EduXBot is a longitudinal feasibility study based on a mixed methods approach. The mixed methods approach combines aspects of quantitative and qualitative research in order to examine the research topic from different perspectives and to answer questions that could not be answered by a purely quantitative or purely qualitative study [46]. The researchers initiated the study with a quantitative phase, which was followed by a second qualitative phase that provided a more in-depth explanation of the initial results [47].

A total of 3 care institutions (1 inpatient, 1 outpatient, and 1 life-sustaining) that support people in their health care and daily life were recruited to participate in the project. The institutions were contacted as partner institutions in previous DAT research projects and agreed to participate in this project. Therefore, it can be assumed that the participating nurses are generally aware of the topic of DAT in nursing. The facility managers will be informed in detail about the project and will be asked to act as gatekeepers to motivate the nurses and relatives in their facility to participate in the research project.

Participants are invited to take part in the study as test subjects by the facilities with which they are affiliated. The participants are as varied as possible in terms of age, gender, length of professional experience, and so forth [48]. The inclusion of nurses from diverse backgrounds within the facilities serves to enhance the heterogeneity of the study sample, thereby contributing to the study’s overall value. It should be noted that some participants may be unable to commit to the study due to scheduling conflicts or other commitments, while others may choose to withdraw (Textbox 1). It is essential to emphasize that no financial compensation is provided for participation in the study. However, those who elect to take part receive certain benefits, including complimentary refreshments during the workshop sessions and, upon request, recognition of their institution and name in the project acknowledgments.

In the context of research involving human participants, it is imperative to consider pragmatic factors such as financial resources and the maximum feasible number of participants in accordance with the guidelines for designing and evaluating feasibility pilot studies [49]. According to the guideline, achieving saturation in a study occurs at 30 participants. Statistical principles suggest that 30 individuals are sufficient to ensure the normality of the sample size [50,51]. Therefore, the objective is to recruit a minimum of 30 caregivers. It is important to acknowledge that the implementation of blinding is not a viable option in this context, as the participants are required to actively engage with the technologies under study.

The primary objective of this study is to design the human-technology interaction between the platform and users in a manner that ensures a high level of usability for all users. The objective is to assess whether the accessible platform alters the intended use of DAT by offering a simplified application option for caregivers. This inquiry seeks to ascertain the potential for DAT to be integrated as a supplementary resource in the planning of individual care processes. The central question guiding this study is whether the EduXBot platform offers an interface that enhances the autonomy of caregivers’ DAT use. The objective is to enhance usability by a minimum of one level over the course of the study. The primary objective is to ascertain the extent to which the platform affects the usability of DAT (System Usability Scale [SUS]), the intention to use it (Technology Usage Inventory [TUI]), and the satisfaction of caregivers’ needs through technology (Technology-Based Experience of Need Satisfaction [TENS]).

The secondary objective is to develop a multilevel benefit concept for EduXBot. This will provide caregivers with different affinities for electronic media, as well as different levels of technical expertise, and will offer them windows of opportunity to use and experience the platform. The objective is to conduct a direct assessment of the EduXBot platform by the participating nurses and to evaluate the platform with the aim of continuously improving the content and application options of the platform. A secondary objective is to record user satisfaction in order to obtain information on potential improvements to the platform through the analysis of structured feedback. The collection of qualitative data will be facilitated through the implementation of think-aloud (TA) and focus group methods. The aim is to empower caregivers to articulate their subjective experiences and perceptions regarding the utilization of EduXBot in an unstructured manner.

The third objective is the long-term evaluation of the EduXBot platform, encompassing not only its readiness for implementation but also its potential for long-term integration, utilization, and management of DAT. A further objective is to examine the interrelationship between the perceived ITU, usability, and user acceptance of the platform, which is currently being developed for the application. The tertiary target variable is the long-term willingness to use DAT in a sustainable manner and to integrate the digital assistive support offering into the existing care provision framework.

The measurement times are based on the development process shown in Figure 2. This allows changes in improving ITU to be measured. Since there are different versions with different application scenarios, especially at TUI, Table 1 is intended to provide an overview of the measurement times and evaluation methods used.

The TUI is a valid measuring instrument that is based on the established Technology Acceptance Model [52] and its further developments [53]. It comprises 9 subscales, encompassing a total of 33 questions. These questions facilitate the estimation of technology use based on technology-specific and psychological factors. Each scale comprises 3-4 items, which are answered using a 7-point Likert scale. The intention to use scale uses a visual analog scale with a length of 10 cm. The subsequent Table 2 offers a comprehensive overview of the TUI scales.

The internal consistencies of the scales can be rated as good overall (Cronbach α=0.70 to α=0.89). Furthermore, the TUI scales (with the exception of accessibility) have been found to be valid indicators of stress and relaxation based on heart rate and heart rate variability [53]. The wording of the individual questions can be adapted to the specific technology being evaluated, with the exception of the technology anxiety and interest scales. The TUI’s modular design allows for the exclusion of individual scales contingent upon the investigative objective. For this study, for instance, the immersion scale was excluded, as no technology aimed at immersion was used.

This study uses the questionnaire at each measurement time (Table 1). Before the development of the platform, the original TUI questionnaire pre-post version was used to assess the caregivers’ needs, functionalities, and DAT application scenarios. The “pre” version is administered before the introduction of the DAT, with the objective of collecting data on technology usage tendencies. The “post” version is administered after the introduction of the DAT, with the objective of establishing a baseline for measurement. Subsequent to these measurements, the TUI II parallel questionnaire (complete version) will be administered in its entirety. The measurement enables the formulation of statements regarding changes in the ITU.

For the purpose of evaluation, a cumulative value was formulated for each scale. The cumulative value commences at 3 or 4, representing the lowest level of the construct, and, contingent on the number of items, ranges from 21 (3 items) to 28 (4 items) at the highest level. The ITU scale constitutes an exception in this regard. The scale is evaluated by measuring the distance from the right endpoint (full rejection) to the answer at the intersection of the line. The distance in millimeters was ascertained and added to all 3 items. The maximum scale value that can be attained is 300. The interpretation of test values is facilitated by the scale description. A high test value is indicative of a high level of the respective construct. In instances where the data does not conform to a normal distribution, it is advised to use stanines (standardized calibration scale with a minimum of 1 and a maximum of 9) and percentile ranks (relativizing the test characteristic value in relation to the reference population) [53]. Standard tables have been developed for this purpose (Table 3). Consequently, a statement regarding the proportion of participants who attained the same or lower values is attainable.

The SUS assesses the usability of a system as perceived subjectively by the user and is proven to be technology-independent; that is, it can be used for a wide range of systems and technologies [54,55]. The 10 items are divided into 5 positive and 5 negative statements and are each represented on a Likert scale from 0 to 5. The participants’ answers result in the SUS item score, which must then be converted into the SUS overall score (from 0 to 100) [56].

The calculation of the overall SUS score entails the subtraction of 1 from the raw value of all odd items in the initial step, whereas the raw value of 5 is to be subtracted from the raw value of all even items. To illustrate this calculation, consider the following example. If item 1 had a raw score of 4, the calculated score would be 3 (obtained as 4 minus 1). For item 2, if the raw score was 2, the score was 3 (derived as 5 minus 2). Subsequently, the sum of these scores was calculated and multiplied by 2.5 to derive the overall SUS score [56].

The average overall score of all studies (68) can be used to interpret your overall score [55,56]. Bangor et al [57,58] also introduced a rating scale using adjectives and letters analogous to the American school grading system as an aid to interpretation (Table 4).

Deci and Ryan’s [59] assertion posits that the use of the platform is contingent upon the satisfaction of specific fundamental human needs, including autonomy, competence, and connectedness [35]. Psychology posits that the more fundamental psychological needs are met through interaction with the system, the more end users engage with technology [60]. The objective of the TENS-Interface questionnaire is to ascertain the extent to which direct interaction with a technology fulfills the fundamental psychological needs for autonomy, competence, and connectedness [35]. The questionnaires, initially developed in English, underwent translation into German under the guidance of a professional linguist for their usage within the project. The translation process was guided by the “Translation Guidelines and Translation Documentation of the European Social Survey (ESS, 2020)” [61] with a focus on maintaining the validity and comprehensibility of the translation.

In the TENS-Interface, the items are each assigned to the basic needs of competence, autonomy, and relatedness. However, they are presented randomly in the questionnaire. The objective is to ascertain the extent to which direct interaction with a technology fulfills the fundamental psychological needs for autonomy, competence, and relatedness [35]. The application is designed to be universally accessible, ensuring autonomy of action, competence in handling, and a connection with the technology facilitated through the EduXBot platform interface. The collected data are then subjected to evaluation according to these objectives.

The descriptive statistics of the quantitative data are presented depending on the distribution, such as mean or median. Categorical data describing the sample are presented as absolute and relative frequencies. In order to answer the main question about the change in ITU, the difference between the measurement times of the individual test subjects, the facilities, and the overall sample is described as an absolute and relative mean difference. Possible further group differences are examined using the parameters “qualification” and “work experience.” IBM SPSS is used as a tool for data management and data analysis.

The TA methodology is used to qualitatively evaluate the platform. The idea is to ask subjects to express their thoughts and emotions out loud while testing the prototype in order to document them [62,63]. The advantage of this method is that it collects problems with the technology that has been tried by the end users, as retrospective surveys can lead to incomplete information about the problems with a technology. This means that TA protocols are helpful in understanding the thinking strategies of end users [64].

Testing sessions can be conducted on participants’ own equipment or in a controlled environment. In protocols, participants think out loud as they complete a series of predetermined tasks. Participants are asked to say anything that comes to mind as they complete the task. This may include what they see, think, do, and feel. Sessions are often audio and video recorded so that developers can review what participants did and how they responded. Raw data comprise the verbalization of the thoughts, perceptions, and feelings that participants articulate as they complete a defined task. In a formal research protocol, all verbalizations are transcribed and then analyzed. In the context of usability testing, observers are asked to take notes on what participants say and do without trying to interpret their actions and words, and in particular, to note where they encounter difficulties [65].

The focus groups conducted during the iteration loops are partially structured with the help of guidelines Multimedia Appendix 1. The overall goal is to generate specific knowledge to answer the research question during the development process. The main factors that play a role are the adaptation of the selected application scenarios to the respective facilities as well as aspects of control and usability.

The composition of the focus groups is specifically based on theoretically based preselection in the sense of theoretical sampling according to the indicator caregiver [66]. In addition to these criteria, the most important thing is the willingness to talk about the respective needs and needs in the context of the platform development. No further criteria are defined in advance, but after the sample pool has been generated, the composition is determined exactly with regard to the contrasting of the groups. The number of participants in the focus groups is limited to 6 to a maximum of 10 people [67,68]. The selection of queries takes place according to the principle of theoretical sampling; achieving statistical representativeness is not intended [69].

All qualitative data were subjected to Kuckartz and Rädiker’s [70] qualitative content analysis, which provides a structured framework for the content (Figure 4). Given the project’s position within an emerging field of research, the material was coded inductively. This entailed the creation of categories based on the TA protocols and focus groups, thereby facilitating an exploratory evaluation of the material. This methodological approach enabled the systematic organization of the data material according to its content-related characteristics.