Validation of the Korean version of the death anxiety beliefs and behaviours scale

Abstract

Death anxiety is a complex psychological construct that includes affective, cognitive, and behavioural reactions to the awareness of mortality. The Death Anxiety Beliefs and Behaviours Scale (DABBS) is designed to capture multiple dimensions of death anxiety. This study aimed to adapt the DABBS into Korean and evaluate its psychometric properties in Korean adults. A sample of Korean adults (N = 752) participated, and the data were analyzed to examine reliability, factor structure, and measurement invariance across gender and age groups. The theoretically proposed three-factor structure (affect, beliefs, and behaviours) was supported by factor analyses, providing strong evidence for the reliability and construct validity of the scale. Measurement invariance was supported across gender; however, configural invariance was not supported across age groups, indicating that the scale may not allow for direct comparisons of death anxiety across age groups. This finding may reflect developmental differences in how death anxiety is structured across the lifespan. Overall, the Korean version of the DABBS appears to be a reliable and valid tool for assessing multidimensional death anxiety in Korean adults, with potential applications in both research and clinical settings, although caution is warranted when comparing scores across age groups.

Introduction

The awareness of one’s mortality and the existential fear it evokes constitute a core aspect of human existence. Across millennia, death has remained a central theme across philosophical, religious, and psychological discourse, often approached with a complex blend of dread and reverence [1, 2]. Death anxiety is broadly defined as a negative emotional response triggered by recognizing the inevitable existential reality of death [3], and emotions such as fear, helplessness, and even disgust are frequently associated with the experience of death anxiety, reflecting its multifaceted psychological nature. Death anxiety emerges when individuals become aware of their own finiteness and is regarded as a universal psychological phenomenon that persists across the life span [4].

Within the cultural context of Korean society, death remains a deeply uncomfortable subject, often associated with avoidance and silence. Traditionally, the mention of death has been regarded as inauspicious, leading to a general reluctance to engage in related conversations [5]. Death has historically been understood within a framework shaped by Confucianism, Buddhism, Taoism, and shamanistic beliefs, which conceptualize death as a natural and inevitable part of life rather than a subject for explicit discussion [6]. Despite this philosophical acceptance, contemporary Korean society continues to treat death as a taboo topic, limiting opportunities for open discussion [7]. This avoidance may restrict individuals’ ability to cognitively and emotionally process death-related experiences, thereby contributing to heightened ambiguity and fear.

Cross-cultural evidence further suggests that individuals in Korean contexts are generally more reluctant to engage in death-related conversations compared to those in Western cultures [8], highlighting a culturally embedded barrier that may intensify death-related anxiety. Such cultural characteristics may have differential implications for the components of death anxiety assessed by the DABBS. Limited discussion about death may intensify affective responses by increasing ambiguity and fear [7]. At the cognitive level, insufficient opportunities for reflection may allow maladaptive or exaggerated beliefs about death to persist. At the behavioral level, cultural avoidance of death-related discourse may reinforce avoidance patterns, such as reluctance to engage in conversations about end-of-life planning or funeral arrangements [7]. This sociocultural climate may contribute to elevated levels of death anxiety and hinder adaptive processes of death acceptance. A lack of healthy acceptance of death can have significant adverse implications for psychological well-being. For instance, findings from the validation of the Chinese version of the Death Anxiety Beliefs and Behaviours Scale (DABBS) suggest that distorted beliefs and maladaptive attitudes toward death are associated with elevated anxiety levels and reduced quality of life [9].

Death anxiety extends beyond a transient feeling of fear and has been proposed as a potential transdiagnostic construct associated with a wide spectrum of psychopathological conditions, thereby highlighting its potential clinical significance [10]. Empirical studies indicate that death anxiety constitutes a core cognitive-affective process observed across multiple psychiatric disorders, including panic disorder, obsessive–compulsive disorder, health anxiety, post-traumatic stress disorder (PTSD), and major depressive disorder [11, 12]. In addition to its association with clinical diagnoses, death anxiety has been linked to a range of maladaptive psychological outcomes, such as low self-esteem [13], attachment insecurity [14], a diminished sense of meaning in life [15], and reduced quality of life, particularly among individuals with chronic illnesses [16, 17]. A salient example of the widespread impact of death anxiety was observed during the COVID-19 pandemic, during which heightened death-related fears were found to mediate the development of anxiety, depressive symptoms, and post-traumatic stress responses on a global scale [18, 19]. In this context, death anxiety should be considered not just as an abstract philosophical concern but as a clinically relevant psychological structure.

As the significance of death anxiety has become increasingly evident, various instruments have been developed to quantify and assess death anxiety. Templer’s [20] Death Anxiety Scale (DAS) remains one of the most widely used classical measures, noted for its strong internal reliability and widespread international application, with numerous adaptations and validation studies conducted in Korea. Despite its broad usage, the DAS has been criticized for focusing predominantly on emotional reactions to death, thereby insufficiently capturing death-related cognitive beliefs and avoidance behaviors [21]. To complement this, Templer and colleagues [22] developed the Death Depression Scale (DDS), which was later validated in Korean by An et al. [23]. The DDS was designed to assess depressive affect related to death (e.g., helplessness, feelings of loss, despair). Nevertheless, similar to the DAS, the DDS has been noted to fall short in adequately addressing the cognitive and behavioral components of death anxiety.

In response to such limitations, several instruments targeting death anxiety—such as the Fear of Death Scale (FODS; [24]), the Multidimensional Fear of Death Scale (MFODS; [25]), and the Revised Death Attitude Profile (DAP-R; [26])—have also been created. While these instruments attempt a multidimensional framework, persistent issues have been reported, such as insufficiently addressing aspects related to beliefs and behaviors. Furthermore, there have been issues raised regarding the stability of their factor structures and the suitability of these instruments for various cultural contexts. Zuccala et al. [27], for instance, highlighted that most commonly used assessments of death anxiety were created during the years spanning from the 1950 s to the 1980s. Consequently, many of their items are obsolete and disconnected from current contexts. One item from Templer’s DAS, for example, states, “I shudder when I hear people talk about a World War III” [21], which has faced criticism for inadequately reflecting how individuals experience death anxiety in contemporary populations. This temporal mismatch has been considered a significant concern regarding validity, especially in rapidly changing societies and across non-Western cultural contexts.

Recent research has increasingly emphasized that death anxiety involves more than just emotional reactions such as fear or disgust. It also contributes to the formation of irrational beliefs about death and reinforces avoidant behavioral patterns [21]. This perspective closely resembles the theoretical foundation of Cognitive Behavioral Therapy (CBT), which interprets psychological distress as arising from distorted thoughts and the maladaptive avoidance behaviors that result [28]. Therefore, the main aim of CBT is to recognize and alter these beliefs and behaviors. Empirical research further supports the relevance of this method in clinical settings, with cognitive-behavioral approaches addressing death anxiety showing effectiveness, and evidence indicating that death anxiety may even contribute to treatment resistance across various psychiatric illnesses [1]. As a result, researchers have increasingly emphasized the need for assessment tools that not only evaluate emotional response to death but also examine cognitive and behavioral components as well.

To address this need, Menzies and Sharpe [21] proposed the Death Anxiety Beliefs and Behaviours Scale (DABBS) based on cognitive behavioral therapy (CBT) theory. DABBS is the newly developed measurement that collectively evaluate three aspects of death anxiety: affect, beliefs, and behaviours. The scale has demonstrated its reliability and validity in studies across both clinical and non-clinical populations, and by presenting the clinical cutoff point, the possibility of being used in actual diagnosis and treatment plans was presented. Validation studies of the DABBS have also been conducted in different cultural contexts and have yielded mixed findings. For instance, the Persian version validated among Iranian adolescents generally supported the intended three-factor model, although item-level modification was required to improve model fit [29]. Similarly, the Chinese version validated in community-dwelling older adults retained the original 18-item, three-dimensional structure, while also requiring minor cultural adaptation at the item level [9]. In contrast, the Romanian adaptation did not confirm the original three-factor model and instead yielded a revised five-factor structure, suggesting that the underlying dimensions of death anxiety may be configured differently across cultural contexts [30]. Taken together, these findings indicate that although the DABBS is grounded in a theoretically coherent framework, its factorial structure may not be fully invariant across cultures. This underscores the importance of examining the factor structure and psychometric properties of the DABBS within the Korean context.

To date, a Korean version of the Death Anxiety Beliefs and Behaviours Scale (DABBS) has not yet been developed. Given the cultural tendency in Korea to avoid direct discussions of death, it is essential to adapt and validate the instrument within the local context. Adaptation of a culturally appropriate assessment tool would contribute to more accurately evaluating death anxiety and verifying the effect of psychological interventions, including CBT.

Therefore, this study attempts to verify the reliability and validity of DABBS in the in a sample of Korean adults by adapting and validating it in Korean. The findings are expected to contribute to the development of a culturally appropriate tool for assessing multidimensional death anxiety in Korean clinical and research settings.

Methods

Participants and procedure

An online survey was used to obtain responses from adults across South Korea. Participants were selected using stratified random sampling by age to enhance demographic representativeness.

In order to examine potential age-related differences, participants were categorized into four groups: young adults (19–34 years), middle-aged adults (35–49 years), late middle-aged adults (50–64 years), and older adults (65 years and above). Age groups were defined based on the age classification criteria commonly used in Korea (Korea Institute for Health and Social Affairs, KIHASA), which distinguishes developmental stages across adulthood. A total of 800 participants (200 per group) were initially recruited. However, after applying Mahalanobis distance outlier detection, 48 responses (6.0%) were excluded due to invalid or inconsistent responses. The final analytical dataset included 752 participants. Participants ranged in age from 19 to 84 years (M = 51.13, SD = 14.68). The sample included an equal proportion of males (N = 376, 50.0%) and females (N = 376, 50.0%). Educational levels varied, with the majority holding a university degree. Detailed demographic characteristics are presented in Table 1.

Table 1.

Demographic characteristics of the sample (N = 752)

Variable	Category	n	%
Age (years)	Mean (SD)	51.13(14.68)	-
Gender	Female	376	50.0
	Male	376	50.0
Education Level	Elementary school	1	0.1
	Middle school	4	0.5
	High school	125	16.6
	University (incomplete)	19	2.5
	University graduate	487	64.8
	Graduate school	116	15.4

The present study received ethical approval from the university’s Institutional Review Board (IRB No: 2025—06—013—001). Data was collected using a professional online survey platform, and individuals were requested to review the study purpose and procedures before participating. Thus, participants were informed in advance about the inclusion of sensitive items and proceeded only after their informed consent was obtained. All responses were anonymous, and participants received a reward of 1,000 KRW upon survey completion. The survey required approximately 10 min to complete.

A systematic translation process was adopted before data collection. The original English version was translated by the first author. The back-translation was conducted by a bilingual individual with expertise in psychology and proficiency in both Korean and English. Discrepancies between the back-translated and original versions were carefully compared and reviewed through discussion. Consequently, a clinical psychology expert reviewed the final version to ensure linguistic clarity, cultural appropriateness, and conceptual equivalence.

Before proceeding with the main analyses, a pilot test was performed to assess the comprehensibility and cultural appropriateness of the translated items. A pilot test was conducted with 20 adults, consisting of 12 males and 8 females, with ages ranging from 20 to 80 years (M = 44.15, SD = 17.29). Based on participant feedback, no items were identified as difficult to understand or as causing emotional discomfort. Therefore, no modifications were made, and the final version of the scale was established. No formal cognitive interviewing or committee-based review was conducted beyond the pilot testing procedure.

Measures

Death Anxiety Beliefs and Behaviors Scale (DABBS)

The Death Anxiety Belief and Behaviors Scale (DABBS), developed by [21], assesses maladaptive components of death anxiety within a cognitive-behavioral framework. Each domain corresponds to a distinct component of death anxiety, encompassing affective responses, cognitive-belief processes, and behavioral avoidance patterns. The DABBS is composed of 18 items distributed across three domains: four assessing affect, seven assessing beliefs, and seven assessing behaviours, all rated on a 5-point Likert scale. Higher total scores indicate greater death anxiety, with scores derived from the sum of all item responses. The scale was translated into Korean for use with an adult sample in the current study. The finalized Korean version of the DABBS, including all translated items, is presented in the Appendix. The original DABBS demonstrated excellent internal consistency, with Cronbach’s α = 0.90 for the total scale and Cronbach’s α = 0.83 to 0.94 across subscales.

Death Anxiety Scale (DAS)

The Death Anxiety Scale (DAS), developed by Templer [20], remains among the most widely utilized instruments for assessing death anxiety. The original scale consists of 15 true–false items designed to assess anxiety arising from various death-related experiences, such as the inevitability of death, the dying process, and the afterlife. Several items are reverse-scored, and higher total scores indicate greater death anxiety. The scale has a unidimensional structure, with internal consistency reported at Cronbach’s α = 0.76. The Korean version, translated and validated by Ko, Choi, and Lee [31], was modified into a 5-point Likert-type scale with internal consistency of Cronbach’s α = 0.83, indicating higher reliability than the original scale. In the present study, the DAS demonstrated good internal consistency (Cronbach’s α = 0.87).

Acquired Capability for Suicide–Fearlessness About Death (ACSS-FAD)

To assess reduced fear of death, a key construct in suicide research, the Acquired Capability for Suicide Scale–Fearlessness About Death (ACSS-FAD; [32]) was used. The tool includes seven items assessed on a 5-point Likert scale. Several items are reverse-scored, and total scores are coded such that higher scores indicate greater fear of death. Psychometric evaluations have supported a single-factor structure, with internal consistency estimates reported between Cronbach’s α = 0.77 and 0.83. The Korean version of ACSS-FAD, translated and preliminarily validated by Seo and Kwon [33], confirmed the unidimensional structure of the original scale, with Cronbach’s α = 0.88, indicating strong internal consistency. In the present study, the ACSS-FAD demonstrated good internal consistency (Cronbach’s α = 0.87).

Death Attitude Profile–Revised (DAP-R)

The DAP-R, developed by Wong, Reker, and Gesser (1994), was designed to measure multidimensional attitudes toward death. The original version includes 32 items assessed on a 7-point Likert scale, representing five dimensions: fear of death, death avoidance, neutral acceptance, approach acceptance, and escape acceptance. Several items are reverse-scored, with higher scores reflecting greater levels of the respective death-related attitudes. Reported internal consistency coefficients ranged from Cronbach’s α = 0.65 to 0.97. The Korean version, validated by Oh and An [34], confirmed the five-factor structure through structural equation modeling. In the present study, only the fear of death (7 items) and death avoidance (5 items) subscales were used, capturing emotional anxiety toward death and avoidance of death-related thoughts, respectively. Internal consistency was high, with Cronbach’s α = 0.89 for fear of death and Cronbach’s α = 0.85 for death avoidance. In the present study, the DAP-R demonstrated good internal consistency (Cronbach’s α = 0.86).

Multidimensional Fear of Death Scale (MFODS)

The Multidimensional Fear of Death Scale (MFODS), developed by Hoelter [25], was designed to assess death anxiety from a multidimensional perspective. The original 42-item instrument encompasses eight factors covering a range of fears about self-versus others, process versus outcome, and physical, psychological, and existential dimensions. The items are rated on a 5-point Likert scale ranging from 0 to 4. Reported internal consistency was α = 0.75. The Korean version, validated by Lee, Lee, Ko, and Kim [35], resulted in a 29-item, seven-factor structure through exploratory factor analysis and demonstrated good internal consistency (Cronbach’s α = 0.89). Several items are reverse-scored, and higher scores indicate greater fear of death. In the present study, the MFODS demonstrated good internal consistency (Cronbach’s α = 0.90).

Death Depression Scale (DDS)

The Death Depression Scale (DDS), developed by Templer, Lavoie, and Nightingale [22], was designed to assess depressive affect associated with death, including hopelessness, helplessness, and a sense of loss. The original scale comprises 17 true–false items and has a unidimensional structure, with internal consistency reported as KR-20 = 0.86. The Korean version, validated by An et al. [23], confirmed the unidimensional structure, with internal consistency of KR-20 = 0.84. Several items are reverse-scored, and higher scores indicate greater death-related depressive affect. In the present study, the DDS demonstrated good internal consistency (Cronbach’s α = 0.88).

The Center for Epidemiologic Studies-Depression Scale (CES-D)

Depressive symptoms were examined using the Center for Epidemiologic Studies Depression Scale (CES-D), which was initially created by Radloff [36]. The CES-D is a self-assessment tool aimed at evaluating the occurrence of depressive symptoms experienced over the previous week. The scale comprises 20 items, with each item evaluated on a 4-point Likert scale from 0 (seldom or never) to 3 (often or always). The CES-D has demonstrated high internal consistency in the original study (Cronbach’s α = 0.85 in the general population and 0.90 in clinical samples; [36]). The Korean version, standardized by Jeon et al. [37], demonstrated good internal consistency (Cronbach’s α = 0.91). In the present study, the CES-D demonstrated excellent internal consistency (Cronbach’s α = 0.92).

Statistical analyses

All statistical analyses were conducted using R version 4.4.0. Mahalanobis distance was employed to identify multivariate outliers that could adversely affect model fit due to extreme response patterns. Using a significance level of 0.001, cases exceeding the chi-square critical value of 155.496 were classified as outliers [38]. As a result, 48 cases (6.0%) were excluded, leaving a final analytic sample of 752 participants.

Before conducting the factor analyses, multivariate normality was assessed using Mardia’s test. Given the potential violation of normality, all subsequent analyses, including exploratory factor analysis (EFA), confirmatory factor analysis (CFA), and measurement invariance testing, were conducted using the robust maximum likelihood estimator (MLR), which provides parameter estimates, standard errors, and fit indices that are robust to non-normality.

Exploratory factor analysis (EFA) was initially performed for the subsample (N = 376) using the maximum likelihood estimation method with oblique rotation to investigate the underlying factor structure of the DABBS. Factor loadings of 0.40 or higher were considered meaningful [39], and the number of factors was determined based on scree plot inspection and parallel analysis. Following the factor structure determined by EFA, confirmatory factor analysis (CFA) was carried out to evaluate the suitability of the proposed measurement model.

CFA was conducted on the second subsample (N = 376). Model fit was evaluated using multiple indices: χ² statistics, CFI (Comparative Fit Index), TLI (Tucker–Lewis Index), GFI (Goodness of Fit Index), AGFI (Adjusted GFI), and RMSEA (Root Mean Square Error of Approximation). CFI and TLI values ≥ 0.90 were considered as an acceptable model fit [40]. In addition, SRMR and RMSEA values were evaluated according to established guidelines, with values below 0.08 indicating acceptable fit and values between 0.08 and 0.10 considered indicative of mediocre fit (Browne & Cudeck, 1992).

To determine the relationship between the DABBS and relevant psychological constructs, the conceptual framework of the scale was investigated. The Pearson correlation coefficients were computed between the DABBS subscale scores and established tools that evaluate death-related beliefs and attitudes, which include the Death Anxiety Scale (DAS), the Acquired Capability for Suicide–Fearlessness About Death (ACSS-FAD), the Death Attitude Profile–Revised (DAP-R), the Multidimensional Fear of Death Scale (MFODS), the Death Depression Scale (DDS), and the Center for Epidemiologic Studies Depression Scale (CES-D). The magnitude and direction of the correlations were interpreted in accordance with theoretical models of death anxiety, focusing on the differential associations among affective, cognitive, and behavioral components and related psychological constructs.

Analyses of measurement invariance were also conducted to evaluate if the Korean version of the DABBS operates similarly across age and gender categories. Participants were divided into four age categories (19–34, 35–49, 50–64, and 65 years and above) as well as into male and female categories. A confirmatory factor analysis for multiple groups was conducted to assess measurement invariance among these groups. Measurement invariance was assessed in a sequential manner by evaluating configural invariance, metric invariance, and scalar invariance.

The reliability of the K-DABBS was evaluated using Cronbach’s alpha coefficients. Following conventional criteria [41], α ≥ 0.70 was considered acceptable, α ≥ 0.80 good, and α ≥ 0.90 excellent. Alpha values were calculated for each subscale (Affect, Beliefs, and Behaviours) as well as for the total scale to assess internal consistency.

Results

Reliability of the DABBS

Descriptive statistics and normality indices for the DABBS items are presented in Table 2. While the distributions of scores across the Affect, Beliefs, and Behaviours items showed no substantial deviations from normality at the univariate level, Mardia’s test indicated a violation of multivariate normality (skewness = 2369.24, p < 0.001; kurtosis = 34.55, p < 0.001). Accordingly, subsequent analyses were conducted using the robust maximum likelihood estimator (MLR).

Table 2.

Descriptive Statistics and Normality Indices for DABBS items (N = 752)

Items	M (SD)	Skewness	Kurtosis
1	3.10 (1.01)	−0.23	−0.54
2	2.88 (1.06)	−0.03	−0.75
3	3.06 (1.02)	−0.23	−0.56
4	3.11 (1.04)	−0.27	−0.59
5	2.86 (0.86)	0.12	0.01
6	2.81 (0.95)	0.31	−0.09
7	2.46 (0.90)	0.61	0.36
8	2.66 (0.96)	0.24	−0.30
9	2.78 (0.94)	0.19	−0.05
10	2.66 (0.88)	0.30	−0.02
11	2.94 (0.91)	0.22	−0.16
12	2.47 (0.86)	0.45	0.04
13	2.66 (0.92)	0.32	−0.22
14	2.31 (0.95)	0.67	0.14
15	2.86 (0.97)	0.22	−0.33
16	2.33 (0.95)	0.68	0.26
17	2.59 (0.97)	0.36	−0.21
18	2.44 (0.99)	0.65	0.08

Abbreviation: DABBS Death Anxiety Beliefs and Behaviors Scale

The internal consistency for the total scale was excellent, as indicated by Cronbach’s alpha (α = 0.93) and McDonald’s omega (ω = 0.95). At the subscale level, reliability estimates were also high, with Cronbach’s alpha values of 0.93, 0.88, and 0.91, and corresponding omega coefficients of 0.95, 0.91, and 0.93 for Affect, Beliefs, and Behaviours, respectively.

Factor structure of the DABBS

To conduct an exploratory factor analysis (EFA) and a confirmatory factor analysis (CFA), the entire sample was randomly divided. Prior to conducting the EFA, sampling adequacy was assessed using the Kaiser–Meyer–Olkin (KMO) measure and Bartlett’s test of sphericity. The results indicated that the data were suitable for factor analysis (KMO = 0.92; Bartlett’s test: χ²(153) = 4589.55, p < 0.001), with excellent sampling adequacy. Using the first random subsample (N = 376), parallel analysis and scree plot inspection suggested a three-factor structure. These factors were conceptually consistent with the theoretically proposed dimensions of the DABBS. All items loaded primarily on their intended factors, with standardized factor loadings exceeding 0.40. No substantial cross-loadings exceeding the 0.40 criterion were observed. The EFA model demonstrated an acceptable fit to the data, χ²(102) = 300.30, p < 0.001, CFI = 0.95, TLI = 0.92, and RMSEA = 0.08, 90% CI [0.07, 0.09]. Specific factor loadings are presented in Table 3.

Table 3.

Exploratory Factor Analysis of the DABBS Items (N = 376)

Item	Factor loadings
	1	2	3
1. Feeling anxious about death	.03	.64	.15
2. Being upset by the fact of one’s death	.04	.70	.13
3. Fear of dying	.01	1.00	-.06
4. Finding death frightening	.00	.96	.01
5. Belief that life is too short to experience what one desires	.05	.16	.51
6. Belief that dying alone would be distressing	-.05	.13	.68
7. Belief that one’s death will be painful	.06	.25	.55
8. Belief that losing a loved one would be unbearable	.08	.05	.53
9. Belief of being emotionally devastated by the sudden loss of a loved one	.04	-.08	.72
10. Belief of being unable to face death courageously	.02	.16	.67
11. Belief that a loved one’s severe illness would be unbearable	.03	-.15	.76
12. Avoiding death-related news or media	.66	.02	.08
13. Avoiding imagine of being diagnosed with a terminal illness	.48	.10	.26
14. Avoiding novels involving death	.87	-.08	.02
15. Avoiding reminders of a loved one’s death	.60	-.01	.22
16. Avoiding movies or dramas about death	.88	-.01	-.09
17. Avoiding imagining one’s own death	.56	.24	.13
18. Avoiding reading about terminally ill patients	.79	.06	-.09

Bold values indicate factor loadings ≥ .40

In the EFA, Item 3 (“Fear of dying”) showed a standardized loading of 1.00. To examine this result, additional diagnostic analyses were conducted. Item 3 was highly correlated with Item 4 (“Finding death frightening”) (r = 0.93), suggesting substantial overlap between the two items. Importantly, the residual variance of Item 3 remained positive (θ = 0.056), indicating that this was not a Heywood case. A sensitivity analysis was conducted by removing Item 3, and the three-factor structure remained stable with comparable model fit indices (CFI = 0.95, TLI = 0.92, RMSEA = 0.08, 90% CI [0.07, 0.09], SRMR = 0.03). These findings suggest that the overall factor structure was not dependent on this item.

Based on the factor structure identified in the EFA, CFA was conducted on the second subsample (N = 376) to evaluate the adequacy of the three-factor measurement model. The model demonstrated an acceptable to marginal fit to the data, χ² (132) = 453.16, p < 0.001, CFI = 0.92, TLI = 0.91, RMSEA = 0.09, 90% CI [0.08, 0.10], and SRMR = 0.07, with RMSEA indicating relatively higher misfit compared to other indices. Each item loaded significantly on its respective factors, with standardized factor loadings ranging from 0.66 to 0.97. These results indicate that the three-factor structure of the DABBS is structurally valid. In addition, the average variance extracted (AVE) values were 0.79 for Affect, 0.55 for Beliefs, and 0.59 for Behaviours, and the composite reliability (CR) values exceeded 0.89 for all factors, supporting convergent validity. Figure 1 presents the three-factor model estimated based on the split sample (N = 376).

Fig. 1.

3-factor model of the DABBS (N = 376). Straight arrow, standardized regression weights, curved arrows, correlations between factors

Invariance testing

Measurement invariance across gender

MG-CFA was used to examine measurement invariance across gender. The suggested factor structure is comparable across gender groups, according to the configural invariance model's acceptable to marginal model fit (CFI = 0.918, TLI = 0.905, RMSEA = 0.088, 90% CI [0.082, 0.094], SRMR = 0.067). Then, models of scalar and metric invariance were tested. In comparison to the configural model, the changes were △CFI = −0.001 and △CFI = −0.004, respectively, meeting the △CFI ≤ 0.01 criterion. This implies that latent variable means can be compared between genders because factor loadings and item intercepts function similarly across gender groups. Detailed results of measurement invariance testing across gender are presented in Table 4.

Table 4.

Results of measurement invariance testing across gender

	CFI	TLI	RMSEA	SRMR	ΔCFI
Configural Invariance	.918	.905	.088	.067	-
Metric Invariance	.918	.910	.086	.069	-.001
Scalar Invariance	.913	.910	.086	.070	-.004
Strict Invariance	.912	.914	.084	.070	-.001

CFI comparative fit index, TLI Tucker-Lewis index, RMSEA root mean square error of approximation, SRMR standardized root mean square residual

Measurement invariance across age groups

The study then examined measurement invariance across age groups. The configural invariance model did not demonstrate adequate fit to the data, with TLI values below 0.90 and RMSEA exceeding 0.09. These findings suggest that the factor structure may not be equivalent across age groups. Since configural invariance was not established, further tests of metric and scalar invariance were not conducted, and direct comparisons of latent variable levels across age groups were not performed. Detailed results of the configural invariance model are presented in Table 5.

Table 5.

Results of measurement invariance testing across age groups

	CFI	TLI	RMSEA	SRMR
Configural Invariance	.908	.894	.093	.072

CFI comparative fit index, TLI Tucker-Lewis index, RMSEA root mean square error of approximation, SRMR standardized root mean square residual

Given the lack of configural invariance across age groups, exploratory factor analyses were conducted separately for each age group to further examine potential differences in factor structure.

In the young adult group, most items loaded in accordance with the theoretically proposed factor structure. However, a small number of items showed cross-loadings across two factors. In particular, item Q17 demonstrated a marginal cross-loading between factors. Overall, the factor structure was largely consistent with the theoretical model.

For the middle-aged group, most items loaded in accordance with the theoretically proposed factor structure. However, some items originally belonging to the belief factor showed loadings on the affective factor or exhibited cross-loadings. Despite these variations, the overall factor structure remained largely consistent with the theoretical model.

In late middle-aged and older adults, no significant cross-loadings were observed. The majority of items exhibited moderate to strong loadings on theoretically hypothesized factors.

Correlations among the DABBS subscales and with other measures

Inter-factor correlations among the three DABBS subscales were examined. The Affect subscale was moderately correlated with Beliefs (r = 0.62) and Behaviours (r = 0.50), while Beliefs was also moderately correlated with Behaviours (r = 0.57). These results indicate that the three factors are related but distinct constructs.

The DABBS total score showed strong positive correlations with other death-related measures, including the DAS (r = 0.74), DAP-R (r = 0.75), ACSS-FAD (r = 0.71), MFODS (r = 0.70), and DDS (r = 0.63), all ps < 0.001. Notably, ACSS-FAD scores were coded such that higher scores reflected greater fear of death. In addition, the DABBS total score was moderately correlated with CES-D (r = 0.35), indicating an association with general depressive symptoms.

At the subscale level, all three dimensions of the DABBS were significantly correlated with other measures. The Affect subscale showed relatively strong correlations with death-related variables (e.g., ACSS-FAD: r = 0.73; DAS: r = 0.72; DAP-R: r = 0.70), whereas the Beliefs and Behaviours subscales demonstrated moderate to strong correlations across measures (rs = 0.45–0.65).

In addition, Similar magnitudes of association with CES-D were observed across other death-related measures, including DAS (r = 0.29), DAP-R (r = 0.36), MFODS (r = 0.32), DDS (r = 0.30), and ACSS-FAD (r = 0.29), indicating that the relationship with depressive symptoms is not unique to the DABBS. Detailed correlation coefficients are presented in Tables 6 and 7.

Table 6.

Correlation between the DABBS and other measures (N = 752)

	1	2	3	4	5	6	7
1. DABBS	1.00
2. DAS	.74	1.00
3. DAP-R	.75	.77	1.00
4. ACSS-FAD	.71	.84	.79	1.00
5. MFODS	.70	.78	.73	.72	1.00
6. DDS	.63	.66	.69	.66	.67	1.00
7. CES-D	.35	.29	.36	.29	.32	.30	1.00

All correlations were significant at p <.001

DABBS Death Anxiety Beliefs and Behaviours Scale, DAS Death Anxiety Scale, DAP-R Death Attitude Profile–Revised, ACSS-FAD Acquired Capability for Suicide Scale–Fearlessness About Death, MFODS Multidimensional Fear of Death Scale, DDS Death Depression Scale, CES-D Center for Epidemiologic Studies Depression Scale

Table 7.

Correlations between DABBS subscales and other measures (N = 752)

	Affect	Beliefs	Behaviours
DAS	.72	.65	.52
DAP-R	.70	.65	.56
ACSS-FAD	.73	.61	.50
MFODS	.62	.63	.53
DDS	.60	.56	.45
CES-D	.24	.40	.25

All correlations were significant at p <.001

DABBS Death Anxiety Beliefs and Behaviours Scale, DAS Death Anxiety Scale, DAP-R Death Attitude Profile–Revised, ACSS-FAD Acquired Capability for Suicide Scale–Fearlessness About Death, MFODS Multidimensional Fear of Death Scale, DDS Death Depression Scale, CES-D Center for Epidemiologic Studies Depression Scale

Discussion

This study was conducted to verify the psychometric validity of the Korean version of the Death Anxiety Beliefs and Behaviours Scale (K-DABBS) and to confirm the measurement invariance according to gender and age.

The factor structure of the K-DABBS was found to be theoretically coherent, and its internal consistency and construct validity were supported by correlations with established death-related scales. The high loading of Item 3 likely reflects its conceptual overlap with Item 4, particularly the close semantic relationship between fear of dying and fear of death, rather than a model estimation issue. These results suggest that K-DABBS is a reliable and valid tool to evaluate death anxiety in multiple dimensions in Korean adults. Strong reliability estimates at both the total-score and subscale levels support the multidimensional assessment of death anxiety provided by the K-DABBS. Overall, the findings support the applicability of the K-DABBS for use in both clinical settings and research contexts.

The K-DABBS showed a significant positive correlation with the CES-D, indicating that death anxiety is meaningfully associated with general depressive symptoms. This finding is consistent with previous research suggesting that heightened awareness of mortality and death-related concerns are linked to broader emotional distress [11, 42].

However, the magnitude of the correlations between the K-DABBS and CES-D was moderate rather than strong, suggesting that death anxiety and depression, while related, are not interchangeable constructs. Importantly, similar levels of association with the CES-D were observed across other death-related measures, including the DAS, DAP-R, MFODS, DDS, and ACSS-FAD. This pattern indicates that the relationship with depressive symptoms is not unique to the K-DABBS but reflects a general connection between death-related psychological constructs and negative affect.

K-DABBS were found to have an identical factor structure and item functions across men and women. Although previous research has reported gender differences in the level and specific aspects of death anxiety [43], the present findings suggest that the underlying structural organization of death anxiety is comparable across gender.

In contrast, configural invariance across age groups was not supported. This indicates that the K-DABBS may not currently support direct comparisons of death anxiety across age groups and suggests that death anxiety's fundamental factor structure varies throughout the lifespan. The absence of configural invariance does not necessarily indicate a limitation of the K-DABBS; rather, it may reflect developmental variation in death anxiety as the construct changes alongside shifts in social roles, life experiences, and perceived proximity to death [44]. Differences in factor structure may be attributable to changes in the psychological meaning and salience of death-related thoughts, feelings, and behaviors as individuals progress through different stages of adulthood. To account for age-related variation, the EFA findings are interpreted within a developmental context, with comparisons drawn across young, middle-aged, late middle-aged, and older adults.

In the young adult group, item Q17 (“imagining one’s own death”), which requires concrete and vivid mental imagery, demonstrated a marginal cross-loading on both the behavioural and affective factors. Unlike more abstract or indirect death-related items, this item prompts participants to vividly simulate their own mortality, thereby intensifying mortality salience. This pattern can be interpreted within the framework of Terror Management Theory [45, 46]. The activation of self-referential mortality-related thoughts may elicit existential threat, leading individuals to simultaneously engage in belief-level appraisals and avoidance-oriented behavioural responses [45]. As a result, cognitive–affective evaluations and behavioural tendencies may be concurrently activated, producing an overlap between components. Consistent with this interpretation, young adults may be more likely to blur the distinction between appraisal and behaviour when confronted with vivid death imagery, whereas more abstract stimuli tend to remain associated with the behavioural domain. Furthermore, the emotional impact of death-related thoughts depends not merely on their occurrence, but on how such thoughts are interpreted and evaluated [47]. Taken together, the cross-loading observed in Q17 is unlikely to reflect measurement error, but rather a developmentally meaningful feature of young adults’ death anxiety, in which vivid, self-referential imagery simultaneously activates affective responses and behavioural tendencies. Importantly, this pattern was limited to a single item and did not undermine the overall factor structure.

In the middle-age group, items assessing beliefs related to one's own death tended to load more strongly on the affective factor. Specifically, items related to fear, pain, and emotional difficulty in facing one's own death Q7 (“My death will be painful”) and Q10 (“I will not be able to face death courageously”) demonstrated either stronger loadings on the affective factor or cross-loadings between affect and belief. In contrast, belief items concerning the death of others and relational loss (Q8, 9, and 11) remained more clearly associated with the belief factor. These findings indicate that beliefs about others’ death may remain relatively more cognitively structured, whereas beliefs about one’s own death become more closely intertwined with affective responses. This pattern can be interpreted within the framework of Socioemotional Selectivity Theory, which suggests that motivational priorities shift as perceived future time becomes more limited [48, 49]. As individuals enter midlife and begin to perceive their remaining time as more limited, death becomes a more personally relevant and imminent concern. Consequently, cognitive appraisals of one’s own mortality may become increasingly emotionally salient, resulting in a reduced differentiation between belief and affect. Importantly, this pattern does not indicate measurement instability, but rather reflects a developmentally meaningful reorganization of death anxiety. The partial convergence of belief and affect in middle adulthood may represent a transitional phase in which mortality is no longer processed solely as an abstract concept, but is increasingly experienced as an emotionally significant reality.

A significant level of clarity and stability of factor structure was exhibited in late middle-aged and older adults. The items in both groups showed consistent loading onto the theoretically organized factors. It demonstrates the well-differentiated organization of affective, belief, and behavioral components of death anxiety. It could be understood as this stable factor structure may represent a stage in which death is acknowledged as an integral part of life and is more directly integrated into one's life narrative. Previous research has suggested that older adults tend to show greater acceptance of death and a more coherent organization of death-related attitudes compared to younger age groups [26, 44]. In line with previous studies and distinctive patterns of factor loadings found in this group, the result could be interpreted as developmental integration rather than instability in measurement.

Limitations

Despite its contributions, the present study is subject to several limitations. First, the sample used in this study is solely on a general adult population, which may have limited applicability to clinical populations, adolescents, or the elderly. In particular, the result of this study found that death anxiety may vary in structure and expression depending on the developmental stage or clinical condition. Therefore, it is recommended for future research to re-examine the factor structure and validity of the DABBS across different population groups. Second, capturing death anxiety through a cross-sectional approach limits insight into its persistence and developmental change across time. The cross-sectional design precluded examination of causal relationships and temporal stability, and the use of self-report measures may have elicited response bias related to social desirability. It is suggested for future researchers to include interviews, clinical evaluations, and behavioral observations. Furthermore, test–retest reliability was not evaluated in this study, limiting the ability to assess the temporal stability of the scale. A further limitation is that although the translation process involved back-translation and pilot testing, it did not include formal committee review or cognitive interviewing procedures, which may have further enhanced the linguistic and cultural equivalence of the scale. Additionally, although the present study considered cultural aspects of death anxiety within the Korean context, the sample consisted of a general adult population and may not fully capture variability across different cultural subgroups. Individuals from more traditional or rural backgrounds, as well as those with differing levels of acculturation, may exhibit distinct patterns of death-related beliefs and behaviors. Future research is needed to examine how such cultural and contextual factors may influence the structure and expression of death anxiety. Future research may also benefit from testing alternative structural models, such as bifactor models, to further examine the dimensionality of the DABBS.

Conclusion

Although some limitations of the DABBS were highlighted in this study, it demonstrated its significant role in evaluating death anxiety. The scale showed psychometric robustness, enabling a multidimensional evaluation of death anxiety across affective, cognitive, and behavioral domains. The DABBS is anticipated to be useful in advanced research and clinical settings.

Abbreviations

DABBS

Death Anxiety Beliefs and Behaviours Scale

DAS

Death Anxiety Scale

ACSS-FAD

Acquired Capability for Suicide–Fearlessness About Death

DAP-R

Death Attitude Profile–Revised

MFODS

Multidimensional Fear of Death Scale

DDS

Death Depression Scale

CES-D

The Center for Epidemiologic Studies-Depression Scale

Appendix

The Korean version of the DABBS

다음은 귀하가 죽음에 대해 가지고 있을 수 있는 생각이나 감정을 알아보기 위한 문항입니다. 각 문항을 읽고 자신의 경험을 가장 잘 반영한다고 생각되는 숫자에 표시해주시기 바랍니다.

번호	문항	전혀 동의하지 않는다	동의하지 않는다	보통이다	동의한다	매우 동의한다
1	나는 죽음에 대해 불안을 느낀다	①	②	③	④	⑤
2	언젠가 내가 죽는다는 사실은 끔찍하다	①	②	③	④	⑤
3	나는 죽는 것이 두렵다	①	②	③	④	⑤
4	나는 죽음이 두렵다	①	②	③	④	⑤

다음은 귀하가 죽음에 대해 가지고 있을 수 있는 믿음이나 신념을 알아보기 위한 문항입니다. 각 문항을 읽고 자신의 경험을 가장 잘 반영한다고 생각되는 숫자에 표시해주시기 바랍니다.

번호	문항	전혀 생각하지 않는다	거의 생각하지 않는다	가끔 생각한다	자주 생각한다	항상 생각한다
5	원하는 것을 경험할 시간이 충분하지 않다는 것은 끔찍하다	①	②	③	④	⑤
6	혼자 죽음을 맞이한다면 끔찍할 것이다	①	②	③	④	⑤
7	나의 죽음은 고통스러운 경험일 것 같다	①	②	③	④	⑤
8	사랑하는 사람과 함께가 아니라면, 나이 들어가는 것을 감당할 수 없을 것 같다	①	②	③	④	⑤
9	사랑하는 사람을 갑작스럽게 잃게 될 것이고, 그로 인해 무너질 것이다	①	②	③	④	⑤
10	죽음을 앞두고, 용감하게 죽음을 마주하지 못할 것 같다	①	②	③	④	⑤
11	내가 아끼는 사람이 불치병에 걸린다면 감당하지 못할 것 같다	①	②	③	④	⑤

다음은 귀하가 죽음에 관련된 행동들을 얼마나 회피하는지 알아보기 위한 문항입니다. 각 문항을 읽고 자신의 경험을 가장 잘 반영한다고 생각되는 숫자에 표시해주시기 바랍니다.

번호	문항	전혀 회피하지 않는다	거의 회피하지 않는다	가끔 회피한다	자주 회피한다	항상 회피한다
12	죽음에 관한 뉴스나 기사, 또는 다큐멘터리 등을 접하는 것	①	②	③	④	⑤
13	불치병 진단을 받는 것을 떠올리는 것	①	②	③	④	⑤
14	죽음을 앞둔 인물이 나오는 소설을 읽는 것	①	②	③	④	⑤
15	사랑하는 사람의 죽음을 떠올리는 것	①	②	③	④	⑤
16	등장인물의 죽음을 앞둔 영화나 드라마를 시청하는 것	①	②	③	④	⑤
17	나의 죽음을 상상하는 것	①	②	③	④	⑤
18	불치병 환자가 쓴 수필이나 회고록을 읽는 것	①	②	③	④	⑤

Authors’ contributions

SP led the conception and design of the study, carried out all statistical analyses, and wrote the first draft of the manuscript. JS provided academic guidance on the study design and interpretation of the results and contributed to the critical revision of the manuscript. All authors reviewed and approved the final version of the manuscript.

Funding

This research was supported by the Brain Korea 21 (BK21) Program.

Data availability

The data supporting the findings of this study are not publicly accessible due to ethical considerations and the need to protect participant confidentiality. However, the datasets may be made available by the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The present study received ethical approval from the Institutional Review Board of Jeonbuk National University (IRB No. 2025—06—013—001). All research procedures involving human participants were carried out in compliance with the ethical standards of the institutional research committee and the principles outlined in the Declaration of Helsinki. Written informed consent was obtained from all participants prior to their inclusion in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.