Impact statement
This cross-sectional study provides validated tools for assessing various mental health conditions among youths in Hong Kong, offering a foundation for both clinical practice and research. By confirming the reliability and validity of eight widely used scales including measures for depression, anxiety, insomnia, internet and smartphone addiction and life satisfaction, the research ensures that professionals have robust instruments to identify and monitor mental health challenges among youths.
The broader impact lies in enabling early detection and intervention. With these scales, healthcare providers such as counsellors and psychologists can screen youths more effectively, tailoring preventive and therapeutic programs to their needs. Researchers can also employ these measures to track mental health trends over time, generating evidence that informs public health strategies and policy decisions. More importantly, these scales apply to older youths with different ethnic backgrounds residing in Hong Kong. Most of these older youths are female, from Chinese Mainland, university students, more digitally engaged and more comfortable with English.
Beyond Hong Kong, the findings contribute to the global effort to standardise youth mental health assessments. As technology continues to shape youths’ lives, these validated tools can be integrated into digital platforms, expanding access to mental health monitoring and support. Ultimately, this work strengthens the capacity of communities, schools and health systems to promote resilience and well-being among youths, ensuring that mental health challenges are addressed proactively rather than reactively.
Introduction
Mental health conditions among youths are on the increase globally, especially after 2019 due to COVID-19 (Liu et al., Reference Liu, Zhang, Chen, Li, Huang, Zhao, Chen, Qu and Li2025; Wang et al., Reference Wang, Dou, Yang, Guo, Ma, Zhou and Zhang2025b). These increasing conditions are a significant public health challenge as they translate into societal burden and detrimental impact at the personal level in the areas of schooling or working, health and functioning (Kaman et al., Reference Kaman, Devine, Erhart, Napp, Reiss, Moeller, Zoellner, Behn and Ravens-Sieberer2025; Wang et al., Reference Wang, Dou, Yang, Guo, Ma, Zhou and Zhang2025b). The reported prevalence estimates among youths are between 15.2% and 31%, with anxiety and depression disorders being the two leading disorders (Armoon et al., Reference Armoon, Ghadipasha, Mohammadi, Lesage, Harooni and Griffiths2025; Castaldelli-Maia et al., Reference Castaldelli-Maia, Ventriglio and Torales2025). This statistical information is important as 62.5% of most mental health conditions begin before age 25 years (Solmi et al., Reference Solmi, Radua, Olivola, Croce, Soardo, de Pablo, Il Shin, Kirkbride, Jones, Kim, Kim, Carvalho, Seeman, Correll and Fusar-Poli2022), suggesting an urgent need for periodic assessment and preventive mental healthcare among youths (i.e., people aged 10–29 years; Ahorsu et al., Reference Ahorsu, Sanchez Vidaña, Lin, Martin, Patel, Rajendram and Preedy2026b), although the present study specifically focuses on older youths (i.e., those between 16 and 29 years). Specific assessment of the biological, psychological and social factors underlying youths’ mental health conditions and their interactions with each other can help to provide a comprehensive understanding of the overall mental health of youths (Tripathi et al., Reference Tripathi, Das and Kar2019; Kirkbride et al., Reference Kirkbride, Anglin, Colman, Dykxhoorn, Jones, Patalay, Pitman, Soneson, Steare, Wright and Griffiths2024). Therefore, this study has carefully chosen factors of mental health conditions that are most prevalent among youths based on each of the factors of the biopsychosocial model and especially among older youths in Hong Kong.
Another important issue associated with mental health conditions among youths is comorbidity and multimorbidity (Fisher, Reference Fisher2022). In this regard, a study among youth found that 64.7% of participants with one to three risk factors and 21.1% with four or more risk factors had 2.6- and 4.6-fold higher odds, respectively, of experiencing comorbidity (Pigatto et al., Reference Pigatto, Grant, Marks, Walker and Waldie2025). For instance, it has been found that insomnia, obsessive-compulsive disorder (OCD), depression, anxiety and somatic symptoms are associated through the insomnia-obsessive-compulsive-anxiety network (Jurado-González et al., Reference Jurado-González, García-Torres, Contreras, Muñoz-Navarro, González-Blanch, Medrano, Ruiz-Rodríguez, Cano-Vindel and Moriana2024; Yuan et al., Reference Yuan, Wang, Shi, Zhao, Tao, Wu, Tan and Ouyang2025). Therefore, it is recommended that clinicians look at health from a holistic perspective with the assumption that there may be interrelationships between different conditions so as to offer better care (Fisher, Reference Fisher2022). Consequently, there is a need for valid psychometric scales to assess or screen mental health conditions. Particularly, the Generalised Anxiety Disorder (GAD) scale, Patient Health Questionnaire (PHQ), Somatic Symptom Scale (SSS), Florida Obsessive-Compulsive Inventory (FOCI) and Insomnia Severity Index (ISI), which are used for anxiety, depression, somatic symptoms, OCD symptoms and ISI, respectively, were validated in this study due to its interrelatedness (Hybelius et al., Reference Hybelius, Kosic, Salomonsson, Wachtler, Wallert, Nordin and Axelsson2024; Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a). Moreover, the advancement of technological devices has been associated with problematic uses (e.g., smartphone application-based addiction [SABAS], internet gaming addiction and social media addiction) of these devices, especially among youths (Ahorsu, Reference Ahorsu2024; Rahman et al., Reference Rahman, Sarkar, Haque, Rabiul Islam, Haque, Roy, Liza, Tabashsum, Alam, Hasan, Adam, Duc, Mahdee and Asha2025; Siaw et al., Reference Siaw, Chua, Ruckwongpatr, Tung, Gan, Poon, Nadhiroh, Nurmala, Üztemur, Chen, Griffiths and Lin2025; Zhang et al., Reference Zhang, Tao, Zhang, Zhou, Wei, Chen, Hu, Zheng and Wang2025). Additionally, these problematic behaviours are associated with different mental health conditions (Phetphum et al., Reference Phetphum, Keeratisiroj and Prajongjeep2023; Huang et al., Reference Huang, Geusens, Tu, Fung and Chen2024a, Reference Huang, Ahorsu, Adjaottor, Addo, Griffiths, Pakpour and Lin2024b).
Hong Kong, one of the most advanced regions in the world in terms of technology (i.e., infrastructure, smart city and AI and rapid adoption) and social media usage (Lee and Busiol, Reference Lee and Busiol2016; Cole et al., Reference Cole, Stivas, Tran and Lai2023; The Government of the Hong Kong Special Administrative Region, 2025), faces similar issues as most advanced countries. Previous studies revealed that the local prevalence rates of internet addiction among adolescents were between 3.0% and 26.8%, with internet addiction affecting their physical, mental and psychosocial development (Lee and Busiol, Reference Lee and Busiol2016; Chung et al., Reference Chung, Sum and Chan2019). Also, there are significant associations between the mental health variables, including social media addiction, depression, anxiety, stress, somatic symptoms, OCD symptoms and ISI (Wong et al., Reference Wong, Mo, Potenza, Chan, Lau, Chui, Pakpour and Lin2020; Wang et al., Reference Wang, Wong, Wang, Li, Kim and Lee2021; Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a). However, previous psychometric property (i.e., validity and reliability) studies among Hong Kong youths (i.e., including adolescents) on these scales were based on the Simplified or Traditional Chinese characters, which neglects minority groups in the general population (Chung et al., Reference Chung, KK- and Yeung2011; Lai et al., Reference Lai, Mak, Watanabe, Ang, Pang and Ho2013; Lun et al., Reference Lun, Chan, Ip, Ma, Tsai, Wong, Wong, Wong and Yan2018; Chen et al., Reference Chen, Strong, Lin, Tsai, Leung, Lin, Pakpour and Griffiths2020; Leung et al., Reference Leung, Mak, Leung, Chiang and Loke2020a, Reference Leung, Pakpour, Strong, Lin, Tsai, Griffiths, Lin and Chen2020b; Fang et al., Reference Fang, Yuen, Fung, Zhang, Chan and Wu2022). Moreover, some of the measures (e.g., PHQ-9, SSS-8 and ISI) have different factor structures that fit other cultures and age groups better than others (Boothroyd et al., Reference Boothroyd, Dagnan and Muncer2019; Otte et al., Reference Otte, Bakoyannis, Rand, Ensrud, Guthrie, Joffe, McCurry, Newton and Carpenter2019; Li et al., Reference Li, Wei, Fritzsche, Toussaint, Zhang, Zhang, Chen, Wu, Ma, Li, Ren, Lu and Leonhart2022; Blake et al., Reference Blake, Munyombwe, Fischer, Quinn, Van der Feltz-Cornelis, De Man-van Ginkel, Santos, Jeon, Köhler, Schram, Wang, Levin-Aspenson, Whooley, Hobfoll, Patten, Simning, Gracey and Broomfield2025; Hybelius et al., Reference Hybelius, Kosic, Salomonsson, Wachtler, Wallert, Nordin and Axelsson2025; Wang et al., Reference Wang, Shen, Guo, Liu, Chen, Ma and Jian2025a; Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a), and so examining these alternative factor structures will help more in capturing youth-specific symptom expressions (see Table 1). The present study intends to examine older youths with different backgrounds (i.e., students, dependents, ethnic minorities and natives) who consider Hong Kong as their place of abode legally.
Summary of the factor structures available for all the measures
Table 1. Long description
The table consists of four columns: Measures, Factor structure, Model specification, and Reference.
* P H Q dash 9:
- One-factor structure.
- Two-factor (A): 1st factor is Cognitive/affective; 2nd factor is Somatic.
- Two-factor (B): 1st factor is Cognitive/affective; 2nd factor is Somatic.
* G A D:
- One-factor structure.
* S S S dash 8:
- One-factor structure.
- Three-factor: 1st factor is Gastrointestinal and fatigue; 2nd factor is Pain; 3rd factor is Cardiopulmonary.
- Four-factor: 1st factor is Gastrointestinal; 2nd factor is Pain; 3rd factor is Cardiopulmonary; 4th factor is Fatigue.
* S A B A S:
- One-factor structure.
* F O C I:
- One-factor structure.
* I A T dash 6:
- One-factor structure.
* I S I:
- One-factor structure.
- Two-factor (A): 1st factor is Insomnia severity; 2nd factor is Insomnia impact.
- Two-factor (B): 1st factor is Nighttime sleep problem; 2nd factor is Daytime impairment.
* B M S L S S:
- One-factor structure.
References include Kroenke et al. 2001, Boothroyd et al. 2019, Blake et al. 2025, Spitzer et al. 2006, Gierk et al. 2014, Li et al. 2022, Hybelius et al. 2025, Csibi et al. 2018, Storch et al. 2007, Ali et al. 2021, Bastien et al. 2001, Otte et al. 2019, and Wang et al. 2025.
Notes: GAD = severity measure for general anxiety disorder; PHQ-9 = patient health questionnaire-9; SSS-8 = somatic symptom scale-8; FOCI = Florida obsessive-compulsive inventory; SABAS = smartphone application-based addiction scale; IAT-6 = internet addiction test-6; BMSLSS = brief multidimensional students’ life satisfaction scale; ISI = insomnia severity index.
a First factor had items 1, 2, 6 and 9; the second factor had items 3, 4, 5, 7 and 8.
b First factor had items 1, 2, 6, 7, 8 and 9; the second factor had items 3, 4 and 5.
c First factor had items 1, 7 and 8; the second factor had items 2 and 3; the third factor had items 4–6.
d First factor had items 1; the second factor had items 2–4; the third factor had items 5 and 6; the fourth factor had items 7 and 8.
e First factor had items 1–4; the second factor had items 5–7.
f First factor had items 1–3; the second factor had items 4–7.
So far, known psychometric studies included internet addiction (IAT) (Lai et al., Reference Lai, Mak, Watanabe, Ang, Pang and Ho2013), SABAS (Chen et al., Reference Chen, Strong, Lin, Tsai, Leung, Lin, Pakpour and Griffiths2020; Leung et al., Reference Leung, Pakpour, Strong, Lin, Tsai, Griffiths, Lin and Chen2020b), depression (Lun et al., Reference Lun, Chan, Ip, Ma, Tsai, Wong, Wong, Wong and Yan2018; Leung et al., Reference Leung, Mak, Leung, Chiang and Loke2020a), anxiety (Lun et al., Reference Lun, Chan, Ip, Ma, Tsai, Wong, Wong, Wong and Yan2018; Ip et al., Reference Ip, Suen, Hui, Wong, Chan, Lee, Wong and Chen2022) and brief multidimensional students’ life satisfaction (BMSLSS) (Fang et al., Reference Fang, Yuen, Fung, Zhang, Chan and Wu2022) and ISI (Chung et al., Reference Chung, KK- and Yeung2011). However, since Hong Kong is a cosmopolitan region with several groups of people who speak different languages, there is a need for a psychometric property study that uses a language that all residents are familiar with and examines the psychometric properties of the scales based on that language (Lim, Reference Lim and Lim2011; Huang, Reference Huang2023; Keenan, Reference Keenan2023). As English is one of the two official languages in Hong Kong and is used as a medium of instruction at university level, this study chose to design the questionnaire in English. The present study is novel in several ways: 1) there are no reported psychometric properties for the scales somatic symptom (SSS-8), and obsessive-compulsive symptom (FOCI) among youths in Hong Kong; 2) previous studies largely focused on adolescents; 3) some previous psychometric properties of the scales only targeted youths who speak Cantonese or read and write Traditional Chinese characters (e.g., PHQ-9, GAD-7) and 4) some of the scales of previous studies (i.e., FOCI) have different versions. Therefore, this study examines the preliminary psychometric properties of the scales (i.e., GAD-7, PHQ-9, SSS-8, FOCI, ISI, SABAS, IAT-6 and BMSLSS) among older youths residing in Hong Kong to ascertain 1) the psychometric properties (reliability and validity) of the scales and 2) the relationships between these scales.
Methods
Settings and population
Secondary schools and universities in Hong Kong were used as the settings of the present study, although the questionnaire was completed online. Specifically, posters with a QR code for the questionnaire were pasted on the noticeboards at the schools and universities. The population used was youths (i.e., those who have been living in Hong Kong as students, dependents, ethnic minorities or natives and may consider Hong Kong as their place of abode) and, more specifically, older youths aged 16–29 years with a Hong Kong Identity Card (i.e., HKID) as proof of legal residence.
Participants and procedure
A total of 428 participants were recruited from the secondary schools and universities in Hong Kong to respond to an online questionnaire for this cross-sectional survey design study. This online questionnaire was accessed by scanning the QR code on a poster. Upon scanning the QR code, participants were sent to the information page, consent page and the main questionnaire, which was hosted on a REDCap platform. The information page contains detailed information about the study, and the consent page explicitly states the ethical conditions applicable to both the study and study participants. All participants must consent by signing the consent form before they can access the questionnaire page. The data collection took place between January and November 2025. The main inclusion criteria were 1) participants must be between 16 and 29 years old, 2) they must understand, read and write English and 3) they must be residents of Hong Kong. Exclusion criteria included participants who were 1) exchange students and 2) less than 16 years or above 30 years. Generally, a convenience sampling strategy and snowballing strategies were used for all data collection. Participants spent about 25 minutes responding to the full questionnaire. The response rate was 90.49%. The Human Research Ethics Committee of The Education University of Hong Kong issued the ethical approval for the study (approval: Ref. no. 2022-2023-0462) on January 17, 2024. This study complied with the ethical standards and procedures involving the use of human subjects as prescribed by the ethics committee and in accordance with the Helsinki Declaration of 1975, as revised in 2013.
Measures
Patient Health Questionnaire-9 (PHQ-9)
The PHQ-9, a DSM-5-adapted version (Kroenke et al., Reference Kroenke, Spitzer and Williams2001; American Psychiatric Association [APA], 2022), was used to assess the severity of depression. The response for each item was on a four-point Likert scale (i.e., 0 = Not at all to 3 = Nearly every day). A participant’s score was calculated by summing each of the responses to the items. Therefore, the total score can be within 0 and 27, with higher scores indicating greater severity of depression. Moreover, the suggested cut-off scores in terms of depression severity were 0–4 for none-minimal depression, 5–9 for mild depression, 10–14 for moderate depression, 15–19 for moderately severe depression and 20–27 for severe depressive symptoms (Kroenke et al., Reference Kroenke, Spitzer and Williams2001; Yu et al., Reference Yu, Tam, Wong, Lam and Stewart2012).
Generalised Anxiety Disorder-7 (GAD-7)
The GAD-7 was used to assess the severity of GAD over a 2-week period (Spitzer et al., Reference Spitzer, Kroenke, Williams and Löwe2006). The response for each item was on a five-point Likert scale from 4 (nearly every day) to 0 (not at all). A participant’s score was calculated by summing each of the responses to the items. Therefore, the total score can be between 0 and 40, with higher scores indicating greater severity of GAD. In terms of categorising its severity, scores of 0–4 were for minimal anxiety, 5–9 for mild anxiety, 10–14 for moderate anxiety and 15–21 for severe anxiety symptoms, with a total score of 10 or greater generally used to identify cases where a formal diagnosis of GAD may be considered (Spitzer et al., Reference Spitzer, Kroenke, Williams and Löwe2006). However, a score of 7 or greater is proposed to be used as a cut-off score among people in Hong Kong (Ip et al., Reference Ip, Suen, Hui, Wong, Chan, Lee, Wong and Chen2022).
Somatic Symptom Scale-8 (SSS-8)
The SSS-8 was used to assess participants’ somatic symptoms (Gierk et al., Reference Gierk, Kohlmann, Kroenke, Spangenberg, Zenger, Brähler and Löwe2014). The response for each item was on a five-point Likert scale from 4 (very much) to 0 (not at all). A participant’s total score was calculated by summing each of the responses to the items. Therefore, the total score can be within 0 and 32, with higher scores indicating greater severity of somatic symptoms. The severity can also be categorised into no – minimal somatic symptom burden (0–3), low symptom burden (4–7), medium symptom burden (8–11), high symptom burden (12–15) and very high symptom burden (16–32) with a score of 8 and above considered as the threshold of clinical significance and further examination (Gierk et al., Reference Gierk, Kohlmann, Kroenke, Spangenberg, Zenger, Brähler and Löwe2014).
Smartphone Application-Based Addiction Scale (SABAS)
The SABAS, a six-item scale, was used to assess participants’ problematic use of (or addiction to) smartphone applications (Csibi et al., Reference Csibi, Griffiths, Cook, Demetrovics and Szabo2018). The response for each item was on a six-point Likert scale from 6 (strongly agree) to 1 (strongly disagree). A participant’s total score was calculated by summing each of the responses to the items. Therefore, the total score can be between 6 and 36, with higher scores indicating greater smartphone addiction risk. The severity can also be categorised into normal (6–13), low addiction risk (14–21), moderate addiction risk (22–32) and high addiction risk (33–36) (Lin et al., Reference Lin, Imani, Broström, Nilsen, Fung, Griffiths and Pakpour2019), with a score of 23 and above considered as the threshold of clinical significance and further examination (Peng et al., Reference Peng, Chen, Ren, Liu, He, Liang, Tan, Tang, Chen and Liao2023).
Florida Obsessive-Compulsive Inventory (FOCI)
The FOCI, a DSM-5 adapted version of a five-item scale, was used to assess participants’ repetitive thoughts and behaviours (usually of OCD) (Storch et al., Reference Storch, Bagner, Merlo, Shapira, Geffken, Murphy and Goodman2007; APA, 2022). The response for each item was on a five-point Likert scale from 4 (extreme) to 0 (none). A participant’s score was calculated by summing each of the responses to the items. Therefore, the total score can be within 0 and 20, with higher scores indicating greater severity of repetitive thoughts and behaviours. Clinically, a total average score (i.e., total score divided by the completed number of items) could be used, where scores can be categorised into none (0), mild (1), moderate (2), severe (3) and extreme (4) with the moderate category or a score of 8 considered a threshold of clinical significance and further examination (Storch et al., Reference Storch, Bagner, Merlo, Shapira, Geffken, Murphy and Goodman2007).
Internet Addiction Test-6 (IAT)
The IAT-6 was used to assess participants’ internet addiction status (Ali et al., Reference Ali, Hendawy, Almarwani, Alzahrani, Ibrahim, Alkhamees and Kunugi2021). The response for each item was on a six-point Likert scale from 5 (very often) to 0 (not applicable). A participant’s total score was calculated by summing each of the responses to the items. Therefore, the total score can be within 0 and 30, with higher scores indicating greater risks of internet addiction. Clinically, using a monothetic scoring scheme (i.e., scoring 3 or above on all six items), a score of 18 and above (i.e., problematic internet use) may be considered as a threshold of clinical significance and further examination (Ali et al., Reference Ali, Hendawy, Almarwani, Alzahrani, Ibrahim, Alkhamees and Kunugi2021).
Insomnia Severity Index (ISI)
The ISI, a seven-item scale, was used to assess participants’ ISI (Bastien et al., Reference Bastien, Vallières and Morin2001). The response for each item was on a five-point Likert scale from 4 (very much) to 0 (not at all). A participant’s total score was calculated by summing each of the responses to the items. Therefore, the total score can be between 0 and 28, with higher scores indicating greater severity of insomnia. The severity can also be categorised into absence of insomnia (0–7), sub-threshold insomnia (8–14), moderate insomnia (15–21) and severe insomnia (22–28) (Bastien et al., Reference Bastien, Vallières and Morin2001), with a score of 8 and above considered as the threshold of clinical significance and further examination, especially among individuals in Hong Kong (Wong et al., Reference Wong, Lau, Espie, Luik, Kyle and Lau2017).
Brief Multidimensional Students’ Life Satisfaction Scale (BMSLSS)
The BMSLSS, a five-item scale, was used to assess participants’ life satisfaction (Seligson et al., Reference Seligson, Huebner and Valois2003). The response for each item was on a seven-point Likert scale from 7 (Delighted) to 1 (Terrible). A participant’s total score was calculated by summing each of the responses to the items. Therefore, the total score can be between 5 and 35, with higher scores indicating greater overall life satisfaction.
Stress from Depression Anxiety Stress Scale for Youth (DASS-Y)
A seven-item stress subscale of the DASS-Y (Szabo and Lovibond, Reference Szabo and Lovibond2022; Cao et al., Reference Cao, Liao, Gamble, Li, Jiang, Li, Griffiths, I-H and Lin2023) was used to assess participants’ stress levels. This scale was added to the present study to help extend the validity of the other mental health scales. The response for each item was on a four-point Likert scale from 3 (very true) to 0 (not true). A participant’s total score was calculated by summing each of the responses to the items. Therefore, the total score can be within 0 and 21, with higher scores indicating more severe stress. The Cronbach’s α and McDonald’s ω of this scale are 0.884 and 0.883, respectively.
PedsQL Multidimensional Fatigue Scale (PedsQL-MFS)
A six-item cognitive fatigue subscale of the PedsQL-MFS (Varni and Limbers, Reference Varni and Limbers2008) was used to assess participants’ cognitive fatigue severity. The response for each item was on a five-point Likert scale from 4 (almost always a problem) to 0 (never a problem). However, responses are reverse-scored (i.e., “4” now becomes “0”) and rescaled to 0-100 (i.e., 0, 25, 50, 75, 100 for “0”, “1”, “2”, “3” and “4”). Therefore, the total score (i.e., the mean of transformed item scores) ranges between 0 and 100, with higher scores indicating less severe cognitive fatigue. The α and ω of this scale are 0.904 and 0.902, respectively. The addition of this scale helped extend the validity of the other mental health scales.
Data analysis
Demographic characteristics of participants and item properties were analysed using descriptive statistics. Afterwards, the measures (i.e., PHQ-9, GAD-7, SSS-8, SABAS, FOCI, IAT-6, ISI and BMSLSS) were evaluated for their scale properties using confirmatory factor analysis (CFA) and internal consistency. Specifically, CFA’s comparative fit index (CFI), Tucker–Lewis index (TLI), root mean square error of approximation (RMSEA) and standardised root mean square residual (SRMR), with CFI and TLI > 0.9, and RMSEA and SRMR < 0.08 indicative of an acceptable one-factor structure were used (Hoyle and Panter, Reference Hoyle and Panter1995). Also, the CFA used weighted least squares with mean and variance adjusted estimator as all the measures were scored on a Likert-type scale (Li, Reference Li2016). Moreover, further CFAs were done for PHQ-9 (Boothroyd et al., Reference Boothroyd, Dagnan and Muncer2019; Blake et al., Reference Blake, Munyombwe, Fischer, Quinn, Van der Feltz-Cornelis, De Man-van Ginkel, Santos, Jeon, Köhler, Schram, Wang, Levin-Aspenson, Whooley, Hobfoll, Patten, Simning, Gracey and Broomfield2025; Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a), SSS-8 (Li et al., Reference Li, Wei, Fritzsche, Toussaint, Zhang, Zhang, Chen, Wu, Ma, Li, Ren, Lu and Leonhart2022; Hybelius et al., Reference Hybelius, Kosic, Salomonsson, Wachtler, Wallert, Nordin and Axelsson2025; Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a), and ISI (Otte et al., Reference Otte, Bakoyannis, Rand, Ensrud, Guthrie, Joffe, McCurry, Newton and Carpenter2019; Liang et al., Reference Liang, Labarda and Chan2023; Lenderking et al., Reference Lenderking, Savva, Atkinson, Campbell, Chabot, Moline, Meier and Morin2024; Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a) as they have been reported to have multiple factor structures. Additionally, construct validity was examined through average variance extracted (AVE) and composite reliability (CR), with acceptable cut-offs ≥0.50 and ≥0.70, respectively (Fornell and Larcker, Reference Fornell and Larcker1981).
Additionally, CFA’s factor loadings and corrected item-to-total correlations (CITCs) were calculated for each measure’s item to evaluate if each item contributes substantially to the measure it belongs to. Acceptable factor loadings and CITC should have an index greater than 0.3 (Mokkink et al., Reference Mokkink, Terwee, Patrick, Alonso, Stratford, Knol, Bouter and de Vet2010; Tavakol and Dennick, Reference Tavakol and Dennick2011; Boateng et al., Reference Boateng, Neilands, Frongillo, Melgar-Quiñonez and Young2018). Cronbach’s α and McDonald’s ω were calculated for each measure, with indices > 0.7 considered an acceptable reliability coefficient (Taber, Reference Taber2018). Lastly, Pearson’s r correlational analysis with a heatmap was used to examine the intermeasure correlations between all the measures used in this study. JASP version 0.19.3.0 (JASP Team; Amsterdam, The Netherlands) was used to analyse the data.
Result
The demographic characteristics of the participants have been presented in Table 2. There were a total of 428 participants, with the majority of the participants originally from the Chinese Mainland (n = 244, 57%), followed by Hong Kong natives (n = 137, 32%), and then “Others” (n = 47, 11%), which included participants from Angola, Austria, Bangladesh, Germany, Ghana, India, Indonesia, Iran, Kazakhstan, Macau, Malawi, Malaysia, Myanmar, Niger, Pakistan, Philippines, Rwanda, Tanzania, United States of America and Vietnam.
Demographics (n = 428)
Table 2. Long description
The table presents data for 428 participants across several categories:
* General Demographics: Mean age is 22.59 (S D 2.88). Sex distribution is 103 males (24.1 percent) and 325 females (75.9 percent). Education is primarily University level (94.9 percent). Citizenship includes China mainland (57 percent), Hong Kong (32 percent), and others (11 percent).
* P H Q 9 (Depression): 29 percent none-minimal, 40.2 percent mild, 18.5 percent moderate, 9.8 percent moderately severe, and 2.6 percent severe.
* G A D 7 (Anxiety): 35.5 percent minimal, 36.2 percent mild, 18.9 percent moderate, and 9.3 percent severe. 48.1 percent meet the Hong Kong cut-off for anxiety (greater than or equal to 7).
* S S S 8 (Somatic Symptoms): 26.9 percent minimal, 28.5 percent low, 19.4 percent medium, 10.5 percent high, and 14.7 percent very high burden.
* S A B A S (Social Media Addiction): 17.1 percent normal, 41.1 percent low risk, 40.2 percent moderate risk, and 1.6 percent high risk.
* F O C I (Fear of COVID 19): 15.2 percent none, 40.0 percent mild, 35.7 percent moderate, 8.6 percent severe, and 0.5 percent extreme.
* I A T 6 (Internet Addiction): 65.9 percent normal and 34.1 percent problematic use (greater than or equal to 18).
* I S I (Insomnia): 32.2 percent absence, 42.1 percent sub-threshold, 22.9 percent moderate, and 2.8 percent severe.
The sample characteristic data based on the measures shows that, in all, there were 71.1% of participants with various levels of depression severity (i.e., mild to severe depression), 64.4% with various levels of anxiety severity (i.e., mild to severe anxiety), 73.1% with various levels of somatic symptoms (i.e., low to very high symptom burden), 82.9% with various levels of smartphone application-based addiction risks (i.e., low to high addiction risks), 84.8% with various levels of obsessive compulsive symptoms (i.e., mild to extreme symptoms), 34.1% with problematic internet use and 67.8% with various levels of insomnia severities (i.e., sub-threshold to severe insomnia). However, taking into account reported cut-offs that signify potential mental health conditions, there were 30.9% of participants with potential depression condition (i.e., moderate to severe depression), 48.1% with potential anxiety condition, 44.6% with potential somatic condition (i.e., medium to very high symptom burden), 41.8% with potential smartphone application-based addiction risks (i.e., moderate to high addiction risks), 44.8% with potential obsessive compulsive condition (i.e., moderate to extreme symptoms), 34.1% with problematic internet use and 67.8% with potential insomnia condition (i.e., sub-threshold to severe insomnia).
The results of the CFA fit indices for the eight psychological measures (i.e., PHQ-9, GAD-7, SSS-8, SABAS, FOCI, IAT-6, ISI and BMSLSS) have been presented in Table 3. Generally, the CFA fit indices for all eight psychological measures were acceptable (CFI > 0.9, TLI > 0.9, RMSEA < 0.08, and SRMR < 0.08) except for the RMSEA index, which had a few measures exceeding 0.08. See Table 3 for further details.
CFA fit indices and internal consistency of the measures based on their one-factor structure
Table 3. Long description
The table consists of 10 columns and 9 rows. The columns are labeled: Measure, chi-squared (df), p-value, C F I, T L I, R M S E A, S R M R, alpha/omega, A V E, and C R.
* P H Q 9: chi-squared 84.442, p < 0.001, C F I 0.986, T L I 0.981, R M S E A 0.071, S R M R 0.049, alpha/omega 0.878/0.893, A V E 0.558, C R 0.918.
* G A D: chi-squared 50.458, p < 0.001, C F I 0.995, T L I 0.993, R M S E A 0.078, S R M R 0.024, alpha/omega 0.918/0.925, A V E 0.714, C R 0.946.
* S S S 8: chi-squared 95.603, p < 0.001, C F I 0.975, T L I 0.965, R M S E A 0.094, S R M R 0.045, alpha/omega 0.873/0.887, A V E 0.550, C R 0.907.
* S A B A S: chi-squared 34.819, p < 0.001, C F I 0.991, T L I 0.985, R M S E A 0.082, S R M R 0.027, alpha/omega 0.846/0.860, A V E 0.542, C R 0.875.
* F O C I: chi-squared 53.366, p < 0.001, C F I 0.992, T L I 0.983, R M S E A 0.151, S R M R 0.028, alpha/omega 0.899/0.909, A V E 0.724, C R 0.929.
* I A T 6: chi-squared 144.918, p < 0.001, C F I 0.954, T L I 0.924, R M S E A 0.188, S R M R 0.047, alpha/omega 0.863/0.884, A V E 0.565, C R 0.886.
* I S I: chi-squared 283.790, p < 0.001, C F I 0.928, T L I 0.892, R M S E A 0.212, S R M R 0.086, alpha/omega 0.834/0.895, A V E 0.531, C R 0.884.
* B M S L S S: chi-squared 128.087, p < 0.001, C F I 0.964, T L I 0.941, R M S E A 0.176, S R M R 0.046, alpha/omega 0.864/0.879, A V E 0.568, C R 0.886.
Notes: GAD = severity measure for General Anxiety Disorder; PHQ-9 = Patient Health Questionnaire-9; SSS-8 = Somatic Symptom Scale-8; FOCI = Florida Obsessive-Compulsive Inventory; SABAS = Smartphone Application-Based Addiction Scale; IAT-6 = Internet Addiction Test-6; BMSLSS = Brief Multidimensional Students’ Life Satisfaction Scale; ISI = Insomnia Severity Index; CFI = Comparative Fit Index; TLI = Tucker–Lewis index; RMSEA = root mean square error of approximation; SRMR = standardised root mean square residual; AVE = average variance extracted; CR = composite reliability.
The internal reliability coefficients of this study were generally acceptable. That is, all the coefficients for Cronbach’s α and McDonald’s ω were between 0.834 and 0.918 and 0.860 and 0.925, respectively (see Table 3). Convergent validity indices, as measured by both CR and AVE, were found to be acceptable. Specifically, the AVE ranges between 0.531 and 0.724 and CR between 0.875 and 0.946.
Table 4 shows the alternative models for the scales PHQ-9, SSS-8 and ISI. It can be observed that all the models generally had an acceptable factor structure.
CFA fit indices for competing models of measures with different factor structures
Table 4. Long description
The table consists of 8 columns: Measure, chi-squared, p-value, C F I, T L I, R M S E A, S R M R, and Structure.
For P H Q 9:
* One-factor: chi-squared 84.442, p-value less than 0.001, C F I 0.986, T L I 0.981, R M S E A 0.071, S R M R 0.049.
* Two-factor A: chi-squared 82.535, p-value less than 0.001, C F I 0.986, T L I 0.981, R M S E A 0.071, S R M R 0.048.
* Two-factor B: chi-squared 83.312, p-value less than 0.001, C F I 0.986, T L I 0.980, R M S E A 0.072, S R M R 0.049.
For S S S 8:
* One-factor: chi-squared 95.603, p-value less than 0.001, C F I 0.975, T L I 0.965, R M S E A 0.094, S R M R 0.045.
* Three-factor: chi-squared 42.111, p-value less than 0.001, C F I 0.992, T L I 0.986, R M S E A 0.059, S R M R 0.031.
* Four-factor: chi-squared 55.600, p-value less than 0.001, C F I 0.971, T L I 0.945, R M S E A 0.080, S R M R 0.028.
For I S I:
* One-factor: chi-squared 283.790, p-value less than 0.001, C F I 0.928, T L I 0.892, R M S E A 0.212, S R M R 0.086.
* Two-factor A: chi-squared 232.649, p-value less than 0.001, C F I 0.941, T L I 0.905, R M S E A 0.199, S R M R 0.074.
* Two-factor B: chi-squared 86.150, p-value less than 0.001, C F I 0.980, T L I 0.968, R M S E A 0.115, S R M R 0.050.
Note: PHQ-9 = Patient Health Questionnaire-9; SSS-8 = Somatic Symptom Scale-8; ISI = Insomnia Severity Index; CFI = Comparative Fit Index; TLI = Tucker–Lewis index; RMSEA = root mean square error of approximation; SRMR = standardised root mean square residual.
a First factor had items 1, 2, 6 and 9; the second factor had items 3, 4, 5, 7 and 8.
b First factor had items 1, 2, 6, 7, 8 and 9; the second factor had items 3, 4 and 5.
c First factor had items 1, 7 and 8; the second factor had items 2 and 3; the third factor had items 4–6.
d First factor had items 1; the second factor had items 2–4; the third factor had items 5 and 6; the fourth factor had items 7 and 8.
e First factor had items 1–4; the second factor had items 5–7.
f First factor had items 1–3; the second factor had items 4–7.
Factor loadings, CITCs and item means (and SDs) and skewness and kurtosis for each of the psychological measures have been presented in Table 5. Generally, the factor loadings for all the measures’ items were acceptable (i.e., >0.30) with CITCs between 0.412 and 0.795.
Factor loadings, corrected item-to-total correlations (CITCs), item means (SD), skewness and kurtosis for measures based on their one-factor structure
Table 5. Long description
The table consists of five columns: Measure Item number, Factor loading, C I T C (Corrected Item-to-Total Correlation), Mean (S D), and Skew/Kurt (Skewness/Kurtosis). It is divided into sections by measure:
* P H Q 9 (Items 1 to 9): Factor loadings range from 0.612 to 0.841. C I T C values range from 0.517 to 0.735. Means range from 0.362 to 1.355.
* G A D 7 (Items 1 to 7): Factor loadings range from 0.774 to 0.895. C I T C values (all in italics) range from 0.682 to 0.795. Means range from 0.729 to 1.220.
* S S S 8 (Items 1 to 8): Factor loadings (all in italics) range from 0.643 to 0.840. C I T C values (all in italics) range from 0.568 to 0.706. Means range from 0.608 to 1.729.
* S A B A S (Items 1 to 6): Factor loadings (all in italics) range from 0.577 to 0.841. C I T C values range from 0.487 to 0.727. Means range from 2.351 to 3.650.
* F O C I (Items 1 to 5): Factor loadings (all in italics) range from 0.785 to 0.903. C I T C values (all in italics) range from 0.704 to 0.795. Means range from 1.329 to 1.577.
* I A T 6 (Items 1 to 6): Factor loadings range from 0.644 to 0.814. C I T C values (all in italics) range from 0.556 to 0.717. Means range from 2.040 to 2.883.
* I S I (Items 1 to 7): Factor loadings range from 0.460 to 0.863. C I T C values range from 0.412 to 0.743. Means range from 1.178 to 2.101.
* B M S L S S (Items 1 to 6): Factor loadings range from 0.631 to 0.893. C I T C values (all in italics) range from 0.577 to 0.778. Means range from 4.605 to 5.152.
Note: GAD = severity measure for General Anxiety Disorder; PHQ-9 = Patient Health Questionnaire-9; SSS-8 = Somatic Symptom Scale-8; FOCI = Florida Obsessive-Compulsive Inventory; SABAS = Smartphone Application-Based Addiction Scale; IAT-6 = Internet Addiction Test-6; BMSLSS = Brief Multidimensional Students’ Life Satisfaction Scale; ISI = Insomnia Severity Index; Skew/Kurt = skewness/kurtosis.
The results of the correlational analyses revealed that there were significant relationships between all the measures of this study. BMSLSS and cognitive fatigue of PedQL related negatively with all the other measures (ps < 0.001), except between BMSLSS and cognitive fatigue of PedQL (r = 0.349, p < 0.001). Figure 1 provides further details.
Correlations between the measures and two external criterion measures (cognitive fatigue [PedsQL_CF] and stress [DASS_Stress]).
Note: ***p < 0.001. PHQ, Patient Health Questionnaire; GAD, Severity Measure for Generalised Anxiety Disorder; SSS, Somatic Symptom Scale; SABAS, Smartphone Application-Based Addiction Scale; IAT, Internet Addiction Test; BMSLSS, Brief Multidimensional Students’ Life Satisfaction Scale; ISI, Insomnia Severity Index; FOCI, Florida Obsessive–Compulsive Inventory; DASS_Stress, stress subscale of Depression, Anxiety, Stress Scale for Youth; PedsQL_CF, cognitive fatigue subscale of PedsQL Multidimensional Fatigue Scale.
Figure 1. Long description
A 10 by 10 correlation matrix heatmap. The vertical y-axis and horizontal x-axis list the same variables in order: P H Q, G A D, S S S, S A B A S, I A T, B M S L S S, I S I, F O C I, D A S S underscore Stress, and Peds Q L underscore C F.
* The diagonal from top-left to bottom-right consists of empty white squares where variables intersect with themselves.
* Positive correlations are shaded in blue, with darker blue representing stronger correlations. The strongest positive correlation is between G A D and D A S S underscore Stress at 0.843.
* Negative correlations are shaded in red or orange. B M S L S S and Peds Q L underscore C F show consistent negative correlations with most other variables. For example, P H Q and Peds Q L underscore C F have a correlation of minus 0.585.
* All displayed numerical values are followed by three asterisks, indicating statistical significance at p less than 0.001.
* The B M S L S S row and column are predominantly red, indicating it is inversely related to the other psychological distress measures, except for a positive correlation of 0.349 with Peds Q L underscore C F.
Discussion
The present study examined the preliminary psychometric properties of GAD-7, PHQ-9, SSS-8, FOCI, ISI, SABAS, IAT-6 and BMSLSS among Hong Kong older youths and the relationships between these scales. It was observed that the study participants had diverse backgrounds, but the majority of the participants (i.e., residents) were originally from the Chinese Mainland, followed by Hong Kong (i.e., natives) and other parts of the world (i.e., Asia, Australia, Africa, Europe and North America). Importantly, the sample characteristic data based on the measures revealed substantial mental health concerns among participants, with 30.9% potentially experiencing depression, 48.1% potentially experiencing anxiety, 44.6% potentially experiencing somatic symptoms, 41.8% potentially experiencing SABAS, 44.8% potentially experiencing obsessive-compulsive symptoms, 34.1% potentially experiencing problematic internet use and 67.8% potentially experiencing insomnia. This indicates that more youths may be going through various mental health conditions and so there is a need for mental health education, screening and effective coping strategies among them (Storch et al., Reference Storch, Bagner, Merlo, Shapira, Geffken, Murphy and Goodman2007; Yu et al., Reference Yu, Tam, Wong, Lam and Stewart2012; Gierk et al., Reference Gierk, Kohlmann, Kroenke, Spangenberg, Zenger, Brähler and Löwe2014; Wong et al., Reference Wong, Lau, Espie, Luik, Kyle and Lau2017; Ali et al., Reference Ali, Hendawy, Almarwani, Alzahrani, Ibrahim, Alkhamees and Kunugi2021; Ip et al., Reference Ip, Suen, Hui, Wong, Chan, Lee, Wong and Chen2022; Peng et al., Reference Peng, Chen, Ren, Liu, He, Liang, Tan, Tang, Chen and Liao2023).
In general, the validity (i.e., factorial and construct) and reliability (i.e., Cronbach’s α and McDonald’s ω) indices for the eight scales were found to be acceptable. Moreover, there were significant relationships between all the variables (including the scales). Apart from the good reliability indices from all the scales, the results of the CFA analysis revealed that generally, the eight scales (i.e., GAD-7, PHQ-9, SSS-8, FOCI, ISI, SABAS, IAT-6 and BMSLSS) had acceptable one-factor structure. Specifically, the DSM-5-adapted version of PHQ-9 was found to have a stable unidimensional factor structure indicative of a single latent construct that assesses depression among participants, similar to previous studies (see Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a; Leung et al., Reference Leung, Mak, Leung, Chiang and Loke2020a). The alternative bifactor structures were also stable, which indicates that it may also be used to assess two latent constructs (i.e., cognitive/affective and somatic) of depression. This indicates that youths residing in Hong Kong may use this scale to assess their general depression levels or its two sub-domains – cognitive/affective and somatic (Boothroyd et al., Reference Boothroyd, Dagnan and Muncer2019; Blake et al., Reference Blake, Munyombwe, Fischer, Quinn, Van der Feltz-Cornelis, De Man-van Ginkel, Santos, Jeon, Köhler, Schram, Wang, Levin-Aspenson, Whooley, Hobfoll, Patten, Simning, Gracey and Broomfield2025; Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a). Furthermore, the GAD-7 was found to have a single latent construct for assessing anxiety, similar to previous studies (Adjorlolo, Reference Adjorlolo2019; Ip et al., Reference Ip, Suen, Hui, Wong, Chan, Lee, Wong and Chen2022; White and Karr, Reference White and Karr2023). Similar to previous studies (e.g., Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a; Chen et al., Reference Chen, Strong, Lin, Tsai, Leung, Lin, Pakpour and Griffiths2020; Leung et al., Reference Leung, Pakpour, Strong, Lin, Tsai, Griffiths, Lin and Chen2020b), the SABAS had a stable one-factor structure, which indicates that it assesses a single latent construct. Additionally, SSS-8 had a stable one-factor, three-factor and four-factor structures indicating that it can be used to assess a single latent construct (i.e., general somatic symptoms) as shown in previous studies (see Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a; Wu et al., Reference Wu, Wang, Ji, Zhu and Guo2024), three (i.e., gastrointestinal and fatigue, pain, and cardiopulmonary symptoms) or four (i.e., gastrointestinal, pain, cardiopulmonary and fatigue) latent constructs based on the purpose of the user (Li et al., Reference Li, Wei, Fritzsche, Toussaint, Zhang, Zhang, Chen, Wu, Ma, Li, Ren, Lu and Leonhart2022; Hybelius et al., Reference Hybelius, Kosic, Salomonsson, Wachtler, Wallert, Nordin and Axelsson2025; Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a).
Also, the DSM-5-adapted version of FOCI was found to have a fairly stable unidimensional factor structure indicative of a single latent construct, as found in Ahorsu et al. (Reference Ahorsu, Oppong and Lin2026a). Moreover, IAT-6 had a fairly stable single latent construct for assessing internet addiction, confirming the findings of previous studies (see Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a; Lai et al., Reference Lai, Mak, Watanabe, Ang, Pang and Ho2013). Likewise, the ISI had fairly stable one-factor and two-factor structures indicative of a single latent construct or two latent constructs (i.e., ISI/nighttime sleep problem and insomnia impact/daytime impairment) for assessing ISI among youths, confirming the findings of previous studies (see Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a; Chung et al., Reference Chung, KK- and Yeung2011; Otte et al., Reference Otte, Bakoyannis, Rand, Ensrud, Guthrie, Joffe, McCurry, Newton and Carpenter2019; Wang et al., Reference Wang, Shen, Guo, Liu, Chen, Ma and Jian2025a). The BMSLSS also had a fairly stable unidimensional factor structure indicative of a single latent construct for assessing life satisfaction among student participants, similar to previous studies (Fang et al., Reference Fang, Yuen, Fung, Zhang, Chan and Wu2022; Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a). However, as the RMSEA indices of FOCI, IAT-6, ISI and BMSLSS are above the acceptable limit, care should be taken when using them across multiple ethnic groups in Hong Kong.
All the relationships between the variables were significant. There were significant positive relationships between the variables except between BMSLSS and PHQ-9, GAD-7, SSS-8, SABAS, IAT-6, ISI, FOCI and stress (i.e., from DASS-Y) and also between cognitive fatigue (i.e., from PedsQL-MFS) and PHQ-9, GAD-7, SSS-8, SABAS, IAT-6, ISI, FOCI and stress (i.e., from DASS-Y), which had negative relationships. A positive relationship suggests that as one variable increases, the other also increases, while a negative relationship suggests that as one variable increases, the other decreases. Specifically for the negative relationships, as depression, anxiety, somatic symptoms, problematic use of (or addiction to) smartphone applications, problematic use (or addiction to) the internet, insomnia and obsessive-compulsive symptoms increase, there is the likelihood that there would be poor life satisfaction and severe cognitive fatigue. These significant relationships, especially with stress and cognitive fatigue, establish potential convergent validity of the scales in addition to the demonstrated construct validity. Furthermore, the CR and AVE indices were well above the acceptable threshold, which supports the convergent validity indicated by the relationships. These findings are similar to those reported among homogeneous populations, unlike the present study, which examined a heterogeneous population (Wong et al., Reference Wong, Mo, Potenza, Chan, Lau, Chui, Pakpour and Lin2020; Wang et al., Reference Wang, Wong, Wang, Li, Kim and Lee2021; Huang et al., Reference Huang, Ahorsu, Adjaottor, Addo, Griffiths, Pakpour and Lin2024b; Yuan et al., Reference Yuan, Wang, Shi, Zhao, Tao, Wu, Tan and Ouyang2025; Zhang et al., Reference Zhang, Tao, Zhang, Zhou, Wei, Chen, Hu, Zheng and Wang2025; Ahorsu et al., Reference Ahorsu, Oppong and Lin2026a). Therefore, researchers can carefully use English versions of these scales to examine mental health conditions among youths residing in Hong Kong.
Practical implications
This study holds practical implications for clinical and research practice. Clinically, the scales can be used by healthcare professionals to screen youths’ mental health conditions. In terms of research, the scales can be used to examine the mental health conditions of youths with different ethnic backgrounds residing in Hong Kong. This can be done in conjunction with healthcare professionals to monitor different mental health conditions over time. Such research can also be performed in combination with digital platforms/applications (e.g., internet addiction, SABAS and life satisfaction).
Limitations
Despite its value, the study has some limitations. Firstly, self-report measures were used to collect the data, an approach known to be susceptible to social desirability and recall bias. Secondly, a cross-sectional design was used for this study, which limits the changes reflected in fluctuating symptoms and across-time stability as would have been observed in longitudinal studies. Future studies can address these limitations by adopting a longitudinal research design. Thirdly, the majority of the scales cannot capture the full complexity of a mental disorder; hence, clinicians must complement them with their experience and expertise. Fourthly, the participants used were predominantly females, mostly university-educated, more digitally engaged, more comfortable with English and older youths; hence, readers should be cautious about extending the findings beyond the sample characteristics. These substantial differences limited detailed subgroup analysis such as measurement invariance across groups. Fifthly, the present psychometric validation is only preliminary, and thus full validation (e.g., test–retest reliability, measurement invariance, criterion validity, discriminant validity) would be needed among youths and across subgroups.
Conclusion
The present study examined the 1) preliminary psychometric properties of GAD-7, PHQ-9, SSS-8, FOCI, ISI, SABAS, IAT-6 and BMSLSS and 2) the relationships between these scales among youths residing in Hong Kong. In general, the validity (i.e., factorial and construct – convergent validity) and reliability (i.e., Cronbach’s α and McDonald’s ω) indices for the eight scales were found to be fairly acceptable. Also, there were significant relationships between all the variables, emphasising the interdependence of all the measures used in the study. Healthcare professionals (e.g., counsellors, health psychologists, clinical psychologists) and researchers may use these validated scales to promote mental health activities among youths in Hong Kong.
Open peer review
To view the open peer review materials for this article, please visit http://doi.org/10.1017/gmh.2026.10302.
Acknowledgements
None.
Author contribution
D.K.A.: Conceptualisation, Data curation, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualisation, Writing – original draft and Writing – review & editing.
Competing interests
The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Data availability statement
The data that support the findings of this study are available from the corresponding author, (DKA), upon reasonable request.
Ethics and approvals
The Human Research Ethics Committee (HREC) at The Education University of Hong Kong approved this study (approval: Ref. no. 2022-2023-0462) on January 17, 2024.
Funding statement
The author disclosed receipt of financial support for this research from The Education University of Hong Kong’s Start-up Research Grant for Newly Recruited Academic Staff (grant number RG43/23-24R).
Inclusivity statement
None.
Open access
