observed variables in the model, the number of observations equals n [n + 1]/2 when means are not analyzed). Under the T-rule, the case of “known information > unknown information” indicates over-identification, the case of “known information = unknown information” indicates just identification, and the case of “known information unknown information” indicates under-identification (Kline, 2015). Based on the T-rule identification guideline, the three sub-structures were over-identified (known = 28, unknown = 14).

Multiple-Group Invariance Test

The configural model (no equality constraints imposed) was initially evaluated. Global model fit, including chi-square with p-value, Comparative Fit Index (CFI), Root Mean Square Error or Approximation (RMSEA), and Standardized Root Mean Residual (SRMR; Kline, 2015) were also checked. After that, the multiple group analysis was performed to investigate whether “cohesion sub-structures” were well reflected by given items or not, as well as to explicate whether measurement variance issues existed across the four groups. The first step in assessing multiple group equivalence involves ensuring that the number of factors and their loading patterns are the same across all groups, without imposing any equality constraints on the parameters. Then, when assessing measurement invariance, the examination is centered on determining whether the parameters in the measurement aspects of the model are equivalent among the four racial groups. To analyze the data, equality constraints are applied to specific parameters, which means that these parameters are restricted to be identical across all groups. The process involved estimating these parameters for each specific group, and then constraining the estimates for all other groups to be equal to that of the specific group. This process could be facilitated by utilizing the automated models of parameter subsets available in the multiple group analysis function of Amos (Byrne, 2001). As Figure 1 shows, FACES IV cohesion sub-structures were tested separately. In other words, above analytic performance were repeated for three times targeting the “balanced cohesion,”“disengaged,” and “enmeshed.”

Figure 1.

Cohesion sub-structures (balanced cohesion, disengagement, and enmeshed) launched from the FACES IV (Olson, 2011).

The main purpose of the configural model is to establish a baseline that serves as a point of comparison for all subsequent tests of invariance. In traditional methodology, evidence of non-invariance is established through the chi-square difference test, where the difference in values is distributed as a Chi-square with degrees of freedom (Byrne, 2001). If the Chi-square difference value is statistically significant, it indicates that there is evidence of noninvariance. In this particular study, the chi-square difference scores were evaluated and then the standardized regression weights were compared with the p-value to examine the extent of invariance.

Results

Configural Model and Fully Constrained Model

The configural models of “balanced cohesion” and “disengaged,” regardless of the racial groups, showed acceptable model fits based on the guidelines of Byrne (2001) and Bentler (1992). The configural model of “enmeshed” did not show good model fit scores relatively. Specifically, model fit scores of the “balanced cohesion” structure were X² = 75.44, CFI = 0.97, RMSEA = 0.05, SRMR = 0.04. For the “disengaged,”X² = 101.96, CFI = 0.95, RMSEA = 0.06, SRMR = 0.05. Regarding the “enmeshed,”X² = 108.69, CFI = 0.90, RMSEA = 0.06, SRMR = 0.08. The three sub-structures showed acceptable internal consistency scores: cohesion (7 items; α = .88), disengaged (7 items; α = .88), and enmeshed (7 items; α = .74).

As Table 3 shows, the “balanced cohesion” and “disengaged” structures that were constrained with all “factor-loadings-equal” showed that ΔX²(s) were not significantly different across the four racial groups (Whites, Blacks, Hispanics, and Asians). This result indicated the measurement invariance status of the whole models of “balanced cohesion” and “disengaged” structures for the four racial groups (Byrne, 2001; Kline, 2015). However, the “enmeshed” structure that was constrained with all factor loadings showed that ΔX² was significantly different (Table 3). Given these findings on the “enmeshed” structure, an additional evaluation was conducted (Muthén & Muthén, 2017) comparing all pairs of the four groups and then evaluating the magnitude of the factor-loading scores with p-value (weak factor loadings < |0.30|; acceptable factor loadings ≥ |0.40|; p < .05; Pett et al., 2003).

Table 3.

Goodness-of-Fit Statistics for Test of Multiple Group Invariance (N = 291).

Categories: Balanced cohesion, Disengaged, and Enmeshed	χ ²	df	Δχ²	Δdf	p-Value	CFI	ΔCFI
Balanced cohesion
1. Configural model (no equality constraints imposed)	75.44	44			.002	0.97
2. Measurement weights constrained			18.19	18	.44	0.97	0.000
Disengaged
1. Configural model (no equality constraints imposed)	101.96	52			.000	0.95
2. Measurement weights constrained			20.68	18	.30	0.943	0.007
Enmeshed
1. Configural model (no equality constraints imposed)	108.69	56			.000	0.90
2. Measurement weights constrained			43.14	18	.001	0.83	0.07

Note. χ² = Ch-square; Δχ² = changes in Chi-square; Δdf = changes in degree of freedom; CFI = comparative fit index; ΔCFI = changes in comparative fit index. Significant p-value (based on .05 level) is bolded.

Measurement Non-Equivalence on Enmeshed Structure

Comparison findings (i.e., enmeshed structure) indicated that the groups of Whites, Blacks, Hispanics, and Asians showed measurement non-equivalent status. Table 4 presents standardized regression weight comparisons from the SEM multiple group analyses targeting the four groups. Within White students, the seven items strongly reflected the structure of the “enmeshed structure.” For Black students, except for item 4 (“family members have little need for friends outside the family” [factor estimate = .22]), the remaining six items reflected the “enmeshed” structure very well. In Hispanic students, items 2 and 5 (“family members feel pressured to spend most free time together” [factor estimate = .36]; “we feel too connected to each other” [factor estimate = .37]) did not significantly reflect the “enmeshed” structure. Among Asian students, item 4 showed very weak (factor estimate = .10) and non-significant reflection on the “enmeshed” structure Details are presented in the Table 4 (below).

Table 4.

Standardized Regression Weight Comparisons: From Multiple Group Analyses for the Four Groups (N = 291).

Item	Whites (n = 109)			Blacks (n = 48)			Hispanics (n = 67)			Asians (n = 67)
Item	Standardized	Un-standardized	SE	Standardized	Un-standardized	SE	Standardized	Un-standardized	SE	Standardized	Un-standardized	SE
1	0.43	1.00		0.62	1.00		0.34	1.00		0.79	1.00
2	0.66***	2.05	0.55	0.60***	1.04	0.30	0.36	1.28	0.69	0.55	0.63***	0.15
3	0.66***	1.55	0.42	0.87***	1.27	0.27	0.46*	1.35	0.66	0.85	0.90***	0.14
4	0.49***	1.13	0.35	0.22	0.41	0.29	0.65*	2.51	1.10	0.10	0.11	0.16
5	0.31**	1.06	0.44	0.88***	1.35	0.29	0.37	1.53	0.82	0.58	0.78***	0.17
6	0.46**	0.86	0.28	0.65***	1.07	0.28	0.75*	2.38	1.02	0.42	0.44***	0.14
7	0.68***	2.19	0.59	0.77***	1.41	0.33	0.60*	2.30	1.02	0.48	0.64***	0.17

Note. Item 1 = we spend too much time together; Item 2 = family members feel pressured to spend most free time together; Item 3 = family members are too dependent on each other; Item 4 = family members have little need for friends outside the family; Item 5 = we feel too connected to each other; Item 6 = we resent family members doing things outside the family; Item 7 = family members feel guilty if they want to spend time away from the family.

***

p ≤ .001, **p < .01, *p < .05.

As Table 4 and Appendix 1 show, the evaluation findings on each pair of the group indicated a consistent pattern with the findings of the multiple group analysis including the four groups simultaneously. In conclusion, items 2 and 5 out of the seven items in the “enmeshed” structure posed problems in explaining the enmeshed structure for Hispanic students, and item 4 was inadequate in explaining the enmeshed structure for Black and Asian students, based on the factor coefficient magnitude and p-value, as they did not significantly reflect the structures.

Discussion

In this study, our objective was to conduct a measurement equivalence test of cohesion among racially diverse college students, including Whites, Hispanics, Asians, and Blacks. The results indicated that while the “balanced cohesion” and “disengagement” structures exhibited measurement invariance across the groups, the “enmeshed” structure exhibited significant measurement variance across the groups. Specifically, two items, “Family members feel pressured to spend most free time together” and “We feel too connected to each other,” were not adequate (not significant) in explaining enmeshment among Hispanics. Likewise, one item, “Family members have little need for friends outside the family,” was not adequate (not significant) in explaining enmeshment among Blacks and Asians. The following discussions are based on conceptual and methodological approaches.

Conceptual Approach

The study findings for the “enmeshed” structure (i.e., measurement variance status across the racial families) suggest the possible variation in cultural background, identity, or beliefs. One explanation of these findings includes possible differences in the salience of collectivism and individualism across the various groups, particularly in that it shed light on how family enmeshment is interpreted as it relates to differences in construal of the self (Markus & Kitayama, 1991). This is likely because the centrality of closeness within family units is related to how one sees themselves in relation to others (e.g., individualistically or collectivistically; Markus & Kitayama, 1991). In other words, family closeness has the potential to be differently understood across groups because groups vary in terms of their cultural values (Green et al., 2005).

Considering current findings, we could suggest possible reasons of measurement variance status of the “enmeshed structure”: (1) First, for Hispanic students, they might think that the meanings of given items (“family members feel pressured to spend most free time together” and “we feel too connected to each other”) are not significantly reflecting the concept of “enmeshed (i.e., extreme closeness).” According to several studies, many Hispanics consider a strong family relationship is one of the important factors of life, and they could feel obligated (Forst & Lehman, 1997; Ma et al., 2017; Schwartz et al., 2005). In other words, Hispanic students in the current study may recognize the power of familial collaboration, connections, and reliability as an essential component for their self-growth and the cultural groups’ needs. This can be differently seen from White, mainstream points of view which highlight independence, freedom, and autonomy. So, we could carefully comment that Hispanic students’ experiences and values may not be completely captured by items 2 and 5 launched from the “enmeshed” structure; (2) Second, regarding Asian students in this study, we first need to clarify the racial and ethnic composition of this group. The majority of Asians in this study were Korean heritage: approximately 70% of Asian students were Korean and the remaining 30% were diverse Asian subgroups (i.e., Chinese, Filipino, Indonesian, Japanese, Vietnamese, and Taiwanese). Some studies found that Korean Americans were more likely to be bicultural than Chinese Americans (Lee et al., 2001; Yeh et al., 2005). Specifically, Koreans tended to retain original heritage and identity while integrating American culture into their identity and lifestyle. Based on bicultural perception (i.e., not skewed to one side), Korean American students might not think that item 4 (“Family members have little need for friends outside the family”) reflects extreme closeness (i.e., enmeshed). According to Chung (2001), although Korean American students (ages 17–30 years) tended to keep the balance between their original culture and new American culture; they were still significantly influenced by their family members when they were in decision-making process for educational career, dating, or marriage. Compared to other Asian groups such as Japanese, Chinese, or Filipino American students, Korean students were more likely to be influenced by their family members in life trajectories. As Korean Americans’ ways of thinking, believing, or perceiving could be significantly influenced by their original culture and their family members, item 4 might not be considered as a part of enmeshed structure. Often, Asian cultures are thought to be highly cohesive collectivist cultures (Holland & Palaniappan, 2012). Not only Korean American students but also general Asian students tend to simply rely more on their parents for academic and career guidance (Okubo et al., 2007), and in turn, feel like ties to anyone outside of their family is not as important; (3) Last, the current study also found that item 4 poorly reflected Black students’“enmeshed” structure. Prior studies consistently highlighted the importance of family closeness of Black university or college students (Annunziata et al., 2006; Fuhrman & Holmbeck, 1995). For Black college students, family closeness was positively related to self-concept, and the extent to which they internalize racially-biased experiences (Wilson & Constantine, 1999). Therefore, seeking others’ support outside their families could be considered as “disengaged family status.” Black university or college students generally think that support from their family members (or parents) are significantly beneficial for their life course (Annunziata et al., 2006).

Methodological Approach

The findings of the current study (measurement non-equivalence on enmeshed structure) could be possibly related to methodological artifact (e.g., sample characteristics effect, analysis strategies, etc.; Cauce et al., 1998; Gottfredson & Koper, 1997; Harachi et al., 2006; Knight & Hill, 1998; Pett et al., 2003).

To address the issue of invariance in factor loadings, one possible approach is to remove items that exhibit substantial differences across groups. However, it’s essential to be careful when considering the removal of multiple items from a widely-accepted scale like FACES IV (i.e., Gold Standard Scale). Several scholars argue that it is not strongly recommended to eliminate items that are highly likely to capture the variability of interest (Gottfredson & Koper, 1997; Harachi et al., 2006). Regarding the fact that FACES series are widely utilized and considered as gold standard measures capturing family functioning, we should be careful to suggest measurement modification. Therefore, repeated future studies targeting similar or same target groups with the current study (i.e., university or college students) should be performed to determine whether the cohesion structure is equivalent across racially diverse university or college students.

Otherwise, before deciding to drop items that inadequately reflect the construct, it is recommended to perform a reliability test (Pett et al., 2003). In our study, when we conducted a reliability test for the “enmeshed structure” of Hispanics, all seven items showed good reliability (α = .71). However, removing the items “family members feel pressured to spend most free time together” and “we feel too connected to each other” resulted in slightly decreased α scores (from .71 to .69 and from .71 to .70, respectively). The reliability test for the “enmeshed structure” of Blacks showed strong reliability, with all seven items achieving α score of .82. Removing the item “family members have little need for friends outside the family” resulted in a slight increase in α scores (from .82 to .86). The reliability test for the “enmeshed structure” of Asians showed good reliability, with all seven items achieving α score of .73. Removing the item “family members have little need for friends outside the family” resulted in a slight increase in α scores (from .73 to .78). Regarding these findings, while some scholars may suggest removing problematic items, others may argue for keeping them due to their importance in capturing the construct. It is important for future studies to carefully deliberate before removing specific items from constructs.

The presence of underlying heterogeneity across groups can complicate the assessment of measurement equivalence across ethnically diverse groups (Harachi et al., 2006; Knight & Hill, 1998). Some scholars (Cauce et al., 1998; Harachi et al., 2006) have argued that racial variability in measurements can be reduced or eliminated by statistically or experimentally controlling for socio-demographic characteristics. In the current study, significant differences were found across ethnic groups in terms of grade (e.g., 1st–5th), father’s educational attainment, parents’ marital status, and other variables. To mitigate these confounding factors, future studies could consider recruiting ethnic groups with similar socio-demographic characteristics, after weighing the pros and cons of doing so.

Limitations and Future Suggestions

Although this study is significant to bring more discussions about different racial, ethnic family dynamics and functions, there are several limitations. First, in terms of data collection, participants in this study stayed in Ohio areas when the data were collected, so empirical findings may not generalize to students residing in other geographic areas in the U.S. In addition, this study used convenience sampling strategy, which in turn could also cause lower levels of generalizability of the study findings compared to the random sampling. Second, while the full FACES IV scale includes 60 items to capture all constructs of the Circumplex model (i.e., Cohesion, Adaptability [i.e., Flexibility], and Communication), the current study utilized only 21 items to capture the cohesion construct. Last, even though the sample size was not too small to perform the multiple group analysis, it would have been preferable to have a more balanced distribution of larger sample sizes across the four racial groups.

Future studies need to include full items of the FACES IV scale to capture not only cohesion structure but also adaptability (i.e., flexibility) structure and communication structure. And then, the evaluation of the full items of the FACES IV scale for measurement equivalence status may be extended to encompass a broader range of racial groups. To improve generalizability and account for the diverse variability of ethnic groups, future studies should aim to recruit larger and more diverse groups from a wider range of geographic areas.

If measurement non-equivalence issues persist despite implementing the suggested procedures in future research, it may suggest that the items are not effectively reflecting the intended structures. In this case, researchers or practitioners need to pay close attention to cross-cultural comparability when interpreting results from the cohesion measure (launched from FACES IV) so that they can prevent misclassification of target groups’ family functioning status. In this case, it could be imperative to re-develop culturally relevant measurements to analyze family cohesion among various ethnic and racial groups in a broader and comprehensive perspective.

Supplemental Material

sj-doc-1-sgo-10.1177_21582440241253954 – Supplemental material for Measurement Equivalence Test on Family Cohesion Scale: Comparison Models for White, Black, Hispanic, and Asian College or University Students in the U.S.

Supplemental material, sj-doc-1-sgo-10.1177_21582440241253954 for Measurement Equivalence Test on Family Cohesion Scale: Comparison Models for White, Black, Hispanic, and Asian College or University Students in the U.S. by En-Jung Shon, Lena Lee, Youn Ki, Siyoung Choe, Anthony James and Eunice Lee in SAGE Open

Footnotes

Acknowledgements

None.

Declaration of Conflicting Interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the internal grant in the university (i.e., the lead university of the project).

Ethics Approval Statement

We conducted ourselves with integrity, fidelity, and honesty. We openly took responsibility for my actions, and only made agreements, which we intended to keep. We did not intentionally engage in or participate in any form of malicious harm to another person or animal.

ORCID iD

En-Jung Shon

Supplemental Material

Supplemental material for this article is available online.

Data Availability Statement

The research team (the corresponding author is the principal investigator) originally collected data using a web-based survey tool.

References

Altiere

M. J.

von Kluge

(2008). Family functioning and coping behaviors in parents of children with autism. Journal of Child and Family Studies, 18(1), 83. https://doi.org/10.1007/s10826-008-9209-y

Annunziata

Hogue

Faw

Liddle

H. A.

(2006). Family functioning and school success in at-risk, inner-city adolescents. Journal of Youth and Adolescence, 35(1), 100–108. https://doi.org/10.1007/s10964-005-9016-3

Atkin

A. L.

Christophe

N. K.

Stein

G. L.

Gabriel

A. K.

Lee

R. M.

(2022). Race terminology in the field of psychology: Acknowledging the growing multiracial population in the U.S. American Psychologist, 77(3), 381–393. https://doi.org/10.1037/amp0000975

Augustine

D. A.

Koss

K. J.

Smith

E. P.

Kogan

S. M.

(2022). The influence of family cohesion on self-regulation and anxiety problems among African American emerging adults. PLoS One, 17(1), 1–12. https://doi.org/10.1371/journal.pone.0261687

Behnke

A. O.

Macdermid

S. M.

Coltrane

S. L.

Parke

R. D.

Duffy

Widaman

K. F.

(2008). Family cohesion in the lives of Mexican American and European American parents. Journal of Marriage and Family, 70(4), 1045–1059. https://doi.org/10.1111/j.1741-3737.2008.00545.x

Bentler

P. M.

(1992). On the fit of models to covariances and methodology to the Bulletin. Psychological Bulletin, 112(3), 400–404. https://doi.org/10.1037/0033-2909.112.3.400

Burlew

A. K.

Feaster

Brecht

M.-L.

Hubbard

(2009). Measurement and data analysis in research addressing health disparities in substance abuse. Journal of Substance Abuse Treatment, 36(1), 25–43. https://doi.org/10.1016/j.jsat.2008.04.003

Byrne

B. M.

(2001). Structural equation modeling with AMOS: Basic concepts, applications, and programming. Taylor & Francis.

Carle

A. C.

Weech-Maldonado

Ngo-Metzger

Hays

R. D.

(2012). Evaluating measurement equivalence across race and ethnicity on the CAHPS cultural competence survey. Medical Care, 50(9), S32–S36. https://doi.org/10.1097/MLR.0b013e3182631189

10.

Cauce

A. M.

Coronado

Watson

(1998). Conceptual, methodological, and statistical issues in culturally competent research. In Isaacs

M. R.

Hernandez

(Eds.), Promoting cultural competence in children’s mental health services (pp. 305–329). Paul H. Brookes Publishing Co.

11.

Choi

Mericle

Harachi

T. W.

(2006). Using Rasch analysis to test the cross-cultural item equivalence of the Harvard Trauma Questionnaire and the Hopkins Symptom Checklist across Vietnamese and Cambodian immigrant mothers. Journal of Applied Measurement, 7(1), 16–38.

12.

Chun

Y.-J.

MacDermid

S. M.

(1997). Perceptions of family differentiation, individuation, and self-esteem among Korean adolescents. Journal of Marriage and Family, 59(2), 451–462. https://doi.org/10.2307/353482

13.

Chung

R. G.

(2001). Gender, ethnicity, and acculturation in intergenerational conflict of Asian American college students. Cultural Diversity and Ethnic Minority Psychology, 7(4), 376–386.

14.

Craddock

A. E.

(2001). Family system and family functioning: Circumplex model and FACES IV. Journal of Family Studies, 7(1), 29–39. https://doi.org/10.5172/jfs.7.1.29

15.

Curran

P. J.

West

S. G.

Finch

J. F.

(1996). The robustness of test statistics to nonnormality and specification error in confirmatory factor analysis. Psychological Methods, 1(1), 16–29. https://doi.org/10.1037/1082-989X.1.1.16

16.

Danes

Lee

Stafford

Heck

(2008). The effects of ethnicity, families and culture on entrepreneurial experience: An extension of sustainable family business theory. Journal of Developmental Entrepreneurship, 13, 229–268. https://doi.org/10.1142/S1084946708001010

17.

Forst

J. K.

Lehman

W. E. K.

(1997). Ethnic differences in the workplace environment by employees in two municipal workforces. Hispanic Journal of Behavioral Sciences, 19(1), 84–96. https://doi.org/10.1177/07399863970191006

18.

Fuhrman

Holmbeck

G. N.

(1995). A contextual-moderator analysis of emotional autonomy and adjustment in adolescence. Child Development, 66(3), 793–811. https://doi.org/10.1111/j.1467-8624.1995.tb00906.x

19.

Gottfredson

D. C.

Koper

C. S.

(1997). Race and sex differences in the measurement of risk for drug use. Journal of Quantitative Criminology, 13(3), 325–347.

20.

Green

E. G. T.

Deschamps

J.-C.

Páez

(2005). Variation of individualism and collectivism within and between 20 countries:A typological analysis. Journal of Cross-Cultural Psychology, 36(3), 321–339. https://doi.org/10.1177/0022022104273654

21.

Harachi

T. W.

Choi

Abbott

R. D.

Catalano

R. F.

Bliesner

S. L.

(2006). Examining equivalence of concepts and measures in diverse samples. Prevention Science, 7(4), 359–368. https://doi.org/10.1007/s11121-006-0039-0

22.

Holland

A. T.

Palaniappan

L. P.

(2012). Problems with the collection and interpretation of Asian-American health data: Omission, aggregation, and extrapolation. Annals Epidemiology, 22(6), 397–405. https://doi.org/10.1016/j.annepidem.2012.04.001

23.

Kaewkerd

Chaiyasit

Vibulchai

Kenthongdee

Sirisawat

Panput

Kotta

Chakansin

Intarachaisri

(2020). Key factors of family adaptation to the illness of family members: An integrative review. The Bangkok Medical Journal, 16(1), 95. https://doi.org/10.31524/bkkmedj.2020.13.002

24.

Kim

DeCoster

Bryant

A. N.

Ford

K. L.

(2016). Measurement equivalence of the K6 Scale: The effects of race/ethnicity and language. Assessment, 23(6), 758–768. https://doi.org/10.1177/1073191115599639

25.

Kivisto

K. L.

Welsh

D. P.

Darling

Culpepper

C. L.

(2015). Family enmeshment, adolescent emotional dysregulation, and the moderating role of gender. Journal of Family Psychology, 29(4), 604–613. https://doi.org/10.1037/fam0000118

26.

Kline

R. B.

(2015). Principles and practice of structural equation modeling (4th ed.). Guilford Publications.

27.

Knight

G. P.

Hill

N. E.

(1998). Measurement equivalence in research involving minority adolescents. In Steinberg

McLoyd

V. C.

(Eds.), Studying minority adolescents: Conceptual, methodological, and theoretical issues (pp. 183–210). Lawrence Erlbaum Associates Publishers.

28.

Knight

G. P.

Tein

J. Y.

Shell

Roosa

(1992). The cross-ethnic equivalence of parenting and family interaction measures among Hispanic and Anglo-American families. Child Development, 63(6), 1392–1403. https://doi.org/10.2307/1131564

29.

Lee

R. M.

Falbo

Doh

H. S.

Park

S. Y.

(2001). The Korean diasporic experience: Measuring ethnic identity in the United States and China. Cultural Diversity and Ethnic Minority Psychology, 7(3), 207–216. https://doi.org/10.1037/1099-9809.7.3.207

30.

Lomax

R. G.

Hahs-Vaughn

D. L.

(2013). An introduction to statistical concepts: Third Edition. Taylor & Francis.

31.

Lynch

E. W.

Hanson

M. J.

(2004). Developing cross-cultural competence: A guide for working with children and their families. Paul H. Brookes Publishing Company.

32.

Malcolm

L. R.

Diaz-Albertini

Sanchez

J. C.

Simpson

Cortes

Kibler

J. L.

(2017). Cultural assets and substance use among Hispanic adolescents. Health Education & Behavior, 44(2), 326–331. https://doi.org/10.1177/1090198116659440

33.

Manzi

Vignoles

V. L.

Regalia

(2006). Cohesion and enmeshment revisited: Differentiation, identity, and well-being in two European cultures. Journal of Marriage & Family, 68(3), 673–689. https://doi.org/10.1111/j.1741-3737.2006.00282.x

34.

Markus

H. R.

Kitayama

(1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224–253. https://doi.org/10.1037/0033-295x.98.2.224

35.

Muthén

L. K.

Muthén

B. O.

(2017). Mplus user’s guide: Statistical analysis with latent variables (Eighth Edition ed.). Muthén & Muthén Press.

36.

National Research Council. (2001). America becoming: Racial trends and their consequences: Volume I. The National Academies Press.

37.

Okubo

Yeh

Lin

P. Y.

Fujita

Shea

(2007). The career decision-making process of Chinese American youth. Journal of Counseling & Development, 85(4), 440–449. https://doi.org/10.1002/j.1556-6678.2007.tb00612.x

38.

Olson

(2011). FACES IV and the circumplex model: Validation study. Journal of Marital and Family Therapy, 37(1), 64–80. https://doi.org/10.1111/j.1752-0606.2009.00175.x

39.

Olson

D. H.

Sprenkle

D. H.

Russell

C. S.

(1979). Circumplex model of marital and family system: I. Cohesion and adaptability dimensions, family types, and clinical applications. Family Process, 18(1), 3–28. https://doi.org/10.1111/j.1545-5300.1979.00003.x

40.

Olson

D. H.

Waldvogel

Schlieff

(2019). Circumplex model of marital and family systems: An update. Journal of Family Theory & Review, 11(2), 199–211. https://doi.org/10.1111/jftr.12331

41.

Park

Unützer

Grembowski

(2014). Ethnic and gender variations in the associations between family cohesion, family conflict, and depression in older Asian and Latino adults. Journal of Immigrant and Minority Health, 16(6), 1103–1110.

42.

Pett

M. A.

Sullivan

J. S.

Lackey

N. R.

Lackey

N. M.

Sullivan

J. J.

(2003). Making sense of factor analysis: The use of factor analysis for instrument development in health care research. SAGE Publications.

43.

Reeb

Chan

Conger

Martin

Hollis

Serido

Russell

(2015). Prospective effects of family cohesion on alcohol-related problems in adolescence: Similarities and differences by race/ethnicity. Journal of Youth & Adolescence, 44(10), 1941–1953. https://doi.org/10.1007/s10964-014-0250-4

44.

Rivera

F. I.

Guarnaccia

P. J.

Mulvaney-Day

Lin

J. Y.

Torres

Alegría

(2008). Family cohesion and its relationship to psychological distress among Latino group. Hispanic Journal of Behavioral Sciences, 30(3), 357–378. https://doi.org/10.1177/0739986308318713

45.

Russel

C. S.

(1979). Circumplex model of marital and family systems: Empirical evaluation with families. Family Process, 18(1), 29–45. https://doi.org/10.1111/j.1545-5300.1979.00029.x

46.

Schwartz

S. J.

Pantin

Prado

Sullivan

Szapocznik

(2005). Family functioning, identity, and problem behavior in Hispanic immigrant early adolescents. The Journal of Early Adolescence, 25(4), 392–420. https://doi.org/10.1177/0272431605279843

47.

Sequeira

Vicente

H. T.

Daniel

Cerveira

Silva

M. I.

Neves

Santo

H. E.

Guadalupe

(2021). Family adaptability and cohesion evaluation scale—version IV (FACES-IV): Validation study in the portuguese population. Journal of Child and Family Studies, 30(7), 1650–1663. https://doi.org/10.1007/s10826-021-01941-3

48.

U.S. Census Breau. (2022). QuickFacts data. https://www.census.gov/quickfacts/fact/table/US#

49.

Vega

W. A.

Patterson

Sallis

Nader

Atkins

Abramson

(1986). Cohesion and adaptability in Mexican-American and Anglo families. Journal of Marriage and the Family, 48(4), 857–867.

50.

Wilson

J. W.

Constantine

M. G.

(1999). Racial identity attitudes, self-concept, and perceived family cohesion in Black college students. Journal of Black Studies, 29(3), 354–366. https://doi.org/10.1177/002193479902900302

51.

Yeh

C. J.

P.-W.

Madan-Bahel

Hunter

C. D.

Jung

Kim

A. B.

Akitaya

Sasaki

(2005). The cultural negotiations of Korean immigrant youth. Journal of Counseling & Development, 83(2), 172–182. https://doi.org/10.1002/j.1556-6678.2005.tb00594.x

52.

Zinn

M. B.

(1990). Family, feminism, and race in America. Gender & Society, 4(1), 68–82. https://doi.org/10.1177/089124390004001006

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB