Introduction

Measuring family function and health is important. The family is the basic unit of society, and Asian cultures share a tendency toward collectivist social values, with the family group often prioritized over the individual (Ham, 2003). In particular, expectations that family members care for each other when sick are higher in Asian societies than, for example, in the United States (Song et al., 2016). Thus, understanding individual-centered health behaviors is generally insufficient to understanding the overall healthcare needs and preferences in Asian countries. Half (49.65%) of Korean households with family anxieties, worries, and conflicts recently identified family member health problems as the most important cause of family anxiety or conflict. The results of the 2021 Korea Welfare Panel Survey and Analysis Report of 5,996 households support the importance of family health (Jeong et al., 2021). However, few studies have been implemented to measure family health in the Korean context.

Researchers have recently examined Korean family relationships, communication, cohesiveness, flexibility (Dunst, 2021; S. S. Kim et al., 2021; H. Kim et al., 2022), family environments (H. J. Kim et al., 2018), and family resilience (Y. A. Kim et al., 2019). The Korean Family Strengths Scale, designed to assess family health, comprises 68 items covering cohesion, adaptability, communication, and supersystems (G. S. Yoo, 2004). However, the scale contains many items and is challenging to use. Subsequent researchers have applied it to subject groups other than adults or modified/extracted items without assessing the validity of the resultant abbreviated scales. A shortened version of the Korean Family Strengths Scale (30 questions) derived from the original scale (Y. Yoo et al., 2013) is faithful only to positive family-related aspects and strengths. In addition to reducing the number of questions by more than half, this abbreviated scale relied on information that is already a decade old. In addition, some parts cannot be considered valid because of limitations discovered in the data analysis such as the unequal distribution of participants in their 20s and 40s. In addition, the Korean Family Relationship Assessment Scale was recently developed, consisting of 15 items addressing the dimensions of family support, family conflict, and family togetherness. However, generalization of its results may be problematic because the underlying study recruited subjects with college-aged offspring only (S. S. Kim et al., 2021).

Friedemann (1998) developed the Assessment of Strategies in Families-Effectiveness (ASF-E), which uses the domestic situation to measure the functioning and health of families of various cultural backgrounds and educational levels based on systematic organizational theory. This theory and the four factors of person, environment, health, and nursing (representing the four meta-paradigms in nursing science) provide dynamic concepts to explain family and family health as a guideline to explain the functions of the individual, social, and environmental systems as well as the interactions between the systems. According to this theory, all of the systems (individual, society, and environment) seek harmony, and the system's basic structure is applicable to individuals, families, and even larger social systems. Because it relates directly to societal and national health, family health is important for professional nursing because of the positive effect on nursing efficiency of using the family-based approach strategy. The initial version of the ASF-E, developed in 1988, has 25 items and has been translated and verified with 20 and 10 items. It was developed and validated in the United States (Friedemann, 1998) and subsequently validated in Mexico (Chávez Aguilera et al., 2000), Finland (Åstedt-Kurki et al., 2001), Germany and Switzerland (Köhlen & Friedemann, 2006), Colombia (Gómez Bustamante et al., 2013), and Brazil (Lise et al., 2022).

The ASF-E has been used in various population groups to measure family health. Therefore, this study was designed to create a Korean version of the ASF-E (K-ASF-E) and validate its effectiveness in a Korean setting. In this methodological study, the scale was translated and validated using a systematic organizational framework. The psychometric properties of the scale are expected to help accurately measure family effectiveness and to be applicable to family-related research, interventions, and training programs conducted in community and clinical settings.

Methods Study Design

A methodological approach was used to verify the validity and reliability of the K-ASF-E. We verified the scale using content validity, construct validity (EFA, CFA), and reliability (Cronbach's alpha, intraclass correlation coefficient [ICC]) analyses.

Participants and Setting

Participants were recruited from June 2021 to August 2021 via computer-aided web interviews targeting the master sample panel of Korea Research Co., Ltd. Survey subjects prepared a quota table according to region, gender, and age in Korea, and 5 times as many as the survey subjects were sampled and sent an invitation email. Information in the invitation included the survey name, participation period, compensation offered, and estimated time required to complete the survey. All of the participants participated as volunteers. Data collection was terminated when the predetermined quota of participants had been reached. Thus, data from 1,028 adult participants (men and women aged 19 years and older) were collected and used in the analysis.

Instrument

The goal of the family system in South Korea is to foster a healthy family using the following four strategies: stability, growth, control, and spirituality. Stability is the most important factor for creating and maintaining a system and for determining the core characteristics of each system. Threats to stability cause instability and conflict in the system, leading potentially to its destruction. Stability is achieved through consistency and system maintenance. Growth and crisis resolution share the same meaning, occurring when a system seeks stimulation, learns, and adapts to changes. This concept explains why the system should be maintained. Growth and stability are complementary, coexisting concepts. Growth is achieved through individualization and change (Friedemann & Webb, 1995).

Control is a factor that influences feelings of anxiety and helplessness. By exercising control, people gain self-confidence and confirm their self-worth. A variety of strategies are employed to control and eliminate uncertainties and threats, including planning, organizing, decision making, training, scheduling, and limiting. If control within one system is impossible, it may be achieved through the intervention of another via system maintenance and changes. In contrast to control, spirituality and affection are compatible concepts. Neither of them are related to religion and are necessary elements for connecting with external systems to form a larger system, which can be achieved through stability and individualization (Friedemann, 1995; see Table 1).

Table 1 - Targets, Dimensions, and Items in the Original Scale Target/Dimension Item Stability  System maintenance 2, 6, 9, 11, 14, 16, 19  Coherence 1, 5, 10, 15, 18 Growth  System change 4, 8, 13, 17, 20  Individuation 3, 7, 12 Control  System maintenance 2, 6, 9, 11, 14, 16, 19  System change 4, 8, 13, 17, 20 Spirituality  Individuation 3, 7, 12  Coherence 1, 5, 10, 15, 18
Translating the scale and verifying its validity

The process of translating the scale and verifying the validity of the content is as follows. First, four experts reviewed the 20 items of the K-ASF-E to verify the relevance and content validity of the scale questions. These experts included a professor from a nursing college with over 20 years of experience, two professors from a medical school, and one individual experienced in tool development with a bachelor's degree in nursing who had majored in nursing. Second, all of the items and options were translated into Korean in a process that considered the unique context of Korean culture. Through this process, all references to the term “my family” in the three response items in the original English version were changed to the term “our family” in the K-ASF-E. In addition, society was specifically defined to include school and workplace settings. Question 17 was modified as follows in consideration of the Korean culture in which family opinions are important when making individual decisions: (1) My family cannot decide anything alone (Option 1); (2) my family makes decisions independently (from Option 3 to Option 2); and (3) When making personal decisions, the coding is changed to obtain the consent of the family (from Option 2 to Option 3). Third, for each item, respondents could select one of three options, with 1 defined as least effective, 2 as fairly effective, and 3 as most effective. The maximum and minimum possible scores on the scale are 60 and 20, respectively. The higher scores are associated with greater family effectiveness. Finally, we obtained ethics approval from the institutional review board before recruiting participants. After the participants signed the consent form, they completed the self-report questionnaires in Korean translated by researchers.

We calculated the content relevance for each item using content validity index (CVI) based on two calculation methods: the item-level CVI (I-CVI) and the scale-level CVI (Polit et al., 2007). We calculated the I-CVI with reading comprehension, conceptual agreement, and accuracy to verify valid sentences. The coefficients for all items were .80 or higher.

The translated 20-item scale was reviewed by experts for sentence expression, item arrangement, and understanding and then pilot tested on adults over the age of 19 years living in South Korea. The pilot test participants evaluated their understanding of each question and related set of answers and were asked to record any difficulties in understanding. The pilot identified no difficulty in understanding or answering the translated questions.

Data Analysis

We analyzed the collected data using IBM SPSS 24.0 (IBM Inc., Armonk, NY, USA), Jamovi 1.6 (The Jamovi Project, 2021), and Mplus 8.4 (Muthén & Muthén, 2019). The EFA and CFA require different samples, so we employed EFA to explore possible factors and CFA to confirm the hypothesized factor structure (Ay et al., 2022; Harmanci Seren et al., 2018; Kahn, 2006). In this study, we randomly divided the collected data and used it to verify the developed scale's cross-validation. For the random assignment of data, we randomly extracted about 60% of the data using the “select cases” function in the SPSS 24.0 program, classified these as Subsample 1, and conducted the EFA. CFA was subsequently conducted on the remaining 40% of the data (Subsample 2; Ay et al., 2022; Park & Bae, 2013). This process allows verification of the cross-validation and suitability of the model by determining how the factor structure constructed through EFA appears in CFA with different subjects.

Before conducting the EFA, to review whether the collected data were suitable for factor analysis, the Kaiser-Meyer-Olkin (KMO) measure and Bartlett's test of sphericity were used to evaluate the sample fit of the 20 items. To conduct EFA, we used the principal component analysis method with varimax rotation, which is the most widely used orthogonal rotation method (Kline, 2016; Lise et al., 2022). For the CFA, the full information maximum likelihood method was used, and the fit of the factor structure model was verified using the root mean square error of approximation (RMSEA ≤ .08), standardized root mean square (SRMR ≤ .08), and comparative fit index (CFI ≥ .09; Kline, 2016).

Ethical Considerations

This study was approved by the institutional review board of Nambu University, and informed consent was provided by all of the participants (IRB No. 1041478-2019-HR-028). We anonymized all of the data used in the analysis. The researcher provided the participants verbal and written information about the study and its objective and assured them of the voluntary nature of their participation as well as the confidentiality and privacy of their data.

Results Characteristics of Subjects

The sociodemographic data for the participants are presented in Table 2. The male-to-female ratio was roughly equal (517 [50.3%] vs. 511 [49.7%], respectively). The average age was 45.1 (SD = 13.6) years. In terms of education, 709 (69.0%) were either attending college or had graduated from college, 171 (16.6%) were high school graduates, and 137 (13.3%) were attending graduate school or had graduated from graduate school. In terms of marital status, 660 (64.2%) were married and 311 (30.3%) were unmarried (Table 2). The divided data for the EFA (Subsample 1) and CFA (Subsample 2) are presented together.

Table 2 - Sociodemographic Status Variable Subsample 1
(n = 605) Subsample 2
(n = 423) Total
(N = 1,028) n % n % n % Gender  Male 310 51.2 207 48.9 517 50.3  Female 295 48.8 216 51.1 511 49.7 Age (years; Mean and SD) 45.31 13.4 44.74 13.9 45.1 13.6  19–39 220 36.4 152 35.9 372 36.2  40–59 261 43.1 194 45.9 455 44.3  ≥ 60 124 20.5 77 18.2 201 19.6 Education  Elementary or middle school graduate 7 1.2 4 0.9 11 1.1  High school graduate 96 15.9 75 17.7 171 16.6  University/college student 24 4.0 27 6.4 51 5.0  University/college graduate 388 64.1 270 63.8 658 64.0  Graduate school student 14 2.3 5 1.2 19 1.8  Graduate school or higher 76 12.6 42 9.9 118 11.5 Marital status  Single 181 29.9 130 30.7 311 30.3  Married 400 66.1 260 61.5 660 64.2  Divorced 18 3.0 18 4.3 36 3.5  Widowed 5 0.8 12 2.8 17 1.7  Separation 1 0.2 1 0.1  Cohabiting 1 0.2 2 0.5 3 0.3
Content Validity

The final version of the K-ASF-E instrument with all items had an I-CVI and scale-level CVI of 1.00.

Construct Validity

In this study, we reviewed the content validity in the process of translating and modifying the original scale and then conducted EFA and CFA to confirm the appropriateness to the Korean context, which assessed construct validity. We verified cross-validation using different cases concurrently in EFA and CFA.

We conducted the EFA on Subsample 1 (n = 605; Ay et al., 2022). The KMO is the variance ratio of the measured variable explained by the principal component, with values closer to 1.0 deemed more suited to the factor analysis. In addition, significance probabilities less than .05 are considered suitable for factor analysis (IBM Knowledge Center, 2022). The KMO of the data collected in this study was .918, which is close to 1, and the Bartlett's sphericity test result was 3,066.986 (df = 190, p < .001). Because of the .001 significance level, the correlation matrix was assessed as unsuitable for factor analysis.

We present the main results of the EFA in Table 3. Specifically, Factor 1 consists of coherence items, with Item 11 added to system maintenance for six items. Factor 2 consists of system maintenance items, with Item 17 added from system change for seven items. Factor 3 consists of individuation items, with Item 20 added to system change for three items. Finally, Factor 4 consists of system change items, with Item 12 added from individuation for four items (see Table 3).

Table 3 - Structure of Exploratory Factor Analysis of the K-ASF-E (Subsample 1, N = 605) Factor No. (Dimension) Item Content Factor C SM I SC Item 11 (SM) → (C) Satisfaction with family decisions .688 Item 15 (C) Expressing one's opinion .676 Item 18 (C) Feeling about being with the family .619 Item 10 (C) Being helpful/not doing enough .613 Item 1 (C) Feeling in the family .546 Item 5 (C) Family being supportive .529 Item 17 (SC) → (SM) Respecting differences .691 Item 19 (SM) Family being flexible to change .621 Item 16 (SM) Putting the family first .617 Item 14 (SM) Learning from problems .582 Item 6 (SM) Understanding each other .570 Item 2 (SM) Agreement about division of labor .503 Item 9 (SM) Solving problems together .496 Item 7 (I) Participating in congregation/neighborhood .664 Item 20 (SC) → (I) Allowing different interests .517 Item 3 (I) Dealing with voluntary organizations .390 Item 13 (SC) Participating in activities .630 Item 4 (SC) Associating with people .629 Item 12 (I) → (SC) Working out disagreements .509 Item 8 (SC) Helping out in the neighborhood .497 Eigenvalue 5.92 1.44 1.28 1.10 Variance (48.68%) 29.57 7.22 6.41 5.48

Note. K-ASF-E = Korean version of the Assessment of Strategies in Families-Effectiveness; C = coherence; SM = system maintenance; I = individuation; SC = system change.

We constructed the process dimensions and items of the K-ASF-E in the EFA. We excluded Item 11 from system maintenance and added it to cohesion. In addition, we removed Items 17 and 11 from system change and system maintenance, respectively, and respectively appended them to system maintenance and coherence. In addition, we removed Item 12 from individuation and added it to system change and then removed Item 20 from system change and added it to individuation. Finally, we excluded Items 17 and 20 from system change and individuation, respectively, and moved Item 12 from individuation to system change.

Confirmatory Factor Analysis

We conducted CFA using Subsample 2 (n = 423) on the factor structure model confirmed through EFA (Ay et al., 2022). According to the results of the CFA centered on the process dimension, the model fit of the factor model was χ2 = 376.439, df = 164, p < .001, RMSEA = .055 (.048–.063), CFI = .897, and SRMR = .051. Although the CFI was slightly below the fit level, it was considered to be in an acceptable range and thus found to be a reasonably suitable model. The χ2/df value was also 2.30, which is less than the standard of 3 (Kline, 2016), indicating that the overall model is acceptable. According to the results of the CFA centered on the target unit, the fit of the factor model was χ2 = 249.391, df = 144, p < .001, RMSEA = .042 (.033–.050), CFI = .949, and SRMR = .036. The χ2/df value was also 1.73, which is less than 3, the standard, indicating the acceptability of the overall model.

Reliability

We used Cronbach's alpha and ICC to indicate and verify the reliability of the final scale (Mehta et al., 2018). We used all of the data (N = 1,028) for the reliability analysis. Cronbach's alpha is the most representative index for evaluating scale reliability, with higher coefficients supporting higher reliability (Kline, 2016). As shown in Table 4, the internal consistency of all 20 items based on Cronbach's α was .86, and the internal consistency of subfactors centered on the target was relatively good, with Cronbach's α = .59–.86. However, in terms of growth among the subfactors, the value of Cronbach's α was .59, which is relatively low.

Table 4 - Reliability of the Korean Version of the Assessment of Strategies in Families-Effectiveness Dimensions and Targets (N = 1,028) Target/Dimension Item Cronbach's α ICC [95% CI] Stability .86 .85 [.84, .87]  System maintenance 2, 6, 9, 14, 16, 17, 19  Coherence 1, 5, 10, 11, 15, 18 Growth .59 .56 [.51, .61]  System change 4, 8, 12, 13  Individuation 3, 7, 20 Control .79 .78 [.76, .80]  System maintenance 2, 6, 9, 14, 16, 17, 19  System change 4, 8, 12, 13 Spirituality .75 .73 [.70, .76]  Individuation 3, 7, 20  Coherence 1, 5, 10, 11, 15, 18

Note. K-ASF-E = Korean version of the Assessment of Strategies in Families-Effectiveness; ICC = intraclass correlation coefficient; CI = confidence interval.

Next, we conducted a reliability analysis of the scale's participants based on the ICC. We obtained this coefficient using a two-way random model based on absolute agreement using the 95% confidence interval. In general, ICC < .50 is “poor,” .50 ≤ ICC < .75 is “moderate,” .75 ≤ ICC < .90 is “good,” and .90 ≤ ICC is “excellent” (Koo & Li, 2016). The ICC for the overall scale was .85 (.84–.87), revealing a good ICC. The ICC ranges for the subfactors centered on the target were .85 (.84–.87) for stability, .78 (.76–.80) for control, .73 (.70–.76) for spirituality, and .56 (.51–.61) for growth. Thus, the ICC values for stability, control, and spirituality were “good,” whereas the ICC value for growth was “moderate.”

Discussion

This study aimed to validate a family health scale developed based on Friedemann's systematic organizational model by modifying and translating it to fit the cultural context of South Korea. To this end, the content validity of the questions was secured through a literature review and expert advice regarding the theoretical background. As a result of this process, the term “my family” in the original questionnaire was altered to “our family.” This change illustrates the necessity of cultural and contextual verification (Friedemann et al., 2003).

Items 11, 12, 17, and 20 were found to differ from the dimensions of the original scale, and the value of the response items of Item 17 was partially adjusted. Finally, Items 7, 4, 6, and 3 were reassigned to system maintenance, system change, coherence, and individuation, respectively. When considering the items where the dimensions differ from those of the original scale, independent personal decisions are presumed for Item 17 in the ASF-E. However, in Korean society, family advice and consent are considered important to maintain familial systems. Item 11 in the ASF-E involves changing the family system by finding a better way for the family to make decisions. System dimensions have been changed or combined in previous studies to reflect different cultural contexts (Gómez Bustamante et al., 2013)

Decision making is viewed as a realm of cohesion and, therefore, a realm of spirituality rather than of control. In the semantic part, aggregation was compared with the other domains, whereas in the case of aggregation, one item was added and no items were excluded. Items 12 and 20 also included the free domain of the family in the control domain through the system change of the original question. However, they were included in the spirituality domain through individualization (Kotlaja, 2020).

As shown, the outward influence of the family on society (school or work) does not belong to spirituality through individualization but to the realm of control through systemic change. These results confirm that Koreans tend to make decisions together, even when decisions are about an individual and perceive this arrangement as enhancing stability rather than imposing control. In contrast, in the West, people tend to feel that systems undergo change and progress to the spiritual stage via individual decision making and freedom. In Korea, cultural differences achieve and control systemic change more via the larger community (societal) context than via the context of the family (Furstenberg, 2019).

Other studies have grouped factors into different categories according to cultural background, target group, or research period. For example, in a recent study in Brazil, Items 5, 19, and 20 differed from the original scale dimensions, and seven, five, four, and four items were used for system maintenance, system change, coherence, and individuation, respectively. This is thought to have been influenced by the fact that 70.5% of the survey participants were married and between the ages of 18 and 60 years, and more than half had more than 11 years of education and were in stable financial circumstances. In addition, the most recent related research study reflects a change in cultural/social development perspective and emphasizes the importance of having a relatively independent disposition (Lise et al., 2022).

In a prior study of Finnish family members receiving medical treatment, six, five, seven, and two items were used for system maintenance, system change, coherence, and individuation, respectively. Individuation component, one of the process dimensions, was grouped with two items. It is the smallest number compared with other studies such as three or more (Åstedt-Kurki et al., 2001). It is believed that the scope of individual activities or independent decision making was limited because the study participants with diseases were protected by family members based on Finnish cultural expectations.

Using the CFA and dividing the scale into a process dimension and a target unit, both models exhibited acceptable model fit index values, and items and factor structures were well aligned with the data. As a result of the internal consistency test and ICC analysis to verify scale reliability, the internal consistency of the overall scale and the ICC were found to be (Cronbach's α) .86 and .86 (.85–.87), respectively, indicating excellent reliability. The results of this study's reliability analysis on Cronbach's alpha and ICC were similar to those of a study on the Brazilian version of the ASF-E (Lise et al., 2022), with the Brazilian version showing good Cronbach's alpha and ICC values for stability, control, and spirituality and relatively low values for growth.

We validated the K-ASF-E tool for the systemic organization framework common among Korean families. Most tools have been developed by combining the highly related variables mentioned in the literature. On the basis of the model, after identifying all of the systems, we identified the central importance of family belonging. In addition, the tool for Korean family effectiveness was reinforced in a validation study using a large-scale stratified sampling method. To validly apply this U.S.-developed tool to Korean subjects, it must be adapted to the Korean cultural context. Finally, to increase the applicability of the scale developed in this study, we suggest conducting a longitudinal evaluation, analysis of measurement invariance, and standard of the scale as well as the results of the repeated study with samples of various regions and age groups. This scale may be further developed into a more standardized and generalized tool by conducting iterative studies, on issues such as standardization work, in the future.

Conclusions

We developed the 20-item K-ASF-E scale with a four-factor structure comprising seven, four, six, and three questions addressing system maintenance, system change, coherence, and individuation, respectively. This is the only tool for family effectiveness that considers cultural differences in Asia, and it may be used for research, interventions, and training programs in community and nursing-related settings.

Acknowledgments

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF-2019R1F1A1058969) and Nambu University, 2023.

Author Contributions

Study conception and design: HC, BP

Data collection: HC