Abstract
Background
The digital transformation of obesity therapy introduces new demands on exercise therapists. Yet, the therefore required competences remain mostly unassessed, hindering effective implementation. To address this gap, valid instruments are needed to identify competences to plan and implement digital obesity exercise therapy and inform targeted professional training programs. The DigCompThExO model provides a conceptual framework and accompanying questionnaire but has not yet been validated. Therefore, this study examines the factorial structure and psychometric properties to enable its diagnostic use.
Methods
To address the research gap, two consecutive studies were conducted. Study 1 assessed content validity through cognitive interviews with German-speaking digital obesity exercise therapists, using Tourangeau-based probing. Study 2 evaluated structural validity via confirmatory factor analysis (CFA) based on revised questionnaire data from a separate sample within the same target group. Model fit (CFI, TLI, RMSEA, SRMR), validity (AVE, HTMT), reliability (CR), and measurement invariance were evaluated.
Results
Study 1 involved ten therapists (Age: M = 39.3, SD = 10.67, 50% female) in cognitive interviews. Based on results, one dimension and one item were removed; scaling and wording were refined. In Study 2, 205 therapists (Age: M = 32.34, SD = 5.9, 47.3% female, 0.5% divers) completed the revised questionnaire. CFA revealed four latent factors, with good model fit [CFI = .964; TLI = .953; RMSEA = .070, 90% CI (.049;.090); SRMR = .051]. Regarding validity and reliability, measurement invariance was confirmed (RMSEA < .069), along with convergent validity (AVE > .50), discriminant validity (HTMT < .85), and internal consistencies (all CR > .70).
Conclusion
Cognitive interviews and CFA confirm reliability, construct validity, and practical relevance of the DigCompThExO for digital obesity exercise therapy. It enables the assessment of existing and required digital competences and may guide the development of tailored training programs. If embedded in institutional medical education (e.g., university), this could enhance quality and effectiveness of digital obesity exercise therapy. Further research should explore the tool's applicability in other domains of obesity therapy (e.g., nutrition, behavior) or chronic conditions (e.g., diabetes) to promote digital, location-independent healthcare.
Keywords: competence model, curriculum development, digital transformation, exercise therapy, obesity
1. Introduction
Obesity, a multifactorial chronic disease marked by excessive body fat accumulation, is a major global public health issue (1). Projections estimate that by 2030, over one billion people worldwide will be affected (2). Beyond its immediate health impact, obesity is associated with comorbidities (e.g., type 2 diabetes), imposing significant burdens on individuals and healthcare systems (3). Despite evidence-based therapeutic approaches, obesity therapy often fails due to structural barriers including long travel distances, limited local services, and restricted access to specialized treatment centers (4).
Given this problem, digital therapy formats have emerged as a promising solution. They offer time- and location-independent access, reducing infrastructural barriers, and promoting equitable healthcare delivery (5, 6). In line with definitions of digital therapeutics, digital therapy in this study refers to therapeutic interventions delivered via clinically informed, patient-directed digital applications (e.g., video platforms, web-based tools, or mobile software) aimed at supporting the treatment, management, or prevention of disease (7). While initial concepts for digital obesity therapy exist, implementation remains limited. This is partly due to the absence of validated and clearly defined digital competences among therapists (8). Obesity therapy involves a multidisciplinary team of nutritionists, pediatricians, psychologists, and exercise therapists. Although digitalization affects all professional groups, exercise therapists warrant particular attention. Given the pivotal role of exercise in long-term weight management (9), German therapy guidelines assign it as many sessions as the other therapy components (nutrition, behavior, medicine) combined (10). Moreover, exercise therapists must deliver theoretical content and practical exercise sessions, posing specific challenges (e.g., media selection, digital learner support) in digital formats (11), potentially influenced by individual factors like age and gender (12). Beyond these general demands, exercise therapy also requires competences such as digitally demonstrating and correcting movement, adapting exercises to diverse physical abilities, and supporting self-regulated activity through digital monitoring and feedback. These go beyond the scope of general digital competence models, which rarely address motor learning, embodied interaction, or exercise-specific safety in digital settings (13). Addressing these needs requires theory-based, validated models to define competences required for digital obesity exercise therapy (DOET) (14). DOET is understood here, in line with the definition above, as the structured, digitally mediated delivery of exercise-based interventions comprising both physical activity and related educational or behavioral components, specifically targeting people living with obesity. Such competence models enable structured assessment, targeted training and curriculum development, supporting effective digital implementation; particularly for patients in structurally underserved regions.
However, no validated competence models exist for DOET. To address this gap, existing educational frameworks offer a valuable foundation for competence identification through conceptually and content-driven modeling approaches (15, 16). Accordingly, the internationally recognized DigCompEdu has attracted attention as transferable framework originally designed for educators but increasingly relevant for competence development in medical education. It defines digital competences across six dimensions: Professional Engagement, Digital Resources, Teaching and Learning, Assessment, Learning Strategy, and Facilitating Learners’ Digital Competence (17). DigCompEdu is distinguished by its dual methodology: target group input and empirical validation via latent construct analysis. In line with this, current best practices in validation begin with model-driven cognitive interviews to assess content validity. Subsequent factor analyses examine the factor structure and psychometric quality of the instrument (18). Through its methodological triangulation, DigCompEdu serves as diagnostic tool and transferable reference model in educational institutions for the use of digital resources ensuring broad applicability (19). This multi-stage validation approach is also reflected in more recent studies on digital competence, including those conducted in healthcare settings (20, 21).
Accordingly, transferring the DigCompEdu validation approach to DOET (22) requires reference to the only established competence model for exercise therapists in digital obesity therapy (DigCompThExO) by Pawellek and Wulff, with the questionnaire development process detailed elsewhere (23). Conceptually, DigCompThExO followed a two-step lineage: a) DigCompEdu as the overarching theoretical reference framework for digital competence development and b) domain-specific adaptation to the requirements of exercise therapy in digital obesity care. Unlike general digital health frameworks (24), DigCompThExO explicitly integrates didactic-motor aspects such as exercise demonstration, exercise feedback, and self-monitoring; elements largely absent from broader models. It addresses key therapeutic tasks including digital planning, motor instruction, behavior regulation, and patient-centered support in digitally mediated exercise contexts. Based on this foundation, DigCompThExO comprises five dimensions (Selection Criteria, Teaching Strategy, Learning Support, Media Reflection and Media Development) and is operationalized through an 18-item questionnaire. Competence is assessed on a six-point Likert scale, measuring two cognitive competence levels of knowledge and application. Thus, the DigCompThExO represents a novel instrument to identify competences to plan and implement DOET. It has potential to inform training programs, and guide professional development. However, adaptation and validation of the DigCompThExO's factor structure remain pending due to limited empirical groundwork, leaving its diagnostic utility unconfirmed.
Against this background, the central problem arises that, although a theory-based DOET competence questionnaire exists, its interpretive value depends on adaptation to the target group and statistical validation. Therefore, the present study pursued the following obejectives: 1. to examine whether DOET therapists interpret the questionnaire items as intended, thereby supporting content validity; 2. to test whether the hypothesized factor structure of DigCompThExO is confirmed through empirical data; and 3. to evaluate whether DigCompThExO demonstrates valid and reliable psychometric properties.
2. Materials and methods
To address these research questions, selected methods from the validation approach by Boateng and Neilands (18) were applied based on available data and study aims. Validation followed a two-step approach: qualitative cognitive interviews for questionnaire refinement (Study 1), followed by quantitative assessment of internal consistency, composite reliability, discriminant validity, and gender invariance (Study 2). All steps adhered to the Consensus-based Standards for the Selection of Health Measurement Instruments (COSMIN) guidelines for their specificity and international applicability (25).
2.1. Study 1: cognitive interviews
2.1.1. Study Design
To assess content validity and identify comprehension issues, cognitive interviews were conducted with the target group (18), guided by the model of Tourangeau (26). Probing techniques (think-aloud, paraphrasing, process probing) addressed the cognitive stages of comprehension, retrieval, judgment, and response of items. Key foci included item clarity, relevance, and scale design. Data were collected between January and March 2024 in Germany.
2.1.2. Participants
Following cognitive interview guidelines, a sample of five to 15 DOET therapists was defined to balance feasibility and comprehension coverage (27). Recruitment was conducted using a quota sampling approach, drawing from a convenience sample, with participants being contacted through exercise therapy institutions and sports clubs (28). Selection was guided by sociodemographic variables shown in literature to impact competence development (29): age was classified using the threshold proposed by Bachmann, Hertweck (30), distinguishing young and older adults (> 36 years) and gender (male and female).
2.1.3. Administration
Interviews were conducted individually via Zoom by a trained interviewer to ensure procedural consistency. Following informed consent and a mandatory entry test (31), participants reviewed the questionnaire item by item (20–45 min) using a structured guide. All sessions were recorded for analysis.
2.1.4. Data analysis
Interview data were transcribed via Whisper and analyzed using qualitative content analysis (32). A structured codebook guided the coding process, developed based on Tourangeau's cognitive interview framework and refined through peer debriefing with colleagues. Categories were derived deductively from the interview guide and relevant literature, and complemented inductively with data-driven codes. Although coding was conducted by a single researcher, peer debriefing supported methodological transparency and interpretive reliability. Modifications were made when ≥30% of therapists independently identified the same issue, indicating consensus. This threshold aligns with established practices in cognitive interviewing, where recurring issues identified by approximately one third of participants are considered indicative of systematic comprehension problems rather than individual idiosyncrasies (18, 31).
2.1.5. Bias
Potential sources of bias include social desirability effects and interviewer-related assumptions, which may have caused confusion or influenced participants’ responses. Furthermore, recall bias cannot be fully excluded, given the retrospective nature of participants’ professional experiences.
2.2. Study 2: confirmatory factor analysis
2.2.1. Study design
The cross-sectional study was designed to evaluate validity and reliability of the DigCompThExO. Data were collected between May 2024 and January 2025. In addition to competence-related responses, participants provided demographic information (age, gender).
2.2.2. Participants
For statistical validation, a 10:1 respondent-to-item ratio required a minimum of 180 participants based on the DigCompThExO's 18 items; to ensure power and account for exclusions, a target of 200–300 responses was set (33, 34). In line with psychometric guidelines for confirmatory factor analysis, this target range is considered adequate for models of moderate complexity and acceptable item quality (35), particularly given the DigCompThExO's focused four-factor structure and theoretical foundation. Eligible participants were certified exercise therapists (degree, vocational training, or other) with experience in obesity care, and use of any media in therapy. Eligibility was verified using professional background data (e.g., qualification, experience) according to Moosbrugger and Kelava (14). Therapists were recruited via email from therapy and rehabilitation centers in Germany, Austria, and Switzerland, partly identified through listings of certified centers by the German Obesity Society.
2.2.3. Administration
DigCompThExO was administered anonymously online via LimeSurvey. Participants were informed about study objectives, confidentiality, voluntary participation, and lack of financial incentives. Response formats were explained in advance; completion took approximately ten minutes. While participation was recommended via laptop for optimal graphical display, the survey was also accessible on phones and tablets.
2.2.4. Statistical analysis
To assess the structural and construct validity of the DigCompThExO, a Confirmatory Factor Analysis (CFA) was conducted in R (Version 4.4.2) using the lavaan package (36). Following Boateng et al.'s framework (18), CFA was chosen over Exploratory Factor Analysis (EFA) due to the instrument's strong theoretical foundation, predefined structure, and limited item count (18 items), which rendered EFA inappropriate given the low item-to-factor ratio and the associated risk of model instability (35, 37, 38). Model fit was evaluated using the Comparative Fit Index (CFI), Tucker–Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA), Standardized Root Mean Square Residual (SRMR), and the chi-square statistic (χ²/df). Thresholds for acceptable model fit were set at CFI and TLI > .95, and RMSEA and SRMR < .08, in line with recommendations by literature (39). While Hu and Bentler (39) suggest a stricter RMSEA cutoff of .06 for good fit, values up to .08 are commonly considered acceptable (40), particularly in models of moderate complexity and sample size. Thresholds for acceptable fit were CFI and TLI > .95, RMSEA and SRMR < .08. Internal consistency was assessed via Composite Reliability (CR ≥ .70 acceptable), which offers improved accuracy over Cronbach's alpha for digital competence constructs (41). Convergent validity was tested using Average Variance Extracted (AVE > .50), construct validity through correlation analysis (r > .50), and discriminant validity via the Heterotrait-Monotrait Ratio (HTMT < .85) (42, 43). Gender invariance was examined following Vandenberg and Lance (44) using RMSEA ≤ .08 as an indicator of acceptable model fit.
2.2.5. Missing data
Incomplete questionnaires were excluded, rendering missing data handling unnecessary.
2.2.6. Post-hoc analysis
post-hoc model modifications followed the planned validation strategy, guided by modification indices and theoretical coherence. As recommended in CFA, adjustments were only made when theoretically justified, and were clearly reported and cross-validated to ensure robustness (37).
2.2.7. Bias
Potential biases may stem from convenience and self-selection, as therapists were recruited via email from certified centers in German-speaking countries. This likely favored younger and more digitally engaged professionals, limiting generalizability to older therapists or those in less connected or international settings. Future studies should use broader sampling strategies to validate subgroup differences.
3. Results
3.1. Study 1: cognitive interviews
Ten out of 13 invited exercise therapists from Leipzig participated in the interviews (Age: M = 39.3, SD = 10.7, 50% male). Participants had prior experience working with patients affected by obesity.
Findings from the interviews indicated overall alignment with intended item meanings, though minor interpretational differences emerged. Based on Tourangeau's (26) four cognitive stages of item response (comprehension, retrieval, judgment, and response), areas for improvement were identified. The following sections provide a detailed breakdown of each stage.
3.1.1. Comprehension
Participants showed adequate comprehension of the items and could paraphrase them accurately. However, four participants misinterpreted “digital media”, referring only to hardware rather than both hardware and software, an issue further complicated by the term “digital resources” in the Media Development dimension. For example, one participant stated: “When I read digital media, I mainly think of devices like tablets or cameras, not software applications.”
3.1.2. Retrieval
Most participants retrieved relevant information quickly, though five therapists needed reflection for soft- and hardware development examples. Two also struggled to recall data protection strategies. Six therapists were unsure whether the item referred to feedback from therapists to participants or vice versa regarding media use: “I wasn’t sure whether the item refers to my feedback to patients or feedback I receive from them.” This quotation illustrates recurring comprehension ambiguities that guided item revision.
3.1.3. Judgment
While the overall structure of the questionnaire was perceived as intuitive, participants’ ability to judge the relevance and applicability of Media Development dimension revealed uncertainty. Specifically, questions about hardware development were deemed irrelevant to therapeutic work: “I mean, I studied sport science, not computer science.” Five participants raised concerns about software development, though three acknowledged their competence in software adaptation after further reflection.
3.1.4. Response
The division of response options into knowledge (“know it”) and application (“use it”) caused confusion. While the six-point Likert scale for application was perceived as clear, seven therapists viewed knowledge as a dichotomous trait, making its assessment complex. Moreover, they noted that knowledge is a prerequisite for application, rendering its separate measurement partially redundant.
3.1.5. Revisions
The following revisions to the questionnaire were made based on the key cognitive errors:
Clarity: The questionnaire's introduction was adjusted to emphasize that digital media includes both hardware and software, and all related terms (e.g., digital resources) were standardized accordingly. The direction of the feedback was clarified, to enhance recipients’ understanding.
Relevance: Questions regarding hardware development were reduced, so that the Media Development dimension was limited to software development.
Scale units: In the revision process, the original knowledge items (“Know it”) were removed due to limited discriminatory power and high overlap with the corresponding application items, resulting in conceptual redundancy and reduced psychometric utility.
The revised dimensions and items are presented in Table 1.
Table 1.
The DigCompThExO dimensions and included items following cognitive interviews.
| Dimension | Question | Item | Item abbrev. |
|---|---|---|---|
| Selection Criteria (SC) | I use Selection Criteria for… | Software in my DOET area. | SC1 |
| Hardware in my DOET area. | SC2 | ||
| The protection of personal data. | SC3 | ||
| Teaching Strategy (TS) | I use digital media for… | The individual activation of participants. | TS1 |
| The demonstration of practical DOET content. | TS2 | ||
| The explanation of theoretical DOET content. | TS3 | ||
| Providing feedback to participants. | TS4 | ||
| Learning Support (LS) | I use digital media that enable participants to… | Plan their DOET-related behavior. | LS1 |
| Document their DOET-related behavior. | LS2 | ||
| Monitor their DOET-related behavior. | LS3 | ||
| communicate. | LS4 | ||
| Media Reflection (MR) | I reflect before each session with digital media on… | Why I use them in DOET. | MR1 |
| How I use them in DOET. | MR2 | ||
| When I use them in DOET. | MR3 | ||
| For whom I use them in DOET. | MR4 | ||
| Media Development (MD) | I can… | Create software. | MD1 |
| Adapt software. | MD2 | ||
| Combine different software. | MD3 |
DOET, digital obesity exercise therapy; digital media, hardware and software.
This content validation provides the basis for subsequent statistical validation (CFA) to develop a standardized diagnostic tool and address the identified research gap.
3.2. Study 2: confirmatory factor analysis
The online study conducted via LimeSurvey yielded a total of 246 collected questionnaires. After excluding incomplete responses, 205 complete questionnaires were analyzed. The separate sample, independent of the cognitive interview participants, had a mean age of 32.45 years (SD = 5.9). The gender distribution included 52.2% men (n = 107), 47.3% women (n = 97), and 0.5% diverse (n = 1). The majority of participants were of German nationality (88.8%, n = 182), followed by Austrian (8.3%, n = 17), Dutch and Polish (1.0% each, n = 2), Italian and Swiss (0.5% each, n = 1).
The assumptions for confirmatory factor analysis were met, and despite non-normal data distribution, the MLR estimator provided robust parameter estimates. The initial five-dimensional model (M0) showed a moderate fit [χ²(125) = 351.650; p < .001; CFI = .917; TLI = .898; RMSEA = .086, 90% CI [.073;.100]; SRMR = .087]. Anomalies were observed with Item MD2, including a factor loading > 1 and a negative variance (.073), indicating a Heywood case. In addition to this statistical irregularity, the cognitive interviews also revealed issues related to the Media Development dimension, as therapists did not consider software creation or further development to fall within their therapeutic scope of practice. Therefore, Media Development dimension was excluded from further analyses.
An adjusted four-dimensional model (M1) showed improved fit [χ²(84) = 172.013; p < .001; CFI = .944; TLI = .929; RMSEA = .083, 90% CI [.065;.101]; SRMR = .074]. As the RMSEA exceeded the.08 threshold, correlations between dimensions were examined. A high covariance between Selection Criteria and Teaching Strategy (Cov = .952; p > .001) indicated conceptual overlap. Beyond statistical fit, their merging was theoretically supported by the close functional link between media selection and instructional planning, as emphasized in media didactics (45) and outlined in the introduction. The combined dimension (Media Application Strategy) reflects this integrated decision-making process.
The three-dimensional model (M2) showed good fit [χ²(87) = 181.672; p < .001; CFI = .939; TLI = .926; RMSEA = .085, 90% CI [.060;.102]; SRMR = .076]. However, a Likelihood Ratio Test revealed that the four-dimensional model [χ²(84) = 231.52] fit the data significantly better than the three-dimensional model [χ²(87) = 245.75; Δχ² = 9.23; p = .026].
Yet, to improve fit of the four-dimensional model, modification indices were calculated following Saris, Satorra (46), considering only values exceeding > 10, as these indicate a statistically significant improvement in model fit [p < .01; χ²(1) improvement] while minimizing the risk of overfitting (see Table 2) (47).
Table 2.
Modification indices for optimizing the four-dimensional model.
| Dimension | Item | MI |
|---|---|---|
| Selection Criteria (SC) | LS4 | 44.865 |
| TS1 | 16.662 | |
| LS2 | 11.923 | |
| MR4 | 10.511 | |
| Teaching Strategy (TS) | LS4 | 44.875 |
| MR4 | 19.167 | |
| MR3 | 18.597 | |
| LS2 | 10.834 | |
| Learning Support (LS) | MR4 | 25.168 |
| TS1 | 15.395 | |
| MR3 | 11.720 | |
| Media Reflection (MR) | LS4 | 39.964 |
| LS2 | 10.179 |
MI, modification indices.
Item LS4 (digital communication), which cross-loaded all factors, was identified as a key source of misfit. As communication is conceptualized as a prerequisite for learning rather than a distinct competence (48), the item was removed and the model was re-estimated. The revised four-factor model (M3) showed improved fit: CFI = .964; TLI = .953; RMSEA = .070, 90% CI [.049;.090] and SRMR = .051. Standardized factor loadings for all 14 items ranged from 0.59 to 0.93 (M = 0.78), confirming strong item-factor relationships. A full list of loadings is provided in Appendix A. The final model is shown in Figure 1. Values on arrows represent standardized factor loadings; residual variances are shown in parentheses.
Figure 1.
The final DigCompThExO model.
To evaluate the impact of model modifications, the M3-model was compared to the previous M1-model using a Likelihood Ratio Test with fit assessed via the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). The M1-model had an AIC of 7,695.1 and a BIC of 7,814.7, while the M3-model showed improved fit (AIC = 7,100.6; BIC = 7,213.6), indicating a better modeling of the construct. Table 3 summarizes all models and their fit indices in the order of modifications made during the research process.
Table 3.
Summary of calculated models and their fit indices.
| Model | Code | χ 2 | df | p | CFI | TLI | RMSEA [90%CI] | SRMR |
|---|---|---|---|---|---|---|---|---|
| 5-factor model | M0 | 351.650 | 125 | <.001 | .917 | .898 | .086 [.073;.100] | .087 |
| 4-factor model | M1 | 172.013 | 84 | <.001 | .944 | .929 | .083 [.065;.101] | .074 |
| 3-factor model | M2 | 181.672 | 87 | <.001 | .939 | .926 | .085 [.060;.102] | .076 |
| 4-factor model without LS4 | M3 | 124.289 | 71 | <.001 | .964 | .953 | 0.70 [.049;.090] | .051 |
χ2, chi-square statistic; CFI, comparative fit index; TLI, Tucker–Lewis index; RMSEA, root mean square of approximation; SRMR, standardized root mean square residual.
Convergent validity was supported by AVE values above.50 (SC = .62; TS = .55; LS = .78; MR = .59), indicating adequate variance explanation. Internal consistency was confirmed by CR values exceeding.70 (SC = .83; TS = .83; LS = .91; MR = .85).
To ensure minimal interdimension correlation, discriminant validity was assessed using the Heterotrait-Monotrait ratio (HTMT). All HTMT values were below the threshold of.85, confirming adequate discriminant validity between the dimensions of Selection Criteria (SC), Teaching Strategy (TS), Learning Support (LS), and Media Reflection (MR) (SC-TS = .79; SC-LS = .62; SC-MR = .77; TS-LS = .71; TS-MR = .75; LS-MR = .54).
The questionnaire revealed a clear four-factor structure with 14 items and strong construct validity. Measurement invariance across gender was tested using configural (I0), metric (I1), and scalar (I2) models. Due to the limited sample of diverse participants (n = 1), only male and female groups were analyzed. The configural model (I0) showed an acceptable fit [RMSEA = .074; 90% CI (.051;.096)], providing a baseline. The metric model (I1) indicated good fit [RMSEA = .069; 90% CI (.045;.091)], supporting equivalence in factor loadings. The scalar model (I2) also showed good fit, with the same RMSEA value, 90% CI [.046;.090], confirming scalar invariance across gender. Since model comparisons revealed no substantial loss in fit (ΔCFI ≤ .005; ΔRMSEA = .000), full scalar invariance was established, and partial invariance testing was not required. These results suggest equivalent construct measurement across genders (compare Table 4).
Table 4.
Measurement invariance across gender.
| Model | Df | AIC | BIC | χ 2 | Δχ 2 | Δdf | p(Δχ2) |
|---|---|---|---|---|---|---|---|
| I0 | 142 | 7,108.3 | 7,426.8 | 256.16 | |||
| I1 | 152 | 7,096.2 | 7,381.6 | 264.13 | 6.0214 | 10 | .8135 |
| I2 | 162 | 7,090.9 | 7,343.0 | 278.76 | 13.2281 | 10 | .1628 |
l0, configural model; l1, metric model; l2, scalar model; Df, degrees of freedom; AIC, akaike information criterion; BIC, bayesian information criterion; χ2, chi-square statistic; Δχ2, difference in chi-square statistics; Δdf, difference in degrees of freedom; p(Δχ2), p-value for chi-square difference.
4. Discussion
To tackle the challenges of digital transformation, this study validated the theory-based DigCompThExO questionnaire to assess digital competences of exercise therapists, guided by three research questions. First, cognitive interviews confirmed item clarity, relevance, and target-group appropriateness; resulting revisions (e.g., clarifying terminology, response simplification) improved content validity. Second, confirmatory factor analysis supported the hypothesized structure, with the revised model demonstrating good fit and thereby evidencing structural validity. Third, psychometric testing confirmed reliability and convergent validity. These findings establish the DigCompThExO as a diagnostic tool for competence development in DOET and contribute to advancing digital transformation within medical education and therapeutic practice. In practice, DigCompThExO can be used as a diagnostic entry tool in academic curricula and continuing professional development to identify individual competence profiles. Based on these profiles, modular training formats (e.g., workshops on digital feedback or self-monitoring tools) can be tailored to therapists’ specific developmental needs.
4.1. Interpretation of results
Due to limited data on exercise therapists in digital obesity therapy, professions with comparable implementation tasks (e.g., psychotherapy, physiotherapy) and digitally competent educators from non-medical contexts (comparable knowledge transfer) were included to underpin four key discussion points on the DigCompThExO's revision.
First, the DigCompThExO was revised based on cognitive interviews and statistical analyses, leading to the removal of the Media Development dimension due to its limited perceived relevance. Similar challenges have been reported in physiotherapy, a discipline grounded in exercise and practical interaction. In this field, limited media development is linked to the disconnect between pedagogical technology use and therapeutic practice (49, 50). Comparable issues arise in digital instruction within educational institutions, where the focus has shifted towards effective use and pedagogical adaptation of existing tools, rather than creating new media (51, 52). Despite this shift, DigCompEdu promotes digital content creation, assuming that educators (or therapists) should act as media designers to support individual learning (17, 53). While it remains unclear whether such expectations are realistic in therapeutic settings, growing demands for media development in exercise-related fields suggest relevance. This is particularly for true for the context examined here, where individualized digital tools are central to behavioral change (54). Consequently, given current challenges in adapting technology to individual needs, future studies should explore user acceptance and pilot selected media in RCTs to evaluate feasibility, efficacy and impact. Such evidence is crucial before recommending extensive media-related training, especially given the highly diverse ways digital media are used in therapy.
Second, reduction of the communication item was necessary due to poor model fit, highlighting persistent ambiguity around its role in digital education as well as therapy. While DigCompEdu defines communication as a core competence for structuring group interactions and learner engagement (17, 55, 56), therapeutic communication relies on nuanced, interpersonal exchanges within one-on-one or small-group interactions (48, 57, 58). Digital formats disrupts this by limiting nonverbal cues and tactile feedback (59, 60), forcing therapists to verbalize even implicit aspects of movement and interaction (61). Beyond interpersonal aspects, digital therapy also demands technical-exploratory communication, such as clarifying the purpose and function of digital tools (62). Yet, DOET therapists examined here struggled even to define “digital resources”, revealing striking conceptual uncertainty. Also, it is unresolved whether digital competence is a perquisite for digital communication or vice versa (63). Consequently, existing competence models may insufficiently reflect the specific communicative demands of digital therapy. Future digitalization efforts should define key conditions (e.g., group size, interaction level) and embed them in didactic design to compensate for the lack of physical presence. Only then can training address both empathic and technical dimensions of digital communication.
Third, the six-point Likert scale was reduced to competence application, as knowledge aspects were deemed overly complex and unclear during cognitive interviews. This marks a deliberate shift from assigning competences to broad cognitive stages towards capturing their nuanced expression within a single process via metrically structured scaling. This approach contrasts with the widely adopted DigCompEdu (64, 65). While enabling self-placement, it lacks psychometric rigor due to undefined level intervals and limited granularity; constraining statistical comparability and precise measurement of competence progression (17, 66). In contrast, the Likert scale used here enables fine-grained profiling of applied competences, supports metric approximation and is well suited to heterogeneous therapeutic settings (67), especially in addressing digital competences (68, 69). Ideally, each nuanced expression of competence across cognitive levels would be captured; however, such granularity places respondent burden (70). Consequently, this highlights the need for empirical studies to evaluate the cost-benefit trade-off, thereby informing evidence-based decisions regarding both the design and application of questionnaires. To ensure systemic alignment, these decisions should be developed in dialogue with key stakeholders (e.g., therapy providers, associations).
Fourth, contextual delimitation of competence frameworks is critical for valid interpretation and use. DigCompThExO's streamlined, four-dimensional structure (14 items) aligns with physical education instruments prioritizing parsimony and feasibility over item volume (71). This approach supports cognitive efficiency and reduced respondent burden without compromising content validity (72). In contrast, educational models such as DigCompEdu have expanded to include contextual dimensions (e.g., digital environment, extrinsic digital engagement), reflecting the resource-bound nature of education-based practice (73). Comparable relevance has been shown in healthcare, where digital competence is shaped by systemic factors, prompting dual-instrument models to capture both skills and enabling conditions (24, 74). Similar dependencies may also affect DOET therapists examined here, whose competence is shaped by institutional and national-level constraints (e.g., choice or funding for digital infrastructure) (75). Consequently, future extensions of DigCompThExO should incorporate context-related dimensions and stakeholder perspectives to clarify decision-making processes around digital infrastructure in relation to cost-benefit considerations. When measurable, contextual factors (e.g., digital infrastructure) could be examined through statistical analysis to explore their influence on therapists’ competence development, thereby informing more targeted and setting-specific training strategies.
4.2. Strengths and limitations
As a cross-sectional self-report questionnaire, the DigComThExO is inherently susceptible to biases such as recall errors, social desirability, and interviewer influence; potentially limiting the accuracy and objectivity of competence ratings. Standardized, experience-based interviews were used to mitigate these effects; however, future studies should incorporate objective measures (e.g., observation, third-party ratings).
Furthermore, DigCompThExO may be applied longitudinally to track competence development over time, for example before and after targeted training interventions or during professional socialization. In the present study, the relatively young, self-selected sample may have led to a systematic overrepresentation of digitally confident participants, potentially inflating overall competence scores and limiting generalizability to older or less digitally experienced therapists.
Potential item bias by gender or professional experience also cannot be ruled out. Future studies should examine differential item functioning to ensure measurement equivalence across subgroups. Such bias could affect group comparisons and obscure true competence differences.
Another limitation is the lack of test-retest reliability assessment, which prevents conclusions about the temporal stability of the scale and the reproducibility of individual competence scores. This was not feasible due to several constraints, including the cross-sectional study design, the sample size required for longitudinal follow-up, and the professional context of participants, which limited the feasibility of repeated measurement. However, measurement invariance across gender supports structural consistency. Future studies should implement longitudinal or repeated-measures designs (e.g., pre-post evaluations of trainings) to assess stability over time.
The removal of the knowledge-level items (“Know it”) during scale revision represents an additional limitation. This decision was based on high empirical overlap with the application items and participant feedback from cognitive interviews, where the distinction was perceived as redundant and burdensome in clinical routines. While this improved feasibility and focus, it may reduce conceptual differentiation between knowing and applying digital competences. Future studies could explore more efficient ways to capture knowledge aspects.
Finally, the questionnaire was developed and validated in a German-speaking sample. For cross-national and broader international applicability, translation and cultural adaptation following established guidelines (e.g., forward-back translation, cognitive pretesting) are required. Subsequent validation should confirm construct validity and measurement invariance across cultural or institutional contexts. This is essential to ensure generalizability and contextual relevance across health systems.
Despite limitations, this study offers key strengths. Methodological triangulation across qualitative and quantitative data ensured comprehensive construct validation. CFA further confirmed structural robustness in handling complex datasets. Finally, COSMIN-aligned methods ensured validation rigor.
5. Conclusion
DigCompThExO provides foundation for identifying existing competences and developmental needs, assuming future pilot applications enable the identification of current competence profiles among DOET therapists. Deriving from this, training programs should be developed and evaluated through RCTs, assessing competence acquisition and subsequent therapeutic outcomes. If effective, integration into academic curricula and continuing education for DOET professionals is imperative to ensure consistent qualification standards. Such integration may further support the establishment of accreditation criteria for DOET, enabling regulatory bodies and professional associations to align certification, licensing requirements, and reimbursement structures with validated competence standards. Only through trained professionals can DOET advance from pioneering interventions, such as the German KLAKSonline program (8), into a scalable, evidence-based standard of care. At the policy level, the instrument offers a foundation for defining minimal digital competence standards in national treatment guidelines and quality assurance frameworks. This could inform strategic workforce planning, support reimbursement negotiations for digital services, and guide the development of accreditation pathways for DOET providers. However, strategies for integrating the questionnaire into practice and embedding derived training into structured education remain to be determined.
Beyond its immediate use, DigCompThExO offers a methodological and conceptual blueprint for future research. Cross-cultural adaptation may enable valid assessment of exercise therapist competences beyond German-speaking countries. This presupposes that the previously discussed challenges, such as media effects, digital communication, and contextual influences, are adequately addressed. Comparative insights could inform international training strategies. Additionally, given the multidisciplinary nature of obesity therapy, the tool may be tailored to other obesity therapy professionals (nutritionists, psychologists) to support discipline-specific competence training. Beyond obesity-specific contexts, the questionnaire may be applied in broader rehabilitation settings and in the treatment of other chronic conditions (e.g., asthma, diabetes). It also holds potential for use in physiotherapy, sport science, and health promotion, where digital interventions and emerging technologies such as artificial intelligence are gaining importance.
In summary, this study closes a critical gap in the assessment of digital competences in obesity exercise therapy by introducing the DigCompThExO, the first validated instrument in DOET. Such questionnaires are essential to meet the evolving competence demands, where exercise therapists must translate obesity-specific content into digital formats across technical, pedagogical, and therapeutic domains. However, how the questionnaire should be implemented and which requirements are needed to derive effective training programs remains unclear, particularly in the light of the unsolved research questions outlined above. If these questions are systematically addressed, the study lays the groundwork to support the integration of digital competencies into medical education and promote more equitable digital access to obesity therapy.
Acknowledgments
The authors acknowledge the participating therapists for their voluntary participation and insights.
Appendix A: Standardized factor loadings (Std.all) for the final 14-item, 4-factor model.
| Factor | Item | Loading (Std.all) |
|---|---|---|
| Selection Criteria (SC) | SC1 | 0.862 |
| SC2 | 0.878 | |
| SC3 | 0.592 | |
| Teaching Strategy (TS) | TS1 | 0.765 |
| TS2 | 0.779 | |
| TS3 | 0.661 | |
| TS4 | 0.760 | |
| Learning Support (LS) | LS1 | 0.858 |
| LS2 | 0.931 | |
| LS3 | 0.862 | |
| Media Reflection (MR) | MR1 | 0.686 |
| MR2 | 0.772 | |
| MR3 | 0.794 | |
| MR4 | 0.820 |
Std.all, standardized factor loadings.
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by a doctoral scholarship awarded to SP by Friedrich Ebert Foundation (Friedrich-Ebert-Stiftung). The funding body had no role in the design of the study, data collection, analysis, interpretation, or manuscript writing. The publication was supported by the Open Access Publishing Fund of Leipzig University.
Footnotes
Edited by: Maura Pilotti, Prince Mohammad bin Fahd University, Saudi Arabia
Reviewed by: Siddiqua Aamir, Prince Mohammad bin Fahd University, Saudi Arabia
Maryam Bojulaia, Prince Mohammad bin Fahd University, Saudi Arabia
Diego Marcial, University SEK, Chile
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
Ethical approval was not required because, in accordance with the German Research Foundation's (DFG) ‘Guidelines for Safeguarding Good Research Practice’ and the recommendations of the German Council for Social and Economic Data (RatSWD), this study did not fall under procedures requiring formal ethics committee review. It involved a fully anonymous, non-invasive online survey of adult health professionals, with no interventions, foreseeable risks, or vulnerable populations. Participation was entirely voluntary, and electronic informed consent was obtained from each participant (Study 1 and Study 2) prior to data collection. By providing consent online, participants confirmed that they had been fully informed about the study's purpose, procedures, and data management. All procedures adhered to the Declaration of Helsinki and complied with the General Data Protection Regulation (GDPR). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
SP: Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Visualization, Writing – original draft, Writing – review & editing. TW: Supervision, Writing – review & editing. HW: Conceptualization, Supervision, Writing – original draft, Writing – review & editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
- 1.Bowman-Busato J, Schreurs L, Halford JCG, Yumuk V, O’Malley G, Woodward E, et al. Providing a common language for obesity: the European association for the study of obesity obesity taxonomy. Int J Obes. (2024) 49(2):182–91. 10.1038/s41366-024-01565-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lobstein T, Brinsden H, Neveux M. World Obesity Atlas (2022). London: World Obesity Federation; (2022). Available from: https://www.worldobesity.org/resources/resource-library/world-obesity-atlas-2022 [Google Scholar]
- 3.Klein S, Gastaldelli A, Yki-Järvinen H, Scherer PE. Why does obesity cause diabetes? Cell Metab. (2022) 34(1):11–20. 10.1016/j.cmet.2021.12.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ocloo J, Garfield S, Franklin BD, Dawson S. Exploring the theory, barriers and enablers for patient and public involvement across health, social care and patient safety: a systematic review of reviews. Health Res Policy Syst. (2021) 19(1):8. 10.1186/s12961-020-00644-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Erku D, Khatri R, Endalamaw A, Wolka E, Nigatu F, Zewdie A, et al. Digital health interventions to improve access to and quality of primary health care services: a scoping review. Int J Environ Res Public Health. (2023) 20(19):6854. 10.3390/ijerph20196854 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Habermann K. Eltern-Guide Social Media: Instagram, Snapchat, Ticktok und Co.—kinder und Jugendliche Unterwegs im Internet. Berlin, Heidelberg: Springer; (2021). p. 41–63. [Google Scholar]
- 7.Fürstenau D, Gersch M, Schreiter S. Digital therapeutics (DTx). Bus Inf Syst Eng. (2023) 65(3):349–60. 10.1007/s12599-023-00804-z [DOI] [Google Scholar]
- 8.Pawellek S, Wagner A, Wulff H. Digitalization of juvenile obesity therapy: intervention protocol of KLAKSonline. Germ J Sports Med. (2025) 76:7–14. 10.5960/dzsm.2024.619 [DOI] [Google Scholar]
- 9.Pojednic R, Arpino E, Halliday I, Bantham A. The benefits of physical activity for people with obesity, independent of weight loss: a systematic review. Int J Environ Res Public Health. (2022) 19(9):4981. 10.3390/ijerph19094981 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wabitsch M, Moß A. Evidenzbasierte (S3-) Leitlinie der Arbeitsgemeinschaft Adipositas im Kindes-und Jugendalter (AGA) der Deutschen Adipositas-Gesellschaft (DAG) und der Deutschen Gesellschaft für Kinder-und Jugendmedizin (DGKJ). AWMF-Leitlinienregister, Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften e. V.; Berlin, Germany; (2019). AWMF-Nr. 050-002. Available online at: https://www.awmf.org/leitlinien/detail/ll/050-002.html [Google Scholar]
- 11.Roitenberg N, Ben-Ami N. Qualitative exploration of physical therapists’ experiences providing telehealth physical therapy during COVID-19. Musculoskeletal Sci Pract. (2023) 66:102789. 10.1016/j.msksp.2023.102789 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Estel K, Scherer J, Dahl H, Wolber E, Forsat N, Back D. Potential of digitalization within physiotherapy: a comparative survey. BMC Health Serv Res. (2022) 22(1):496. 10.1186/s12913-022-07931-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Merolli M, Vallance P, McCreesh K, O'Sullivan C, Ahmed O, Kerry R, et al. Entrustable professional activities for physiotherapists to demonstrate digital health competency: an international delphi study. Physiother Theory Pract. (2025) 41(10):2075–88. 10.1080/09593985.2025.2495130 [DOI] [PubMed] [Google Scholar]
- 14.Moosbrugger H, Kelava A. Qualitätsanforderungen an Tests und Fragebogen („Gütekriterien“). In: Moosbrugger H, Kelava A, editors. Testtheorie und Fragebogenkonstruktion. 3rd edition. Berlin, Heidelberg: Springer; (2020). p. 13–38. [Google Scholar]
- 15.Batt A, Williams B, Rich J, Tavares W. A six-step model for developing competency frameworks in the healthcare professions. Front Med (Lausanne). (2021) 8:789827. 10.3389/fmed.2021.789828 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Redmond C, Farrell R, Cunningham C, Dineen A, Foley S, O’Donnell D, et al. Development of the EVIBEC learning outcomes framework to support the delivery of evidence-based practice curricula in health care professional programmes: a codesign approach. BMC Med Educ. (2024) 24(1):3. 10.1186/s12909-023-04972-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Redecker C. European Framework for the Digital Competence of Educators: DigCompEdu. In: Punie Y, editor. EUR 28775 EN. Luxembourg: Publications Office of the European Union; (2017). ISBN 978-92-79-73494-6, 10.2760/159770, JRC107466 [DOI] [Google Scholar]
- 18.Boateng G, Neilands T, Frongillo E, Melgar-Quiñonez H, Young S. Best practices for developing and validating scales for health, social, and behavioral research: a primer. Front Public Health. (2018) 6:149. 10.3389/fpubh.2018.00149 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Horváth L, Pintér T M, Misley H, Dringó-Horváth I. Validity evidence regarding the use of DigCompEdu as a self-reflection tool: the case of Hungarian teacher educators. Educ Inf Technol. (2025) 30(1):1–34. 10.1007/s10639-024-12914-6 [DOI] [Google Scholar]
- 20.Akatsuka E, Taguchi A, Miyagawa S, Sugiyama D. Development of a digital health competency scale for public health nurses. Int J Med Inf. (2025) 204:106076. 10.1016/j.ijmedinf.2025.106076 [DOI] [PubMed] [Google Scholar]
- 21.Golz C, Hahn S, Zwakhalen SMG. Psychometric validation of the digital competence questionnaire for nurses. SAGE Open Nursing. (2024) 10:1–8. 10.1177/23779608241272641 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Titov N, Hadjistavropoulos HD, Nielssen O, Mohr DC, Andersson G, Dear BF. From research to practice: ten lessons in delivering digital mental health services. J Clin Med. (2019) 8(8):1239. 10.3390/jcm8081239 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Pawellek S, Wulff H. Digitale patient: innenschulung: konzeption eines erhebungsinstruments für kompetenzen von therapeut: innen zur planung und umsetzung digitaler adipositastherapie. MedienPädagogik. (2024) 57:1–21. 10.21240/mpaed/57/2024.03.22.X [DOI] [Google Scholar]
- 24.Jarva E, Oikarinen A, Andersson J, Tomietto M, Kääriäinen M, Mikkonen K. Healthcare professionals’ digital health competence and its core factors; development and psychometric testing of two instruments. Int J Med Inf. (2023) 171:104995. 10.1016/j.ijmedinf.2023.104995 [DOI] [PubMed] [Google Scholar]
- 25.Prinsen CA, Mokkink LB, Bouter L, Alonso J, Patrick DL, de Vet H, et al. COSMIN guideline for systematic reviews of patient-reported outcome measures. Qual Life Res. (2018) 27(5):1147–57. 10.1007/s11136-018-1798-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Tourangeau R. Attitude measurement: a cognitive perspective. In: Hippler HJ, Schwarz N, Sudman S, editors. Social Information Processing and Survey Methodology. New York, NY: Springer; (1987). p. 149–62. [Google Scholar]
- 27.Boateng G, Collins S, Mbullo P, Wekesa P, Onono M, Neilands T, et al. A novel household water insecurity scale: procedures and psychometric analysis among postpartum women in western Kenya. PLoS One. (2018) 13(6):1–28. 10.1371/journal.pone.0198591 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Lenzner T, Schick L, Hadler P, Neuert C. Projekt” LoneCovid"-Fragen zum Sozialen Umfeld und Dessen Entwicklung im Verlauf der COVID-19-Pandemie: Kognitiver Pretest. Mannheim: GESIS: Leibniz-Institut für Sozialwissenschaften: GESIS-Projektbericht; (2023). p. 41. [Google Scholar]
- 29.Mosquera Gende I. Digital competence of future English language teachers: influence of age and gender on their self-perception. Forum Linguistic Stud. (2024) 6(3):357–73. 10.30564/fls.v6i3.6613 [DOI] [Google Scholar]
- 30.Bachmann R, Hertweck F, Kamb R, Lehner J, Niederstadt M, Rulff C. Digitale Kompetenzen in Deutschland – eine Bestandsaufnahme. RWI Materialien. Essen: RWI – Leibniz-Institut für Wirtschaftsforschung (2021). [Google Scholar]
- 31.Beatty P, Willis G. Research synthesis: the practice of cognitive interviewing. Public Opin Q. (2007) 71(2):287–311. 10.1093/poq/nfm006 [DOI] [Google Scholar]
- 32.Mayring P, Fenzl T. Qualitative Inhaltsanalyse. In: Baur N, Blasius J, editors. Handbuch Methoden der empirischen Sozialforschung. Wiesbaden: Springer VS; (2019). p. 633–48. [Google Scholar]
- 33.Nunnally JC. An overview of psychological measurement. In: Wolman, editors. Clinical Diagnosis of Mental Disorders. Boston: Springer; (1978). p. 97–146. 10.1007/978-1-4684-2490-4_4 [DOI] [Google Scholar]
- 34.Guadagnoli E, Velicer W. Relation of sample size to the stability of component patterns. Psychol Bull. (1988) 103(2):265. 10.1037/0033-2909.103.2.265 [DOI] [PubMed] [Google Scholar]
- 35.Kline R. Principles and Practice of Structural Equation Modeling. New York, NY: The Guilford Press; (2023). [Google Scholar]
- 36.Field A. Discovering Statistics Using IBM SPSS Statistics. 6th edition. London: SAGE Publications; (2024). [Google Scholar]
- 37.Brown T. Confirmatory Factor Analysis for Applied Research. 2nd edition. New York, NY: The Guilford Press; (2015). [Google Scholar]
- 38.Schreiber J. Issues and recommendations for exploratory factor analysis and principal component analysis. Res Soc Admin Pharmacy. (2021) 17(5):1004–11. 10.1016/j.sapharm.2020.07.027 [DOI] [PubMed] [Google Scholar]
- 39.Hu L, Bentler P. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Structural Equation Model. (1999) 6(1):1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
- 40.Browne MW, Cudeck R. Alternative ways of assessing model fit. Sociol Methods Res. (1992) 21(2):230–58. 10.1177/0049124192021002005 [DOI] [Google Scholar]
- 41.Rosli MS, Saleh NS, Alshammari SH, Ibrahim MM, Atan AS, Atan NA. Improving questionnaire reliability using construct reliability for researches in educational technology. Int J Interactice Mobile Technol. (2021) 15(4):109–16. 10.3991/ijim.v15i04.20199 [DOI] [Google Scholar]
- 42.Homburg C, Baumgartner H. Beurteilung von kausalmodellen: bestandsaufnahme und anwendungsempfehlungen. Marketing. (1995) 17(3):162–76. 10.15358/0344-1369-1995-3-162 [DOI] [Google Scholar]
- 43.Teo T, Srivastava S, Jiang L. Trust and electronic government success: an empirical study. J Manag Inf Syst. (2008) 25(3):99–132. 10.2753/MIS0742-1222250303 [DOI] [Google Scholar]
- 44.Vandenberg R, Lance C. A review and synthesis of the measurement invariance literature: suggestions, practices, and recommendations for organizational research. Organ Res Methods. (2000) 3(1):4–70. 10.1177/109442810031002 [DOI] [Google Scholar]
- 45.Kerres M. Mediendidaktik: Konzeption und Entwicklung Mediengestützter Lernangebote. Berlin: Walter de Gruyter; (2013). [Google Scholar]
- 46.Saris WE, Satorra A, van der Veld W. Testing structural equation models or detection of misspecifications? Structural Equation Modeling. (2009) 16(4):561–82. 10.1080/10705510903203433 [DOI] [Google Scholar]
- 47.Whittaker TA. Using the modification Index and standardized expected parameter change for model modification. J Exp Educ. (2012) 80(1):26–44. 10.1080/00220973.2010.531299 [DOI] [Google Scholar]
- 48.Schmidt J, Lipowsky S, Schmidt H. Kein Lernen ohne Interaktion: Gestaltungsprinzipien und Methoden für effektive Zusammenarbeit in E-Learning-Umgebungen. In: Pfannstiel MA, Steinhoff PFJ, editors. Kein Lernen ohne Interaktion: Gestaltungsprinzipien und Methoden für effektive Zusammenarbeit in E-Learning-Umgebungen. Wiesbaden: Springer Fachmedien Wiesbaden; (2022). p. 633–56. [Google Scholar]
- 49.Røe Y, Vik Torbjørnsen AC, Admiraal W. Educators’ digital competence in physiotherapy and health professions education: insights from qualitative interviews. Digit Health. (2024) 10:1–10. 10.1177/20552076241297044 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Ødegaard NB, Røe Y, Dahl-Michelsen T. “Learning is about being active, but the digital is not really active”: physiotherapy teachers’ attitudes toward and experiences with digital education. Physiother Theory Pract. (2024) 40(3):494–504. 10.1080/09593985.2022.2119907 [DOI] [PubMed] [Google Scholar]
- 51.Eickelmann B. Digitalisierung in der schulischen Bildung: Entwicklungen, Befunde und Perspektiven für die Schulentwicklung und Bildungsforschung. In: McElvany N, Schwabe F, Bos W, Holtappels HG, editors. Digitalisierung in der schulischen Bildung: Chancen und Herausforderungen. Münster: Waxmann; (2018). p. 11–25. [Google Scholar]
- 52.Domingo-Coscollola M, Bosco-Paniagua A, Carrasco-Segovia S, Sánchez-Valero J-A. Fomentando la competencia digital docente en la universidad: percepción de estudiantes y docentes. Rev Investig Educ. (2020) 38(1):167–82. 10.6018/rie.340551 [DOI] [Google Scholar]
- 53.Yang C, Ogata H. Personalized learning analytics intervention approach for enhancing student learning achievement and behavioral engagement in blended learning. Educ Inf Technol. (2023) 28(3):2509–28. 10.1007/s10639-022-11291-2 [DOI] [Google Scholar]
- 54.Napolitano M, Bailey C, Mavredes M, Neighbors C, Whiteley J, Long M, et al. Personalized versus generic digital weight loss interventions delivered on university campuses: a 6-month cost-benefit analysis. Transl Behav Med. (2023) 13(6):358–67. 10.1093/tbm/ibac081 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Touron J, Martín D, Asencio EN, Victoria Íñigo SP. Construct validation of a questionnaire to measure teachers’ digital competence (TDC). Rev Española Pedag. (2023) 76(269):17. 10.22550/REP76-1-2018-02 [DOI] [Google Scholar]
- 56.Oyarzun B, Martin F. A systematic review of research on online learner collaboration from 2012 to 21: collaboration technologies, design. Facilitation Outc Online Learn. (2023) 27(1):71–106. 10.24059/olj.v27i1.3407 [DOI] [Google Scholar]
- 57.Qobilovna AM. Communicative competence as a factor of teacher’s professional competency. Am J Soc Sci Hum Res. (2023) 3(09):32–44. [Google Scholar]
- 58.Egan G. The Skilled Helper: A Problem-management and Opportunity-Development Approach to Helping. Toronto: Nelson Education; (2013). [Google Scholar]
- 59.Watolla A-K. Zwischen wirklichkeitskonstruktion, kommunikation und kompetenz: was verändert sich durch digitalität? Medien Pädag. (2024) 58:59–72. 10.21240/mpaed/58/2024.06.27.X [DOI] [Google Scholar]
- 60.Tremain H, McEnery C, Fletcher K, Murray G. The therapeutic alliance in digital mental health interventions for serious mental illnesses: narrative review. JMIR Ment Health. (2020) 7(8):e17204. 10.2196/17204 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Fernando M, Lange E. New ways of working as a physiotherapist in the digital era: a qualitative study. Physiother Pract Res. (2025) 46(10):103–10. 10.1177/22130683251337337 [DOI] [Google Scholar]
- 62.Räisänen J, Tuovinen T. Digital innovations in rural micro-enterprises. J Rural Stud. (2020) 73:56–67. 10.1016/j.jrurstud.2019.09.010 [DOI] [Google Scholar]
- 63.Guillén-Gámez FD, Mayorga-Fernández MJ. Measuring rural Teachers’ digital competence to communicate with the educational community. J New Approaches Educ Res. (2022) 11(2):323–41. 10.7821/naer.2022.7.1053 [DOI] [Google Scholar]
- 64.García-Vandewalle García JM, García-Carmona M, Trujillo Torres JM, Moya Fernández P. Analysis of digital competence of educators (DigCompEdu) in teacher trainees: the context of Melilla, Spain. Technol Knowl Learn. (2023) 28(2):585–612. 10.1007/s10758-021-09546-x [DOI] [Google Scholar]
- 65.Rutten N, Brouwer-Truijen K. Defining XR-specific teacher competencies: extending the DigCompEdu framework for immersive education. Trends Higher Educ. (2025) 4(1):11. 10.3390/higheredu4010011 [DOI] [Google Scholar]
- 66.Siller HS, Bruder R, Linnemann T. Kompetenzstufen- und Kompetenzentwicklungsmodelle. In: Roth J, Ames J, editors. Beiträge zum Mathematikunterricht 2014. Münster: WTM; (2014). p. 83–4. [Google Scholar]
- 67.Li Q. A novel Likert scale based on fuzzy sets theory. Expert Syst Appl. (2013) 40(5):1609–18. 10.1016/j.eswa.2012.09.015 [DOI] [Google Scholar]
- 68.Yaqoob Mohammed Al Jabri F, Kvist T, Azimirad M, Turunen H. A systematic review of healthcare professionals’ core competency instruments. Nurs Health Sci. (2021) 23(1):87–102. 10.1111/nhs.12804 [DOI] [PubMed] [Google Scholar]
- 69.Longhini J, Rossettini G, Palese A. Digital health competencies among health care professionals: systematic review. J Med Internet Res. (2022) 24(8):e36414. 10.2196/36414 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 70.Aiyegbusi OL, Cruz Rivera S, Roydhouse J, Kamudoni P, Alder Y, Anderson N, et al. Recommendations to address respondent burden associated with patient-reported outcome assessment. Nat Med. (2024) 30(3):650–9. 10.1038/s41591-024-02827-9 [DOI] [PubMed] [Google Scholar]
- 71.Martínez-Rico G, Alberola-Albors M, Pérez-Campos C, González-García RJ. Physical education Teachers’ perceived digital competences: are they prepared for the challenges of the new digital age? Sustainability. (2022) 14(1):321. 10.3390/su14010321 [DOI] [Google Scholar]
- 72.DeVellis RF, Thorpe CT. Scale Development. Theory and Applications: Sage publications; (2021). [Google Scholar]
- 73.Alarcón R, Del Pilar Jiménez E, de Vicente-Yagüe MI. Development and validation of the DIGIGLO, a tool for assessing the digital competence of educators. Br J Educ Technol. (2020) 51(6):2407–21. 10.1111/bjet.12919 [DOI] [Google Scholar]
- 74.Jarva E, Mikkonen K, Andersson J, Tuomikoski A-M, Kääriäinen M, Meriläinen M, et al. Aspects associated with health care professionals’ digital health competence development—a qualitative study. Finnish J EHealth EWelfare. (2022) 14(1):79–91. 10.23996/fjhw.111771 [DOI] [Google Scholar]
- 75.Sailer M, Murböck J, Fischer F. Digital learning in schools: what does it take beyond digital technology? Teach Teacher Educ. (2021) 103:103346. 10.1016/j.tate.2021.103346 [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.