Psychometric Properties of the Smartphone Addiction Scale—Short Version Among Nursing Students in Greece

Abstract

Background: Problematic smartphone use has been increasingly reported among university students, including nursing students, yet the availability of brief, culturally appropriate, and psychometrically sound assessment instruments in Greece remains limited. Aim: This study aimed to translate and culturally adapt the Greek version of the Smartphone Addiction Scale—Short Version (SAS-SV) and to evaluate its psychometric properties, including internal structure, reliability, and convergent validity with the Mobile Phone Problem Use Scale-10, among nursing students in Greece. Methods: In a cross-sectional study, nursing students from multiple departments across Greece (N = 331) completed the Greek SAS-SV, distributed online via official university forums, student groups, and institutional social media pages, between September 2025 and November 2025. We conducted exploratory and confirmatory factor analyses using polychoric correlations, examined convergent validity, performed exploratory comparisons across demographic characteristics, and estimated the reliability of the SAS-SV. Results: Confirmatory factor analysis was consistent with a one-factor structure and showed acceptable model fit. Internal consistency was high (Cronbach’s α was 0.862 and McDonald’s omega was 0.891), with supportive evidence of convergent validity through its correlation with the MPPUS-10 (Spearman’s ρ = 0.772, p < 0.001). Conclusions: The Greek SAS-SV showed acceptable psychometric properties among nursing students and seems appropriate for research purposes in Greece.

1. Introduction

Problematic smartphone use has become a significant public health concern in modern society, particularly among young adults and university students [1]. The versatility of smartphones—serving as communication tools, sources of information, entertainment devices, and social networking platforms—has fundamentally transformed daily life and social interactions [2]. While these technological advances offer notable benefits for productivity, communication, and access to information, they also create conditions that promote excessive and problematic usage patterns [3]. The World Health Organization has included “gaming disorder” in the ICD-11, highlighting increased awareness of certain behavioral disorders related to technology [4]. However, this inclusion specifically addresses gaming disorder and should not be interpreted as recognizing or diagnosing “smartphone addiction” as a separate clinical disorder; therefore, it is more appropriate to refer to it as “problematic smartphone use,” even though “smartphone addiction” remains a common term in the literature.

Problematic smartphone use (often called “smartphone addiction” in research) is seen as a behavioral pattern marked by a loss of control over smartphone use, ongoing excessive usage despite awareness of negative effects, and prioritizing smartphone use over other valued activities and relationships [5]. The literature often notes similarities with substance-related addictions (e.g., increased use and distress when unable to access the phone), but the concept and its behavioral mechanisms are still debated, and language should avoid suggesting there is a definitive clinical consensus [6]. The anxiety and psychological distress experienced when separated from one’s mobile device is known as “nomophobia” (no-mobile-phone phobia), highlighting the psychological dependence that can develop [7]. Research consistently shows meaningful connections between problematic smartphone use and negative mental health outcomes, including anxiety, depression, sleep disturbances, impaired attention and inhibitory control, and lower academic performance [8,9]. Additionally, problematic smartphone use has been linked to impulsivity, reduced social skills, emotional intelligence deficits, and interpersonal issues, suggesting that excessive use can have serious negative effects across various aspects of functioning [10].

University students, especially nursing students, face a higher risk of problematic smartphone use due to various overlapping developmental and occupational factors. This stage of life involves increased exposure to academic and social pressures, and many students turn to smartphones to cope, distract themselves, or regulate emotions [11]. Nursing students are particularly vulnerable because they navigate the dual challenges of demanding coursework and intensive clinical training in high-stress healthcare environments characterized by emotional labor, physical demands, and significant responsibility for patient care [12]. Transitioning into university nursing programs requires substantial social, structural, and psychological adjustments, as students must establish independence, develop professional identities, adapt to new social settings, and learn complex technical and interpersonal skills essential for clinical practice [13]. These combined demands highlight the importance of assessing problematic smartphone use among nursing students, as negative impacts can affect their personal well-being, academic performance, professionalism, and clinical learning [8]. Therefore, it is crucial to identify and measure problematic smartphone use within this group to develop targeted interventions and support strategies tailored to the specific needs and vulnerability of nursing students [10,14].

Epidemiological evidence shows that problematic smartphone use among nursing students is a widespread issue with notable variation across different countries, influenced by cultural differences, healthcare systems, measurement methods, and population traits. A comprehensive scoping review [15], which examined 39 studies from 15 countries, found that smartphone addiction or problematic use among nursing students ranged from 19% to 72%, with an average prevalence of 40–50%. Due to this international variability, estimates from single studies in individual countries should be interpreted with caution, as prevalence can vary greatly depending on the measurement tools, cutoff points, and sampling methods.

Problematic smartphone use—commonly called “smartphone addiction”—is best seen as an umbrella term that describes dysregulated, hard-to-control smartphone engagement that is persistent and connected to distress and/or functional impairment, rather than a single, uniform clinical condition. Modern theories emphasize that problematic smartphone use is varied, with similar outward signs possibly stemming from different psychological mechanisms [16]. In this context, the pathways perspective indicates that problematic use can develop through different routes, such as excessive reassurance seeking or poor self-regulation, which can lead to various problematic symptoms (e.g., addiction-like signs) [17]. This conceptual difference is important for measurement. Therefore, to accurately identify, evaluate, and treat cases of problematic smartphone use, validated and reliable psychometric tools are essential [10].

In recent years, researchers have developed many assessment scales to measure smartphone addiction and problematic mobile phone use, reflecting the growing recognition of this phenomenon as a public health and psychosocial issue. A systematic review by Harris et al. [10] analyzed 78 validated scales created over 13 years to evaluate problematic smartphone use. It found that, despite the large number of assessment tools, many published scales lack proper internal consistency, test–retest reliability, and empirical support for their theoretical basis. Importantly, these tools are not interchangeable because they measure different aspects of problematic smartphone or mobile phone use. Some tools mainly serve as screens for “addiction-like symptoms” (e.g., the Smartphone Addiction Scale—SAS [2] and the Smartphone Application-Based Addiction Scale—SABAS) [18], emphasizing loss of control and related features. Others explicitly include behavior and context domains such as “dangerous use” and “prohibited components” (e.g., the Problematic Mobile Phone Use Questionnaire [PMPUQ] and its shorter version) [19], which can affect prevalence estimates and cross-study comparisons. This variety of instruments, while highlighting the increasing acknowledgment of problematic smartphone use in research, underscores the need for proper validation of existing tools across different populations and cultural contexts. The Smartphone Addiction Scale (SAS) contains 33 items across six dimensions: daily-life disturbance, positive anticipation, withdrawal, cyberspace-oriented relationships, overuse, and tolerance. Later, the same authors created the Smartphone Addiction Scale—Short Version (SAS-SV), a 10-item brief yet comprehensive assessment designed for adolescents and young adults. The SAS-SV captures core addiction features, including loss of control and functional impairment, and has been widely used across various international samples. Other common tools include the Mobile Phone Problem Use Scale (MPPUS), originally developed by Bianchi and Phillips [20] as a 27-item measure, later shortened to the MPPUS-10 by Foerster et al. [21]. The MPPUS-10 assesses problematic mobile phone use through factors such as dependence, withdrawal, and functional impairment. Additional instruments include the Problematic Use of Mobile Phones (PUMP) Scale [22], a 20-item measure, and the Mobile Phone Addiction Craving Scale (MPACS) [23], an 8-item scale.

Among the available tools, the SAS-SV was selected as the primary measure for this study because it aligns with the operational definition of problematic smartphone use as a multidimensional pattern of impaired control and functional impact, and it has published evidence supporting its measurement properties. While the literature includes many instruments with varying psychometric support, the SAS-SV stands out for having clear validation evidence across several international contexts, including Asian, European, and Latin American populations [24,25,26,27]. Therefore, it was prioritized. In this context, a significant gap in the literature is the lack of population- and language-specific psychometric evidence for problematic smartphone use measures among Greek nursing students. Greece’s unique cultural and linguistic context, along with the structure of nursing education and clinical training, may influence both the expression of problematic smartphone use and the psychometric performance of measurement tools. Consequently, this study addresses these empirical gaps by conducting a comprehensive psychometric evaluation of the Greek translation of the SAS-SV among nursing students in Greece. Establishing strong evidence for the Greek SAS-SV in this population is essential for research use within Greece and for enabling meaningful international comparisons.

2. Materials and Methods

2.1. Sample and Data Collection

A convenience sample of 331 nursing students from various departments in Greece participated in this cross-sectional study, conducted between September 2025 and November 2025. An invitation to participate was shared through university student platforms and social media channels, including official university forums, student groups, and institutional social media pages, to reach a broad and diverse group of nursing students. All undergraduate nursing students from first to final year who were willing to participate were eligible. Data was collected using an electronic questionnaire created with Google Forms. The first page featured an informed consent form explaining the study’s purpose and providing the first author’s contact details. The typical completion time was approximately 10–15 min. Since the survey link was distributed via open channels and shared within students’ networks, it was not possible to determine how many students received or viewed the invitation; therefore, a formal response rate could not be calculated.

2.2. Sampling Technique

The sample size was planned to ensure stable factor solutions and accurate validity estimates. Assuming a total population of approximately 3000 students across the nine nursing departments in Greece and using a significance level of 0.05 with 95% confidence; the minimum recommended sample size was 249 students. To improve the precision and stability of the psychometric analyses, we ultimately recruited 331 nursing students, exceeding this threshold and providing an adequate participant-to-item ratio for the planned factor analyses. For convergent validity, a sample of N = 331 yields 99% power to detect a correlation of ρ = 0.30, roughly 96% power for ρ = 0.20, and about 78% for ρ = 0.15 (α = 0.05; two-tailed). Overall, the study was well powered for its primary objectives (factor structure, reliability, and convergent validity).

2.3. Measurements

The survey tool included three sections: (a) socio-demographic characteristics, (b) the Smartphone Addiction Scale-—Short Version (SAS-SV), and (c) the short form of the Mobile Phone Problem Use Scale (MPPUS-10).

2.3.1. Socio-Demographic Characteristics

The socio-demographic characteristics questionnaire created for the study included questions on gender, age, year of study, smoking status, and living arrangements.

2.3.2. The Smartphone Addiction Scale—Short Version (SAS-SV)

The Smartphone Addiction Scale—Short Version (SAS-SV), developed by Kwon et al. [2], is a psychometric tool used to measure problematic or addictive smartphone use, mainly among adolescents and young adults. It is a shorter version of the original Smartphone Addiction Scale (SAS) and includes 10 items rated on a 6-point Likert scale, from 1 (strongly disagree) to 6 (strongly agree). Total scores range from 10 to 60, with higher scores indicating a greater risk of problematic smartphone use. The scale assesses key aspects such as loss of control over smartphone use, sleep disruption, functional impairment, and emotional dependence on the device. It has demonstrated high internal consistency (Cronbach’s α > 0.90) and strong psychometric performance across multiple studies and populations. The SAS-SV has been translated and used internationally, with suggested cutoff points of ≥31 for males and ≥33 for females to identify potential problematic use.

2.3.3. The Mobile Phone Problem Use Scale (MPPUS-10)

The Mobile Phone Problem Use Scale (MPPUS-10) is a short version of the original 27-item Mobile Phone Problem Use Scale [20] created to assess problematic mobile phone use. It contains 10 items rated on a 10-point Likert scale from 1 (not true at all) to 10 (extremely true), resulting in total scores between 10 and 100, with higher scores indicating more severe problematic use. It has been validated among Greek students, and the Greek version uses a Likert scale from 4 (strongly agree) to 1 (strongly disagree) [28]. The MPPUS-10 has shown good internal consistency (Cronbach’s α = 0.85) and strong correlations with the full MPPUS, supporting its reliability and convergent validity in adolescent and young adult populations.

2.4. Translation and Cultural Adaptation

The SAS-SV was translated and culturally adapted following guidelines for cross-cultural adaptation of self-reported measures [29]. The process included two independent forward translations, combining them into one version, blind back-translation, and review by an expert committee to ensure semantic, idiomatic, and conceptual equivalence. Cognitive debriefing interviews with 10 nursing students (about 5–10 min each), who were not part of the study’s sample were then conducted. Participants were asked to comment on item clarity, wording, relevance, interpretability, and cultural appropriateness, and to indicate whether any item was ambiguous or conceptually inconsistent with the intended meaning of the original version. No major comprehension or cultural-equivalence problems were identified; therefore, no wording changes were required.

2.5. Ethical Considerations

The current study established ethical standards, protecting participants’ rights, privacy, and confidentiality throughout the research. All participants were informed about the voluntary nature of their involvement and the time required to complete the survey. They were also told they could withdraw at any time without consequences. The study was conducted following the principles of the Declaration of Helsinki and received ethical approval from the Institutional Review Board of Democritus University of Thrace (Approval 55705/475).

2.6. Statistical Analysis

Descriptive statistics summarize sample characteristics. Quantitative variables were reported as means and standard deviations (SD), while categorical variables were presented as frequencies and percentages. The item-level distribution of the SAS-SV (means, SD, skewness, kurtosis) was analyzed; since several items showed deviations from normality, all items were treated as ordinal.

To explore the underlying structure of the SAS-SV, factorability was first assessed using the Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity. To avoid arbitrary factor retention, we applied parallel analysis and Velicer’s Minimum Average Partial (MAP) test to determine the optimal number of factors. Exploratory factor analysis (EFA) was then conducted using principal axis factoring with oblimin rotation. Items with factor loadings below 0.40 or with significant cross-loadings were excluded from the final model. The factor structure identified through EFA was subsequently tested using confirmatory factor analysis (CFA). Since indicators were ordinal, CFAs were estimated on polychoric correlation matrices with a robust diagonally weighted least squares estimator (DWLS/WLSMV). Model fit was evaluated using the chi-square to degrees of freedom ratio (χ²/df), Comparative Fit Index (CFI), Tucker–Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA), and Standardized Root Mean Square Residual (SRMR), along with standardized factor loadings. Measurement invariance across gender was examined using multi-group CFA in a staged sequence of configural and metric models. Because the items were treated as ordered categorical indicators, strong/scalar invariance would require equality constraints on thresholds in addition to factor loadings; however, the corresponding model could not be estimated stably and was therefore not retained. The internal consistency of the SAS-SV was assessed using McDonald’s omega (ω). Convergent validity was examined through Spearman’s rank correlation (ρ) between SAS-SV total scores and MPPUS-10 total scores, including 95% bootstrap confidence intervals. Exploratory comparisons of SAS-SV scores across demographic characteristics (sex, year of study, and smoking status) were conducted using appropriate parametric or non-parametric tests, with effect sizes reported. All analyses were performed using IBM SPSS Statistics, Version 25 (IBM Corp., Armonk, NY, USA) for descriptive and inferential statistics, and JASP for EFA and CFA. Statistical significance was set at p < 0.05 (two-tailed).

3. Results

A total of 331 nursing students participated in the study. Most were women (74.5%), in their first year (47.5%), living independently (63%), and non-smokers (67.6%). The average age was 23.1 years (±7.79). The demographic details of the participants are shown in

Table 1. Socio-demographic characteristics.

Variables	n (%)
Gender
Male	85 (25.5%)
Female	246 (74.5%)
Age, years	23.1 ± 7.79
Living arrangement
Live alone	208 (63%)
Live with my partner/family	123 (37%)
Smoking status
Yes	107 (32.4%)
No	224 (67.6%)
Year of study
1st	154 (47.5%)
2nd	36 (9%)
3rd	42 (13%)
4th	99 (30.6%)

.

Descriptive statistics for the SAS-SV scores are shown in

Table 2. Descriptive statistics for the SAS-SV.

Items	Mean	SD	Skewness	Kurtosis
Item 1	2.41	1.18	0.585	−0.204
Item 2	2.85	1.28	0.462	−0.108
Item 3	2.52	1.48	0.627	−0.756
Item 4	3.24	1.62	0.176	−1.059
Item 5	2.42	1.37	0.806	−0.192
Item 6	2.22	1.51	0.976	0.616
Item 7	2.77	1.39	0.629	−0.107
Item 8	2.50	1.39	0.591	−0.805
Item 9	3.64	1.37	−0.108	−0.563
Item 10	2.44	1.35	0.883	0.192
SAS-SV total	29.30	9.69

Note: Items scored on a 6-point Likert scale; higher scores indicate greater problematic smartphone use.

. The observed item scores ranged from 2.22 (±1.51) to 3.64 (±1.37). The total SAS-SV score averaged 29.30 (±9.69). According to previously proposed sex-specific cut-offs (≥31 males, ≥33 females), 18.9% of this convenience sample met criteria for potential problematic smartphone use (6.2% male, 12.7% female). This proportion should be interpreted descriptively and not as a population prevalence estimate. The SAS-SV items exhibited mild positive skewness and positive platykurtosis, which is consistent with 6-point ordinal ratings, but they did not significantly deviate from normality. Due to their ordinal nature, the items were analyzed as ordered categorical indicators using polychoric correlations and a robust DWLS/WLSMV estimator.

The factorability of the SAS-SV items was initially assessed to determine their suitability for exploratory factor analysis. The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy was 0.792, indicating acceptable sampling adequacy. Bartlett’s test of sphericity was statistically significant, χ²(45) = 294.02, p < 0.001, supporting the factorability of the correlation matrix. The determinant of the item correlation matrix was 0.045. Although most items showed adequate shared variance, one item had a low communality (h2 = 0.19), suggesting it is less strongly related to the overall factor. However, considering the established content of the SAS-SV and the adequate factor structure seen in our sample, the item was kept to ensure coverage of the construct and comparability across studies. Parallel analysis and Velicer’s MAP indicated that a single factor should be retained. This factor was extracted using principal axis factoring; the first factor had an eigenvalue of 4.34 and explained 43.4% of the total variance. Loadings on this general problematic smartphone use factor ranged from 0.433 to 0.755, with the highest loadings for items SAS-SV 6, SAS-SV 7, SAS-SV 5, SAS-SV 10, and SAS-SV 4 (0.702–0.755), and slightly lower but still acceptable loadings for SAS-SV 2, SAS-SV 8, SAS-SV 9, SAS-SV 1, and SAS-SV 3 (0.433–0.633). No items needed to be removed, and the unidimensional 10-item solution was maintained for reliability testing and future confirmatory analyses. The factor loadings are presented in

Table 3. Exploratory factor analysis of the Smartphone Addiction Scale—Short Version (SAS-SV).

Items	Smartphone Addiction Scale (Factor 1 Loading)	Communalities
Item 1	0.57	0.33
Item 2	0.63	0.40
Item 3	0.43	0.19
Item 4	0.70	0.49
Item 5	0.74	0.54
Item 6	0.76	0.57
Item 7	0.75	0.55
Item 8	0.62	0.39
Item 9	0.61	0.37
Item 10	0.71	0.51

Note: Extraction method: principal axis factoring; factor retention supported by parallel analysis and Velicer’s MAP.

.

A one-factor confirmatory factor model was specified for the SAS-SV items and estimated using a DWLS estimator on the polychoric correlation matrix. Overall model fit was acceptable, although it did not provide especially strong support for a brief unidimensional scale: χ²(35) = 119.58, p < 0.001, CFI = 0.944; TLI = 0.927; RMSEA = 0.081 (90% CI [0.071, 0.136]; SRMR = 0.067; GFI = 0.987; MFI = 0.904. All standardized factor loadings were positive and statistically significant (p < 0.001), ranging from 0.44 to 0.76, with the lowest loadings for item SAS-SV 3 and the highest for item SAS-SV 6. Corresponding R2 values ranged from 0.20 to 0.58, and standardized residual variances from 0.42 to 0.81, aligning with the magnitude of loadings. Internal consistency of the SAS-SV was high, with McDonald’s ω = 0.891 and Cronbach’s α = 0.862. Based on the standardized loadings, the average variance extracted (AVE) for the general factor was 0.45, and the composite reliability (CR) was 0.89. Although the CR value was high, the AVE of 0.45 indicates that the general factor explained slightly less than half of the item variance on average; therefore, this finding supports acceptable but not especially strong item-level convergence.

As expected, SAS-SV total scores were positively associated with problematic mobile phone use as measured by the MPPUS-10. In this sample, the MPPUS-10 showed good internal consistency with McDonald’s ω = 0.792, and Cronbach’s α = 0.787. The correlation between SAS-SV and MPPUS-10 totals was ρ = 0.772 (95% CI [0.71, 0.82]), p < 0.001, indicating strong convergent validity between problematic smartphone use and problematic mobile phone use.

Measurement invariance was examined across gender and year-of-study group using multi-group CFA in a staged sequence of configural and metric models. For gender, the configural model showed good fit (CFI = 0.993, TLI = 0.995, RMSEA = 0.046, SRMR = 0.079). Constraining factor loadings to equality did not worsen fit (metric model: CFI = 0.993, TLI = 0.995, RMSEA = 0.046, SRMR = 0.079, ΔCFI = 0.000, ΔRMSEA = 0.000), supporting metric invariance across gender. For year-of-study group (junior vs. senior), the configural model showed moderate fit (CFI = 0.991, TLI = 0.993, RMSEA = 0.050, SRMR = 0.081). Constraining factor loadings to equality again did not worsen fit (metric model: CFI = 0.991, TLI = 0.993, RMSEA = 0.050, SRMR = 0.081, ΔCFI = 0.000, ΔRMSEA = 0.000), supporting metric invariance across year-of-study group. Because the scalar invariance model for the ordered categorical indicators could not be estimated stably, only configural and metric invariance results are reported (

Table 4. Measurement invariance of the Greek SAS-SV across gender and year-of-study group.

Grouping Variable	Model	CFI	TLI	RMSEA	SRMR	ΔCFI	ΔRMSEA	Interpretation
Gender	Configural	0.993	0.995	0.046	0.079	-	-	Good fit
Gender	Metric	0.993	0.995	0.046	0.079	0.000	0.000	Metric Invariance supported
Year-of-study group	Configural	0.991	0.993	0.050	0.081	-	-	Moderate fit
Year-of-study group	Metric	0.991	0.993	0.050	0.081	0.000	0.000	Metric Invariance supported

Note: Strong/scalar invariance was not retained because the corresponding model for the ordered categorical indicators could not be estimated stably.

).

Mann–Whitney U tests revealed no statistically significant differences in SAS-SV total scores between male and female students (U = 3373, Z = −0.64, p = 0.524). Likewise, there were no significant differences between junior and senior students (U = 3992, Z = −0.95, p = 0.314). These findings suggest that, in this sample, the risk of problematic smartphone use does not differ systematically based on gender or year of study. A small difference was observed between smokers and non-smokers (U = 3275, Z = −2.00, p = 0.045, r = 0.11); given the small effect size and exploratory nature of these comparisons, this finding should be interpreted cautiously.

4. Discussion

The current study aimed to translate, culturally adapt, and assess the psychometric properties of the Greek version of the SAS-SV in a sample of nursing students in Greece. Overall, findings provide evidence that the Greek SAS-SV produces reliable and valid scores in this population, supporting its use as a brief measure of problematic smartphone use risk in Greek nursing education and research settings.

Regarding the factor structure, both exploratory and confirmatory analyses were consistent with a one-factor solution for the SAS-SV in this sample. However, the CFA indices provide acceptable rather than especially strong support for unidimensionality. This finding is broadly consistent with previous studies that have also supported a predominantly unidimensional structure for the SAS-SV. In addition, Item 3 showed the weakest psychometric performance in the present sample, suggesting that it may reflect the underlying construct less strongly than the other items. Nevertheless, it was retained to preserve content coverage and comparability with prior studies using the full 10-item SAS-SV [24,25,26,27,30]. The present study provides, for the first time, a comprehensive psychometric evaluation of the SAS-SV using a sample of nursing students. Although the SAS-SV has been widely used in studies involving nursing students, published research often relies on previously validated language versions and offers limited reporting on measurement within nursing-student samples. Additionally, methodological reviews highlight heterogeneity in language versions and whether local cultural adaptation or revalidation has been documented [31]. To date, we have not identified a psychometric evaluation of the SAS-SV specifically in Greek nursing students, representing a significant measurement gap for research and assessment in this population.

The study findings showed that the SAS-SV had high internal consistency in our sample, supporting the reliability of the total score in Greek nursing students, with Cronbach’s alpha and McDonald’s omega values both exceeding 0.85. Previous studies with nursing student samples also indicate that the SAS-SV demonstrates high internal consistency, with reported Cronbach’s alpha values around 0.87–0.88 [32,33]. Additionally, studies assessing test–retest reliability have reported high intraclass correlation coefficients (ICCs above 0.80 and even 0.90), further confirming the scale’s temporal stability and overall reliability [25,34].

Based on the results, although the mean SAS-SV score was below the scale midpoint, 18.9% of this convenience sample exceeded previously proposed sex-specific cut-offs, indicating a subgroup with elevated risk of problematic smartphone use. Since cut-off score classification is threshold-based, it can identify a high-risk minority even when the overall mean remains below the midpoint. In a recent scoping review on nursing students’ smartphone addiction [15], approximately 40–50% of nursing students showed signs of problematic smartphone use, with a slightly higher tendency observed among females, as also noted in our study.

In our sample, SAS-SV scores did not differ significantly by gender or year of study, indicating that problematic smartphone use did not appear to differ systematically across these subgroups among Greek nursing students. This interpretation is broadly consistent with the measurement invariance analyses, which supported configural and metric invariance across gender and year-of-study group, although the fit for year-of-study group was less strong and strong/scalar invariance was not established. Therefore, subgroup comparisons should be interpreted cautiously. Evidence from student populations varies and generally supports that gender differences are not always clear when measurement is comparable. For example, a gender invariance study in Chinese university students found that the SAS-SV could be used for cross-gender comparisons, with no significant gender differences in addictive symptoms in their sample [26]. Similarly, no gender differences were reported in nursing-student research in Turkey [33,35]. However, a study in Iraq showed highly significant gender differences in problematic smartphone use among nursing students [36].

Meanwhile, the absence of differences by year of study should be viewed cautiously, as the nursing-student literature is mixed—some studies report differences in SAS-SV scores across years, while others do not [36,37]. A likely explanation for our findings is that smartphone use is deeply ingrained in students’ daily routines across all years, involving communication, study-related activities, and stress management; the academic and clinical demands that could lead to problematic use are probably present throughout training rather than limited to a specific year. Additionally, considering the high heterogeneity observed in nursing-student studies using the SAS-SV, subgroup effects may heavily depend on context, such as institutional environment, academic workload, clinical placement structure, and cultural patterns of use [31].

The convergent validity of the Greek SAS-SV was supported by its relationship with problematic mobile phone use, measured by the MPPUS-10. In our sample, the SAS-SV total score had a high correlation with the MPPUS-10 total score, indicating that students who reported more symptoms of problematic smartphone use also reported higher levels of problematic mobile phone use. This pattern aligns with expectations because both tools aim to assess related maladaptive behaviors in technology use. At the same time, because both measures are self-report instruments assessing closely related forms of problematic phone use, this finding should be interpreted as supportive but relatively narrow convergent validity evidence. Importantly, although the association was strong, it was not perfect, indicating that the two measures are related but not identical. This is consistent with the fact that the SAS-SV specifically targets smartphone-related addiction symptoms, while the MPPUS-10 measures problematic mobile phone use more broadly [20].

In our sample, students who reported current smoking had significantly higher SAS-SV total scores than non-smokers. This finding supports evidence that problematic smartphone use often occurs alongside other health-risk behaviors. Among Spanish undergraduate students, smoking was associated with higher odds of problematic smartphone use assessed with the SAS-SV [38]. Meanwhile, broader research indicates that the links between smoking and problematic smartphone use vary depending on the setting and measurement methods. For instance, a Swiss study found that alcohol and tobacco consumption were not related to problematic smartphone use [39]. Overall, our results suggest a modest co-occurrence of risk behaviors in this group, aligning with frameworks that see problematic behaviors as interconnected rather than isolated. Nevertheless, the effect size was very small, the p-value was marginal, and no adjustment for multiple comparisons was applied, this finding should be considered exploratory rather than confirmatory.

Regarding limitations, convenience sampling may have introduced selection bias, potentially affecting the generalizability of the findings. Also, the study relied on convenience online recruitment and included only nursing students, the findings should not be generalized beyond similar student populations without caution. However, the sample’s gender distribution and key demographic traits were mostly similar to those reported for Greek nursing students overall, suggesting any impact of selection bias on the psychometric evaluation may be limited, especially considering the female dominance in the nursing student population [40,41]. In addition, the sample also appeared demographically heterogeneous in age, despite a large proportion of first-year students, suggesting the inclusion of older or non-traditional students; this should be considered when interpreting generalizability and cutoff-based classifications. All measures were self-reported, which could have led to reporting bias and inflated associations due to common method variance. A further limitation is that the exploratory and confirmatory factor analyses were conducted on the same sample rather than in independent subsamples. Accordingly, the CFA should be interpreted as providing within-sample support for the proposed factor structure, rather than as a fully independent cross-validation of the exploratory findings. Additionally, we did not evaluate criterion validity against an external clinical or behavioral standard; therefore, the cut-off-based prevalence should be viewed as screening rather than diagnostic. Another limitation is that test–retest reliability was not assessed; thus, the temporal stability of the Greek SAS-SV scores in this population remains to be established.

In terms of strength, the present study is the first to provide a comprehensive psychometric evaluation of the Greek SAS-SV in a relatively large, multi-department sample of nursing students, enhancing the robustness and applicability of the findings within Greek nursing education settings. Future studies are necessary to assess the test–retest stability of the Greek SAS-SV. Longitudinal research is also needed to determine whether SAS-SV scores predict sleep quality, academic performance, and psychological distress during clinical training. More broadly, future research should also address the continuing conceptual and methodological debate in this field, as recent expert synthesis has highlighted that the literature on smartphone and social media harms remains fragmented, many claims are still limited by inconsistent or non-causal evidence, and constructs such as behavioral addiction remain contested [42].

5. Conclusions

The present study suggests that the Greek SAS-SV shows acceptable psychometric properties in this sample of nursing students, including internal consistency, evidence for internal structure, and convergent validity. The findings support its use as a brief research instrument for assessing problematic smartphone use in similar Greek nursing-student samples. Further studies should examine test–retest reliability, criterion validity, and performance in broader student and non-student populations.