To address the holistic nature of our research interests adequately, a multistage research design was selected, as illustrated in Figure 1. The study followed the principles of interpretative social research and emphasized the subjective experiences and opinions of the stakeholders interviewed.

In the first research stage, an inductive analysis of the data facilitated the construction of new concepts and theories (35). The methodological approach was based on the principles of Grounded Theory (GT), incorporating iterative data collection and analysis, constant comparison and theoretical sampling (36–38). Specifically, the systematic analysis process involved the interpretation and correlation of experts' opinions, meanings and practical findings (37). The analysis prioritized the close examination of the source material through various coding steps rather than relying solely on the researchers' prior knowledge. In addition, the underlying concepts of practical knowledge and its interpretation were based on a constructivist approach (39, 40). The reporting structure follows the Consolidated Criteria for Reporting Qualitative Research (see Supplementary File 1) (41).

Given the existence of established theories in the field of talent development, in the second research stage data were analyzed deductively via qualitative content analysis (42). The participants' responses were categorized into predefined themes on the basis of two frameworks: the Athletic Talent Development Environment (ATDE) (7) and the Psychological Characteristics of Developing Excellence (PCDE) (1). In the final research stage, abductive reasoning was carried out by comparing the inductive and deductive results.

The interviewees for this research were stakeholders involved in the implementation of mental skills development programs, such as coaches, sports or performance psychologists, (sports) educators, teachers, or researchers. These stakeholders are characterized by special technical and processual knowledge and integrated experience in this specific field (43).

Figure 2 illustrates the exact recruitment process in more detail. Recruitment was conducted in two ways: (1) The authors' own network of contacts provided information about potential interviewees with whom the authors were not connected. (2) A respondent-driven sampling approach was chosen for the subsequent recruitment process; this means that at the end of each interview, interview participants were to suggest further interviewees. The selection and final inclusion of any interviewees were discussed and agreed by the research team. Our aim was to maximize diversity in terms of profession, experience, nationality, sports, sport setting (individual and team), institution (school, academy, association, club), and gender.

In the initial recruitment step, 7 participants were invited, and 4 agreed. At the end of the interview, these participants were asked to suggest additional stakeholders to be interviewed, along with their contact details. Using this respondent-driven sampling approach, further participants were invited. In the second recruitment step, 2 out of 5 invited participants agreed to be interviewed; in the third recruitment step, 3 out of 5 agreed; in the fourth recruitment step, 2 out of 2 agreed; and in the fifth recruitment step, 4 out of 5 agreed. This resulted in a total of 15 interviews. The participants were contacted via email, and provided with details in the form of a study information sheet. There were no prior personal relationships between the participant and the researchers who conducted the interviews. The present study protocol was assessed by the Cantonal Ethics Committee KEK Zurich (BASEC No. Req. 2023-001330) and was judged not to fall within the scope of the Human Research Act (HRA). This meant that authorization from the ethics committee was not needed and that written informed patient consent was waived.

Data were collected between February and September 2024 through individual semi-structured interviews in accordance with the recommendations of Gläser and Laudel (44). The interviews were conducted online by the researchers PM (8 interviews) and US (7 interviews), in English, German, or Dutch via Microsoft Teams. The research team designed an interview guide to ensure comparability of the interviews (45). To ensure the linguistic equivalence of the interview guide across different languages, a back-translation was performed.

The interview guide was divided into three sections: (1) In the introduction section, the stakeholders introduced themselves and described their area of responsibility and their relationship with mental skills training in youth sports. This concrete introduction was important for focusing on the main themes of this study and placing the interviewees at the center as experts. (2) In the main section, the participants explained their experiences with various topics in much more detail and independently. This allowed the interviewer to understand the different perspectives and gain deeper insight into the participants' perceptions. Issues raised could then be explored in more depth by asking specific questions (44, 46). (3) The conclusion section contained questions that gave respondents the opportunity to summarize what they felt were the most important elements (44, 46). This led to a reflective process in which they were able to reevaluate, reinforce or relate what they had mentioned in the main section.

Two pilot interviews with a mental coach and a sports psychologist from the personal network of the researcher US were conducted to test the process, technical details, and interview guide in advance (45). The interviewees' feedback and recorded interviews were discussed between US and PM regarding the process, completeness and conduct of the interviews, resulting in one question in the main section being moved to the conclusion section and one question in the introduction being made more general. Data from the pilot interviews were not included in the data analysis.

Based on the concepts of iterative data collection and theoretical sampling, the guideline was adapted after each round of data collection and analysis during the evaluation phase. For example, during the first round of further development, direct questions about gender differences and responsibility for mental skills training were added. The second iteration included specific questions about the necessary steps for successfully implementing mental skills programs.

After 15 interviews, no new categories emerged during data collection and no new codes were generated, indicating data saturation, and theoretical saturation was also reached (47).

The inductive data analysis process is illustrated in Figure 2. The interviews were recorded and transcribed verbatim. The transcripts were checked for accuracy against audio recordings but were not reviewed by the interviewee at this stage. The coding process aimed to identify both similarities in the data and higher-level concepts (48). The data were open coded (26) by coding software (MAXQDA V.24.2.0). In the first phase, the authors familiarized themselves with the content by reviewing the transcripts and highlighting areas of interest (36, 37). Two researchers (US and PM) independently coded the first two transcripts, and the resulting codes were compared and discussed. After alignment, the next two interviews were coded independently and discussed to establish a consistent approach to the categorization system, resulting in a hierarchical structure of codes and categories (26). After the next two coded interviews, a third researcher (JS) who was not familiar with the data, was consulted to discuss the comprehensibility of the preliminary results. The comparison of the coding between US and PM took place after an additional 3 interviews and then 2 and the last 3 interviews. The final results were again discussed with JS. In accordance with the constant comparison approach, new data were collected or analyzed to complete, specify and verify the categories identified. This ongoing comparison of the existing material aimed to avoid premature assumptions, unspecific, banal or redundant categories, uncover possible data gaps, answer the research question and sharpen the development of ideas in terms of theory formation (37, 38). When the data were reanalyzed, no inconsistencies or contradictions with the categories and concepts were found. Finally, to identify the main concepts, categories and subcategories and to create the final conceptual model of the analysis, the results were structured considering similarities, differences and connections. The process of constant comparison with the development of the concepts is shown in Figure 3.

In qualitative research, ensuring trustworthiness is critical to the scientific integrity and credibility of the research findings, and the results authentically reflect the experiences and perspectives of the participants and are methodologically sound (49). The trustworthiness of the results of this study was ensured through several measures. First, the data were collected through semi-structured interviews, which were analyzed via both inductive and deductive methods (50). As a constructivist grounded theory approach was adopted in this study, the research team was aware that the data collected were a product of both the researcher and the subject. To minimize bias and ensure sound interpretation, the data were analyzed independently by two researchers. The results were then cross-checked to ensure consistency and rigor. In addition, the results were critically discussed with a third independent researcher who was not familiar with the raw data. This person acted as a critical friend, asking specific questions and challenging interpretations to uncover blind spots or potential biases. Furthermore, the results (i.e., the inductively derived main concepts and categories with brief descriptions) were then sent to the participants for reflection, who were asked to provide feedback on four questions aimed at the accuracy, completeness and appropriateness of the interpretation (51). This gave them the opportunity to reflect on their own experiences and make additional comments or considerations, which were then incorporated into the overall results. Importantly, the final judgment of transferability rests with the readers, who should assess it on the basis of the contextual information provided (52). This highlights the necessity for a critical and reflective transfer of the results to other settings.

As shown in Figure 4, four main concepts emerged from the consolidation of the expert interviews, which represent the conceptual levels of their current practice in the implementation of mental skills training programs in youth sports. There are multiple relationships between the core concepts and categories, although clear boundaries are apparent.

Following the inductive process, a deductive content analysis was used to categorize the stakeholders' responses into predefined themes on the basis of two common frameworks: the Athletic Talent Development Environment (ATDE) (7) and the Psychological Characteristics of Developing Excellence (PCDE) (1).

A comparison of the inductively and deductively interview analysis outcomes as the foundation for abductive reasoning is summarized in Table 7.

The results of the inductive analysis highlight the importance of practical experience. They reflect the interrelationships between different factors and influencing variables and create a holistic picture of the complexity of long-term mental skill development. In contrast, the findings also highlight the value of deductive content analysis in systematically categorizing participants' experiences into existing frameworks. At the subcategory level, there was almost 88% alignment between the subcategories of the inductive analysis and the ATDE categories. Only 4 out of 33 subcategories could not be categorized. Since mental abilities also formed a separate category in the inductive analysis, they could be compared with the PCDEs at the level of coded segments. Among the 331 segments, 37 (11.2%) could not be classified in the PCDEs. Interestingly, despite the participants' diverse national backgrounds, which spanned North America, Oceania, and various parts of Europe, no clear cultural differences emerged from the analysis. Taken together, the inductively derived themes that are well aligned with ATDEs and PCDEs demonstrate the interconnectedness of structural, cultural and psychological factors in promoting sport and personal development.

Despite the strong alignment of most stakeholder statements and the ATDEs/PCDEs frameworks, however, some statements did not fit neatly into the existing subcategories, for example, process-oriented factors such as education and professionalism. Such unallocated or divergent codes indicate a need for further research and adaptation of existing athlete development frameworks. More specifically, current mental skill development frameworks may be particularly lacking in relation to the themes of “role alignment”, “communication” and “long-term athlete development strategies”, which emerged from our stakeholder interviews. Particularly for these unassigned themes, existing frameworks should be expanded to better capture nuances in talent development and psychological training.

An important aspect of applied research is the transfer into practice. According to our study and supporting literature, the following recommendations can be made for practical implementation. The process model (Figure 5) shows the relationship between the main concepts derived from the study and the steps for implementation.

Embedding mental skills training into the strategic and policy frameworks of sports organizations is crucial to achieve long-term impact and sustainability. This is achieved through top-down commitment, structural support, and clearly defined organizational values.

A resource audit should be conducted at the beginning of the planning phase. Anticipating and addressing common barriers related to time management, staff capacity, and budget constraints can improve the feasibility and sustainability of programs, especially in resource-limited settings. Making resources available effectively can help ensure effective access to services and support long-term, sustainable initiatives.

To improve the effectiveness of mental skills training, practitioners and organizations must establish clear roles and responsibilities within interdisciplinary teams and encourage structured yet flexible collaboration. Key strategies include fostering close communication and collaboration within the supporting team and encouraging parental involvement. Additionally, it is important to consider both the athlete's sports and the school environment.

Finally, to support athletes' mental development effectively, sport organizations should prioritize comprehensive education in sport psychology for all professionals involved. This includes systematically training coaches, parents, and support staff not only in the content but also in the effective delivery of mental skills interventions, using a variety of pedagogically sound methods tailored to organizational contexts and athlete needs. Building a knowledgeable and aligned support network is essential for fostering the quality of mental skill programs, and the long-term well-being and success of athletes.

This study includes a wide range of perspectives on mental skill development. Using a broad sample of stakeholders from different sports and professions captures many relevant areas of the topic. However, our sample is limited geographically to countries across three continents, which could lead to a Western cultural perspective. Moreover, information on the interviewees' countries of birth (not necessarily their countries of employment) was lacking. This may have further influenced the results due to socio-cultural differences. Finally, using the respondent-driven sampling approach may also have resulted in some selection bias.

This study adopted a constructivist approach, acknowledging that the social and professional backgrounds of the researchers may influence both data collection and interpretation. The research team comprised an international group from Switzerland and the United Kingdom, bringing diverse professional perspectives to the project. The team included one woman and four men. PM is a sport psychologist and PhD candidate; US, a sports physiotherapist, contributed to the project as part of her Master's thesis. DC, RF, and JS are senior researchers with experience in (sport) psychology and sport science.