Sleep cognition among adolescents: adaptation of the dysfunctional beliefs and attitudes about sleep-16 (DBAS-16) scale in Indonesian

Abstract

Background

Sleep is associated with psychological factors such as cognition, and negative sleep cognition is related to dysfunctional beliefs and attitudes. Tool is needed to understand sleep-related to cognition to prevent insomnia. The Dysfunctional Beliefs and Attitudes about Sleep-16 (DBAS-16) scale had not been previously translated into Indonesian, and the construct validity had not been tested among Indonesian adolescents. The present study translated the DBAS-16 scale into Indonesian and assessed the construct validity of the Indonesian version of the DBAS-16 scale.

Method

This study employed a cross-sectional and correlational design through a self-reported questionnaire. Data were collected from 1191 adolescents, aged between 13 and 18 in Indonesia. In this study, validity evidence based on test content was obtained through forward translation, expert review, back-translation, and pretesting procedures. For validity evidence based on internal structure, Exploratory Factor Analysis was used to assess factor structure and Confirmatory Factor Analysis was conducted to evaluate the structural model fit of the original DBAS-16 scale and the Indonesian version of the DBAS-16 scale. Validity evidence based on relations to other variables was evaluated by examining the association between DBAS-16 scale scores and depressive symptoms measured using the Beck Depression Inventory-II (BDI-II). The Cronbach's alpha and McDonald's omega coefficients were measured to examine the internal consistency reliability.

Results

The Kaiser–Meyer–Olkin value was 0.893, and the Bartlett's test of sphericity was statistically significant and adequate for Exploratory Factor Analysis. Confirmatory Factor Analysis tested the second-order model with three first-order factors and showed good model fit. The total score of DBAS-16 scale was significantly correlated with the total score of BDI II (r = 0.165, p < 0.05). The Cronbach's alpha and McDonald's omega of the Indonesian version of the DBAS-16 scale was 0.83.

Conclusions

The Indonesian version of the Dysfunctional Beliefs and Attitudes about Sleep-16 (DBAS-16) scale supported validity and reliability for evaluating sleep-related cognition in Indonesia. This implies that is an appropriate tool for cognition of sleep screening among Indonesian adolescents.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s40359-026-05213-x.

Introduction

Sleep habits are easily affected by the physical and social development in adolescence [15, 16]. Sleep duration and adequate sleep among adolescents have decreased over the past 20 years [10]. In a study, the prevalence of insomnia among adolescents between 18.7%—20.8% [22, 33]. 33% −46% of adolescents experience sleep restriction [47]. The prevalence of sleep disturbance is 83%, whereas the behavior of sleep especially in cognitive domain showed less category (71.74%) in Indonesia [49, 54]. Sleep is crucial for adolescent health [1], a study reported that poor sleep affects the mental and physical well-being of adolescents [60, 68].

Sleep is associated with psychological factors such as cognition and negative sleep cognition is related to dysfunctional beliefs and attitudes [24]. Dysfunctional beliefs and attitudes about sleep include faulty beliefs, anxiety, attention bias about the effects of sleep loss, and unrealistic expectations about sleep [32]. Furthermore, people with dysfunctional beliefs and attitude of sleep are unrealistic about the amount of sleep and worry regarding the negative effect of insomnia [32]. These conditions may induce emotional harm [40], insomnia [46], and depression [62]. Studies on etiological models for insomnia have emphasized the influence of dysfunctional sleep beliefs and attitudes on this disorder [61].

Several instruments have been developed to assess sleep-related cognitions. For example, the Anxiety and Preoccupation about Sleep Questionnaire (APSQ) focuses primarily on sleep-related worry [30], whereas the Sleep-Associated Monitoring Index (SAMI) assesses worry about sleep, arousal, and distress [53]. The Dysfunctional Beliefs and Attitudes about Sleep-16 scale (DBAS-16) was selected because it is one of the most widely used instruments for assessing dysfunctional sleep-related cognition [66]. The DBAS-16 scale is a valid and reliable instrument with adequate internal consistency (Cronbach’s alpha: 0.77–0.79) and 2-week test–retest stability with a Pearson correlation coefficient (r) of 0.83 [45]. It measures four factor structures: (1) the perceived consequences of insomnia, (2) worrying about insomnia, sleep expectation, and medication [45]. The DBAS-16 scale is derived from the long version of the DBAS, which tests 30 items, [44] and is more user-friendly than the original version [66]. The DBAS-16 scale has been translated into languages such as Turkish [8], German [38], and Hindi [18], and has favorable reliability and validity. Therefore, the DBAS-16 scale can evaluate sleep-related cognition across general practice good sleepers [38] and clinical populations [45, 66].

Sleep-related cognition must be evaluated before cognitive behavioral therapy [66] and sleep research is employed [38]. Understanding about sleep-related cognition can prevent insomnia [5]. Hence, it was needed tool to measure sleep-related cognition and it is also used for clinical and research purposes of rapid screening on a large scale. To the best our knowledge the DBAS-16 scale had not been previously translated into Indonesian, and the construct validity had not been tested among Indonesian adolescents. Given this background, the present study translated the DBAS-16 scale into Indonesian and assessed the construct validity of the Indonesian version of the DBAS-16 scale.

Methods

Participants and settings

This study used a cross-sectional and correlational design through a self-reported questionnaire. Data were collected from 1191 adolescents in Indonesia, aged between 13 and 18 years. General guidelines for factor analysis suggest having at least 300 participants [59]. We included participants who lived in dormitories and had no history of psychiatric or neurological disorders. The sample comprised male and female adolescents from several regions of Indonesia, predominantly from the western part of the country. Majority of participants resided in dormitory-based settings, with most sharing room with their friend. Detailed demographic characteristics are presented in Table 1.

Table 1.

Characteristics of participants (n = 1191)

Characteristics	Mean (SD)	n	%
Age (years)	16.10 (0.95)
13–15		340	28.5
16–18		851	71.5
Sex
Female		792	66.5
Male		399	33.5
Area
West Indonesia		1105	92.77
Central Indonesia		79	6.63
East Indonesia		7	0.6
Sleep in a Bed
Yes		952	79.9
No		239	20.1
Share a bed with another person
Yes		783	65.7
No		408	34.3
Sleep in a room
Alone		9	0.76
≤ 20		909	76.32
> 20		273	22.92
Have a pillow
Yes		969	81.4
No		222	18.6
Have a bolster
Yes		615	51.6
No		576	48.4
Having a blanket
Yes		675	43.3
No		516	56.7
Regular Exercise over a Week
Once		65	5.5
More than once		1126	94.5
Illness in the previous semester
Yes		645	54.2
No		546	45.8

SD standard deviation

All adolescents provided informed consent for participation in this survey. For participants under the age of 16 years, informed consent was obtained from their legal guardians, as these adolescents resided in dormitories and were under guardian supervision. Participation was voluntary, and confidentiality of the data was ensured throughout the study.

Instruments

Characteristics of the participants were first obtained from a questionnaire, including age, sex, residential area, and the environment of sleep such as whether they sleep on the bed and have a pillow, bolster, blanket [39], or a partner in one room [31]. The questionnaire also includes factors that predict good sleep, including missed classes [4] and exercise [48].

BDI-II

The Beck Depression Inventory-II (BDI)-II. The BDI-II was developed by Aaron-T. Beck [6]. This questionnaire also provides in Indonesia version [23]. We also assessed the participants using this questionnaire. The BDI-II is a valid and reliable questionnaire, it has 21 items, each score from 0 to 3. Total score of BDI-II is 63, the cut off score of Indonesia BDI-II is 17. The Cronbach alpha of BDI-II in this study was 0.82.

DBAS-16 Scale

The DBAS-16 scale was created by Charles M. Morin [45]. It measures, identifies, and assesses sleep-related cognitions. The factor structure of DBAS-16 is similar to the original four dimensions of the DBAS-30: (1) perceived consequences of insomnia (Items 5, 7, 9, 12, and 16), (2) worry and helplessness about insomnia (3, 4, 8, 10, 11, and 14) [9], sleep expectation (1 and 2), and (Sunderland et al.) medication (6, 13, and 15). It applies a 10-point scale (0: strongly disagree,10: strongly agree), and a higher score indicates more dysfunctional beliefs and attitudes about sleep [45].

Previous study mentioned that The DBAS-16 scale was found to be reliable, having a Cronbach’s alpha of 0.77 for clinical setting, 0.79 for research samples and temporal stability (r = 0.83) [45]. Studies in other countries and on other languages have demonstrated adequate Cronbach’s alpha values [Hindi (0.91) [18], German (0.71) [38], and Turkish (0.82) [8]]. Overall, the Cronbach’s alpha of DBAS-16 scale showed good reliability. We selected the DBAS-16 scale because it is a brief instrument and is easy to complete and score.

Data collection and procedures

All procedures were approved by the Ethics Committee. Data were collected from August to September 2014. Permission to use the DBAS-16 scale was granted by its original authors. The English version of the DBAS was translated and adapted into Indonesian following the procedure of the World Health Organization, because they provide a rigorous and internationally recognized framework for achieving semantic, conceptual, and cultural equivalence across languages and cultures [64, 67]. The English version of the DBAS was translated into Indonesian by a translator who is a native speaker of Indonesian, fluent in English, and familiar with the scale’s terminology. We then conducted an expert panel by a mental health nurse and community nurse to evaluate their clarity, relevance, and cultural appropriateness of DBAS-16 scale in Indonesia. The expert panel confirmed that the translated items were clear, culturally appropriate, and consistent with the original meaning of the DBAS-16 scale. No items were removed because all items were considered relevant to assess dysfunctional beliefs and attitudes about sleep. The results of expert panel were translated back to English by a different professional translator who had not been involved in the forward translation process and was blinded to the original version of the instrument. Before providing a final version, we tested the DBAS-16 scale on 10 adolescents to assess the understandability of the questionnaire. The participants were allowed to withdraw from the study after seeing the questionnaires.

Data analysis

All analyses used SPSS for Windows [27]. Descriptive statistics (means, standard deviations, skewness, and kurtosis) were applied in calculations for all DBAS-16 items. Skewness and kurtosis were employed to assess the normality of data [35]. Skewness and kurtosis were used to measure the asymmetry and peakedness of data distribution, respectively [11, 35]. The acceptable range of skewness is from − 1 to 1 [11]. Similarly, the acceptable range of kurtosis is from − 1 to 1 [11, 37].

Validity was evaluated following the Standards for Educational and Psychological Testing [20], which conceptualize validity as a unitary construct supported by multiple sources of evidence. In this study, validity evidence based on test content was obtained through forward translation, expert review, back-translation, and pretesting procedures. Validity evidence based on internal structure was examined using exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Furthermore, validity evidence based on relations to other variables was evaluated by examining the association between DBAS-16 scores and depressive symptoms measured using the Beck Depression Inventory-II (BDI-II) (Fig. 3).

Fig. 3.

Conceptual Framework of Multi-Trait Mono-Method (MTMM) Approach Validity Evidence Based on Relations to Other Variables

Confirmatory factor analysis (CFA) was initially conducted to examine whether the factor structure of the original DBAS-16 could be replicated in the Indonesian adolescent sample. CFA was performed using AMOS version 21.0 with the maximum likelihood (ML) estimator. The original factor structure was tested first before exploratory factor analysis (EFA), followed by comparison of alternative models, including a single-order factor model and a second-order factor model. Confirmatory factor analysis (CFA) was used to assess the structural model fit of the DBAS-16. We used AMOS version 21.0 to test the goodness of fit and used same sample with EFA [13, 21]. The initial and last CFA was performed using the Maximum Likelihood (ML) estimator. The following fit indices were evaluated by chi-square/df (the ratio of the chi-square to the degree of freedom) and p value, however p value is sensitive to large sample size, therefore, the Chi-square test divided by its degree of freedom. The references recommended that χ²/df ≤ 5 is acceptable fit and χ²/df ≤ 3 is good fit [36, 55]. Further, acceptable values indicated in parentheses: normed fit index (NFI > 0.9), goodness of fit index (GFI,> 0.9), adjusted goodness of fit index (AGFI > 0.9), comparative fit index (CFI>0.9), related fit index (RFI > 0.9)), and root mean square error of approximation (RMSEA < 0.08) [26]. A second-order model was used to assess validity. The second-order model was adopted because the DBAS-16 scale consists of multiple related dimensions that represent a broader latent construct. A second-order model allows these first-order factors to be explained by a general factor and provides a more parsimonious representation of the construct [52] and to estimate the effect of the main construct on subconstructs [2].

An exploratory factor analysis (EFA) with principal components extraction and varimax rotation was conducted to evaluate construct validity [29]. Several tests should be conducted to assess data for suitable factor analysis, including the Kaiser–Meyer–Olkin (KMO) and Bartlett’s test of sphericity [65]. The KMO test of sampling adequacy must be > 0.5 and Bartlett’s test of sphericity should be significant (p < 0.05) [34, 58]. The criteria used to determine factor extraction were as follows: an eigenvalue of > 1 [34] and the percentage of variance explained of ≤ 50% or ≤ 60% [50]. Convergent validity of the DBAS-16 was evaluated using factor loadings. Loading variables must be > 0.32. Standardized factor loadings of ≥ 0.32, ≥ 0.55, ≥ 0.63, and ≥ 0.71 are considered to be fair, good, very good, and excellent, respectively [14]. EFA was conducted to examine whether the translated and culturally adapted Indonesian version demonstrated a comparable underlying dimensional structure. Similar to previous cross-cultural validation studies, EFA was employed as an initial step to assess whether the latent structure of the original instrument could be replicated in the target population following translation and cultural adaptation [7, 56].

Validity evidence was used to evaluate relation to other variables based on relationship previous research [17, 20]. Previous study mentioned that depression was significantly associated with dysfunctional belief and attitudes about sleep [45]. Kolmogorov–Smirnov test showed the data were normally distributed so we used Pearson correlation to evaluate the relationship [41] between depression (BDI-II) and DBAS-16 scale. However, as these instruments measure related but distinct constructs, a weak-to-moderate positive correlation was expected.

Reliability was assessed using Cronbach's alpha and McDonald's omega coefficients. Both coefficients were calculated for the original four-factor structure proposed by Morin et al. [45] and the three-factor structure identified in the present study. Researchers have noted that a Cronbach’s alpha of > 0.5 is considered to represent acceptable reliability [3, 43, 57] and omega > 0.70 is considered acceptable [42]. Inter-item and item-total correlations were calculated using Pearson correlation coefficients between all items of the scale. A correlation of > 0.3 between each domain is considered satisfactory [2].

Results

Sample Characteristic

This study sample comprised 1191 adolescents with a mean (standard deviation [SD]) age of 16.10 (0.95) years (range: 13–18 years; Table 1). The study achieved a 75% response rate, which is above that recommended (60%) by one study [19]. The majority of participants were female (66.5%), and the residence area was predominantly the west of Indonesia (Java, Sumatra, and Kalimantan Islands). The majority of participants slept on beds (79.1%), shared their beds with other friends (65.7%), and had pillows (81.4%). Approximately half of the participants had bolsters (51.6%) and less than half had blankets (43.3%). Only 8% of participants had private rooms (Table 1). Regarding factors affecting good sleep, including exercise once per week and sickness in the previous semester, it was revealed that every participant exercised every week, but the participants had different exercise frequencies (5.5% once per week and 94.5% more than once per week). Furthermore, half of participants had been sick in the previous semester (54.2%).

The average scores of the Indonesian version of the DBAS-16 items and dimensions are illustrated in Table 2. The highest mean value for an item was 7.65 for Item 3 (“I am concerned that chronic insomnia may have serious consequences on my physical health”), whereas the lowest mean score was 2.18 for Item 6 (“In order to be alert and function well during the day, I believe I would be better off taking a sleeping pill rather than having a poor night’s sleep”). The skewness for all items of the DBAS-16 scale was between − 1.19 and 1.24 and the kurtosis was from − 0.98 to 1.07 under the assumption of normality. The complete inter-item correlation matrix is presented in Appendix A.

Table 2.

Average scores of the DBAS-16 scale in Indonesian according to items and dimensions

DBAS-16 dimension	Scale item	Min	Max	Mean	SD	Skewness	Kurtosis
I need 8 h of sleep to feel refreshed and function well during the day	DBAS 1	0.00	10.00	6.11	2.78	− 0.35	− 0.71
When I don’t get a proper amount of sleep on a given night, I need to catch up the next day by napping or sleeping longer the next night	DBAS 2	0.00	10.00	5.00	2.86	0.01	− 0.96
I am concerned that chronic insomnia may have serious consequences on my physical health	DBAS 3	0.00	10.00	7.65	2.57	− 1.19	0.72
I am worried that I may lose control over my ability to sleep	DBAS 4	0.00	10.00	6.28	2.76	− 0.47	− 0.62
I know that a poor night’s sleep will interfere with my daily activities the next day	DBAS 5	0.00	10.00	6.26	3.05	− 0.57	− 0.81
In order to be alert and function well during the day, I believe I would be better off taking a sleeping pill rather than having a poor night’s sleep	DBAS 6	0.00	10.00	2.18	2.46	1.24	1.07
When I feel irritable, depressed, or anxious during the day, it is mostly because I did not sleep well the night before	DBAS 7	0.00	10.00	4.57	2.87	0.11	− 0.98
When I sleep poorly on one night, I know it will disturb my sleep schedule for the whole week	DBAS 8	0.00	10.00	3.43	2.87	0.59	− 0.64
Without an adequate night’s sleep, I can hardly function the next day	DBAS 9	0.00	10.00	5.98	3.07	− 0.41	− 0.96
I can’t ever predict whether I’ll have a good or poor night’s sleep	DBAS 10	0.00	10.00	6.15	2.85	− 0.49	− 0.59
I have little ability to manage the negative consequences of disturbed sleep	DBAS 11	0.00	10.00	5.39	2.82	− 0.24	− 0.91
When I feel tired, have no energy, or just seem not to function well during the day, it is generally because I did not sleep well the night before	DBAS 12	0.00	10.00	6.18	2.76	− 0.45	− 0.70
I believe insomnia is essentially the result of a chemical imbalance	DBAS 13	0.00	10.00	5.32	2.56	− 0.26	− 0.45
I feel insomnia is ruining my ability to enjoy life and prevents me from doing what I want	DBAS 14	0.00	10.00	5.84	2.98	− 0.36	− 0.89
Medication is probably the only solution to sleeplessness	DBAS 15	0.00	10.00	3.79	2.870	0.38	− 0.76
I avoid or cancel obligations (social or familial) after a poor night’s sleep	DBAS 16	0.00	10.00	6.26	2.64	0.79	− 0.25

Min Minimum, Max maximum, SD standard deviation

Validity evidence based on test content

The adaptation process of the DBAS-16 scale involved four steps: forward translation, expert panel review, back-translation, and pilot study. The outcome of the forward translation was an Indonesian version of the DBAS-16, which maintained the conceptual meaning of the original instrument. The expert panel review indicated that all items were clear, relevant, and culturally appropriate for Indonesian adolescents; therefore, no items were modified or removed. The back-translation resulted in an English version of the Indonesian DBAS-16 scale, which was compared with the original English version. Finally, pretesting among 10 adolescents showed that all items were understandable and appropriate for the target population.

Validity evidence based on internal structure

Factor structure of the dysfunctional beliefs and attitudes about sleep-16 scale

A Confirmatory Factor Analysis (CFA) was first performed to examine whether the original four-factor structure proposed by the original developers fit the Indonesian data. The fit indices indicated that the original model did not achieve acceptable model fit (χ² = 577,09, df = 98, p = 0.00, χ²/df = 5.88, GFI = 0.93, AGFI = 0.91, CFI = 0.87, RMSEA = 0.06, NFI = 0.85, and RFI = 0.82) (Fig. 1). Further, we conducted EFA before measuring construct validity. EFA was used to analyze all 16 items of the DBAS-16 scale. The value of the Kaiser–Meier–Olkin test was 0.893, and the Bartlett’s test value of sphericity was statistically significant (p < 0.001). As shown in Table 3, the principal component with varimax rotated factor indicated that a 3-factor model was optimal and had eigenvalues of > 1; thus a 3-factor model was retained. The values of rotated factor loading of the Indonesian version of the DBAS-16 for all items were > 0.32, which indicated favorable convergent validity (Table 3). Factor 1 emerged as the dominant dimension, consisting of eight items (DBAS 1, DBAS 2, DBAS 5, DBAS 7, DBAS 9, DBAS 10, DBAS 11, and DBAS 12), with factor loadings ranging from 0.38 to 0.62. Among these, DBAS 1 (0.62) and DBAS 9 (0.61) were the strongest indicators, whereas DBAS 10 (0.38) contributed relatively less to the factor. Factor 2 comprised four items (DBAS 6, DBAS 8, DBAS 15, and DBAS 16), with loadings ranging from 0.52 to 0.69. DBAS 6 (0.69) demonstrated the highest loading, indicating that it was the most representative indicator of this dimension. Factor 3 included four items (DBAS 3, DBAS 4, DBAS 13, and DBAS 14), with loadings ranging from 0.57 to 0.66. DBAS 3 (0.66) exhibited the strongest association with this factor, while the remaining items also showed satisfactory loadings above 0.57. In addition, the first factor was composed of eight items referring to consequences and expectations (explained variance of 28.98%); the second factor was composed of four items referring to medication and disturbed sleep (explained variance of 8.89%); and the third factor was four items referring to worry and helplessness (explained variance of 6.84%).

Fig. 1.

Confirmatory factor analysis of the original DBAS-16 scale factors structure. Model fit indices χ² = 577,09, df = 98, p = 0.00, χ.²/df = 5.88, GFI = 0.93, AGFI = 0.91, CFI = 0.87, RMSEA = 0.06, NFI = 0.85, and RFI = 0.82

Table 3.

Factor loading

	Factor loading
	Factor 1	Factor 2	Factor 3
DBAS 1	0.62
DBAS 2	0.61
DBAS 5	0.51
DBAS 7	0.56
DBAS 9	0.61
DBAS 10	0.38
DBAS 11	0.57
DBAS 12	0.52
DBAS 6		0.69
DBAS 8		0.52
DBAS 15		0.68
DBAS 16		0.64
DBAS 3			0.66
DBAS 4			0.57
DBAS 13			0.61
DBAS 14			0.62
Total
Eigenvalue	4.64	1.42	1.09
Explained variance (%.)	28.98	8.89	6.84

To validate the three structures of the Indonesian version of the DBAS-16 scale extracted through CFA, [13, 21]. We conducted single-factor and second order model. A model with second-order factors was applied, and the goodness of fit indices (χ² = 475.826, df = 101, p = 0.00, χ²/df = 4.71, GFI = 0.95, AGFI = 0.93, CFI = 0.90, RMSEA = 0.05, NFI = 0.88, and RFI = 0.86) are presented in Fig. 2. A comparison between the single-factor and second-order models showed that the second-order model provided a significantly better fit to the data, Δχ² (4) = 56.172, p < 0.001. Furthermore, the second-order model exhibited lower AIC values (431.88 vs 595.88), indicating superior model fit and parsimony (Table 4).

Fig. 2.

Confirmatory Factor Analysis model of The DBAS-16 scale after EFA with standardized factor loadings. Model fit indices (χ² = 475.826, df = 101, p = 0.00, χ.²/df = 4.71, GFI = 0.95, AGFI = 0.93, CFI = 0.90, RMSEA = 0.05, NFI = 0.88, and RFI = 0.86)

Table 4.

Comparison of competing CFA models after EFA

Model	χ2	df	CFI	TLI	RMSEA	AIC
Single-factor	531.998	104	.84	.82	.07	595.99
Second-order	475.826	101	.90	.92	.05	431.88

Validity evidence based on relations to other variables

The total score of DBAS-16 scale was significantly correlated with the total score of BDI II (r = 0.165, p < 0.05). This finding indicates a statistically significant but weak positive association between depressive symptoms and dysfunctional beliefs and attitudes about sleep (Fig. 3). In addition, known-groups validity was examined by comparing DBAS-16 scores between adolescents with normal group and depression risk group. Participants with depression risk group (20.14 ± 5.57 vs 8.28 ± 3.47) had significantly higher DBAS-16 scores than those with normal group (86.50 ± 23.01 vs. 80.32 ± 24.27, t (1189) = −4.424, p < 0.001), indicating that higher depressive symptoms were associated with greater dysfunctional beliefs and attitudes about sleep (Table 5).

Table 5.

Mean difference between adolescents of the normal and depression-risk groups

Depression risk group	n	BDI-II scores (Mean ± SD)	DBAS-16 scores (Mean ± SD)	t	df	p-value	Mean difference (95% CI)
Adolescents normal group	698	8.28 ± 3.47	80.32 ± 24.27	−4.424	1189	< 0.001	−6.18 (−8.92 to −3.44)
Adolescents’ depression risk group	493	20.14 ± 5.57	86.50 ± 23.01

Reliability

The internal consistency of the original DBAS-16 scale four-factor structure was evaluated using Cronbach's alpha and McDonald's omega coefficients (Table 6). The overall scale demonstrated good reliability (α = 0.83; ω = 0.83). At the subscale level, the consequences factor showed acceptable reliability (α = 0.69; ω = 0.70), whereas the worry and helplessness factor demonstrated marginal reliability (α = 0.64; ω = 0.64). Lower reliability coefficients were observed for the expectations factor (α = 0.48) and medication factor (α = 0.53; ω = 0.55). These findings suggest that the reliability of some dimensions of the original DBAS-16 scale structure was limited in the present sample.

Table 6.

Reliability of the Original DBAS-16 Four-Factor Structure

Factor	Items	α	ω
Consequences	5,7,9,12,16	0.69	0.70
Worry and helplessness	3,4,8,10,11,14	0.64	0.64
Expectations	1,2	0.48	-
Medication	6,13,15	0.53	0.55
Total		0.83	0.83

The Cronbach’s alpha for the internal consistency of the Indonesian version of the overall DBAS-16 scale was 0.83. Table 7 demonstrates that the Cronbach’s alpha of each subscale was 0.76 for Factor 1 (consequences and expectations), 0.66 for Factor 2 (medication and disturbed sleep), and 0.60 for Factor 3 (worry and helplessness). The McDonald's Omega of each subscale was 0.76 for Factor 1 (consequences and expectations), 0.65 for Factor 2 (medication and disturbed sleep), and 0.60 for Factor 3 (worry and helplessness). Statistically significant and positive correlations were observed between the global score of the Indonesian version of the DBAS-16 scale and its 13-item scores. Table 7 shows that the range of correlation between each item (r = 0.40–0.65, p < 0.05).

Table 7.

Inter-item and item-total correlations for the three DBAS-16 factors

	Factor 1 (Consequences and expectations)								Factor 2 (Medication and disturbed sleep)				Factor 3 (Worry and helplessness)				Item-total correlations	McDonald’s omega (ω) Indonesia	Cronbach’s alpha Indonesia
	1	2	5	7	9	10	11	12	6	8	15	16	3	4	13	14
Factor 1																		0.76	0.76
DBAS 1	-																0.48
DBAS 2	0.32	-															0.51
DBAS 5	0.24	0.24	-														0.58
DBAS 7	0.19	0.25	0.33	-													0.57
DBAS 9	0.28	0.26	0.32	0.35	-												0.63
DBAS 10	0.16	0.16	0.19	0.15	0.19	-											0.40
DBAS 11	0.27	0.27	0.35	0.35	0.44	0.24	-										0.65
DBAS 12	0.29	0.25	0.35	0.36	0.45	0.19	0.46	-									0.64
Factor 2																		0.65	0.66
DBAS 6									-								0.42
DBAS 8									0.33	-							0.60
DBAS 15									0.38	0.28	-						0.51
DBAS 16									0.29	0.35	0.34	-					0.51
Factor 3																		0.60	0.60
DBAS 3													-				0.44
DBAS 4													0.32	-			0.47
DBAS 13													0.19	0.19	-		0.48
DBAS 14													0.29	0.25	0.38	-	0.61
Total																		0.83	0.83

Discussion

To the best our knowledge, this is the first study to examine validity and reliability of the Indonesian version of the DBAS-16 scale. The Indonesian version of the DBAS-16 scale supported validity and reliability for evaluating sleep-related cognition. Our findings showed that the Indonesian version of the DBAS-16 scale had adequate validity and reliability. The validity was supported by validity evidence based on test content, internal structure, and relations to other variables. However, the finding of EFA the Indonesian version of DBAS-16 scale was different than that the originality DBAS-16 scale. Reliability was supported by Cronbach's alpha and McDonald's omega coefficients.

Validity evidence based on internal structure showed that the Indonesian version of the DBAS-16 scale had significant construct validity. EFA was used to extract factors from the 16 items in the questionnaire. The results of the factor analysis demonstrated changes in the factor structure from the originality DBAS-16 scale. However, the values of factor loading of the Indonesian version of the DBAS-16 scale for all items were greater than 0.32, and these loadings were similar to those of a previous study conducted in German [38]. In the EFA of the Indonesian version of the DBAS-16 scale, the factors “the consequences of insomnia” and “sleep expectation” were integrated into one factor, except Item 16 (“cancel obligation”), and two items were added, Item 10 (“sleep is unpredictable”) and Item 11 (“unable to manage consequences”). The second factor “medication and disturbed sleep” consisted of Item 6 (“better off taking sleeping pills”) and Item 15 (“medication as a solution”). High scores in Items 6 and 15 were associated with medication usage, and this factor also produced another two items: Item 8 (“one poor night disturbs the whole week” and Item 16 (“cancel obligations”). High scores in these items were associated with disturbed sleep. The last factor “worry and helplessness” produced Items 3 (“consequences of insomnia on health”), 4 (“fear of losing control over sleep”), 13 (“insomnia resulting from chemical imbalance”), and 14 (“insomnia destroying life”). In some populations, such as in cancer populations, factor analysis produced different dimensions and items [66]. We assumed that different interpretations of these items were due to differences in phrases used characteristics of populations, and cultures [66].

An important consideration in interpreting the present findings is the adolescent nature of the sample. The DBAS-16 was originally developed and validated primarily among adults with insomnia [45], whereas the current study focused on adolescents from the general population. Adolescence is characterized by substantial biological, environmental, lifestyle and psychosocial factors changes that influence sleep [28]. These developmental characteristics may partly explain why the original four-factor structure was not fully supported in the present sample. Adolescents may not clearly distinguish between certain cognitive domains related to sleep, resulting in the integration of dimensions such as sleep consequences and sleep expectations into a single factor. Furthermore, the participants in this study were recruited from the general adolescent population rather than from a clinical insomnia sample. Patients with insomnia and depression report more unhelpful beliefs about sleep [46]. Consequently, some beliefs related to sleep may have been less salient, which may explain the relatively lower reliability observed for certain factors.

This study conducted CFA to examine the construct validity of the Indonesian version of the DBAS-16 scale. The model structure was evaluated based on several fit indices, specifically GFI, AGFI, CFI, RFI, NFI, and RMSEA. Figure 2 shows that the RMSEA value was 0.05 and indicates that the final the Indonesian version of the DBAS-16 scale was a significant model. The RMSEA should be < 0.08. All factors of the Indonesian version revealed significant factor loading of 0.33 − 0.67. Three first-order factors (consequences and expectations, medication and disturbed sleep, and worry and helplessness) and one second-order factor (DBAS-16 scale) were used to test the second-order model and showed good model fit.

In addition, validity evidence based on relations to other variables were measured using Pearson correlation between total score of DBAS-16 scale and BDI-II. The DBAS-16 scale assesses dysfunctional beliefs and attitudes about sleep, whereas the BDI-II measures depressive symptoms; therefore, these instruments represent two conceptually distinct constructs. Consequently, a modest rather than a strong correlation between the two measures would be expected. Based on theory, people with dysfunctional beliefs and attitude of sleep may induced depression [45]. Previous studies have also reported positive associations between dysfunctional sleep beliefs and depressive symptoms, as cognitive distortions related to sleep may contribute to emotional distress [45, 63]. However, because depression is influenced by multiple biological, psychological, and social factors [51], only a weak association with sleep-related beliefs would be expected.

Reliability analyses revealed differences between the original four-factor structure and the three-factor structure identified in the Indonesian sample. Although the total DBAS-16 scale score demonstrated good internal consistency in both models (α = 0.83; ω = 0.83), several subscales of the original structure showed relatively low reliability, particularly the expectations and medication dimensions. The Cronbach’s alpha values were 0.76 for consequences and expectation, 0.66 for medication and disturbed sleep, and 0.60 for worry and helplessness. The Cronbach’s alpha values for the Indonesian version of the DBAS-16 scale showed adequate internal consistency ranged from 0.60 to 0.76 and McDonald's omega coefficients ranged from 0.60 to 0.76. Researchers have noted that the various DBAS-16 language versions had acceptable Cronbach’s alpha values (0.77 − 0.90) [18, 38, 66]. Studies have noted that a Cronbach’s alpha of > 0.5 is considered to be reliable [43, 57]. The reliability coefficients of the Medication and Worry factor were below 0.60. One possible explanation is that the present study was conducted among general adolescents rather than individuals with clinically significant insomnia. Beliefs regarding sleep medication may have limited relevance for adolescents who have no experience with pharmacological treatment for sleep problems. Similarly, concerns about the consequences of poor sleep, as reflected in the Worry factor, may not be consistently experienced by adolescents without persistent sleep disturbances [12].

Several limitations should be considered when interpreting the findings. First, participants were recruited from the general adolescent population rather than clinical samples. Therefore, the findings may not fully represent adolescents experiencing clinically significant insomnia. Second, a proportion of participants lived in dormitories, where environmental and contextual factors, such as structured daily schedules, shared sleeping environments, limited privacy, and institutional sleep routines, may influence sleep patterns and sleep-related beliefs. These factors should be considered because they may affect DBAS-16 scale responses and limit the generalizability of the findings to adolescents living in other settings. Previous research has also highlighted that heterogeneity in adolescent samples may limit comparisons across different living locations [25]. However, this study supported construct validity for evaluating sleep-related cognition among Indonesia Adolescents. Further study is required to replicate this result in a clinical sample. In addition, this was mainly was a construct study. We did not compare the Indonesian version of the DBAS-16 scale to other scales that measure sleep-related cognition.

Conclusion

Despite these limitations, our study revealed that the Indonesian version of the DBAS-16 scale supported validity and reliability for evaluating sleep-related cognition. The DBAS-16 scale contain 16 items related to cognition of sleep (consequences and expectations, medication and disturbed sleep, and worry and helplessness about sleep), and is an appropriate tool for cognition of sleep screening among Indonesian adolescents.

Authors’ contributions

Anggi Setyowati contributed to conception and design, acquisition, analysis, drafted manuscript, critically revised manuscript. Henny Dwi Susanti contributed to analysis, interpretation and critically revised manuscript. Min-Huey Chung contributed to conception and design, analysis and interpretation, critically revised manuscript. All authors gave final approval and agrees to be accountable for all aspects of work ensuring integrity and accuracy.

Funding

No funding was received by the authors for this study.

Data availability

The datasets used and/or analyzed as part of the present study are available from the corresponding author on reasonable request.

Declarations

Ethics approval consent to participate

The research process began with obtaining ethical clearance. This study received ethical clearance under number All procedures were approved by the Ethics Committee from Universitas Airlangga, number 1190/UN3.14/LT/2014.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.