Giving guidance in the online lives of young people with intellectual disabilities: Challenges and guiding approaches in the daily practice of Dutch care staff

ABSTRACT Background This study reports the findings from a Dutch study, exploring care workers’ challenges and approaches in guiding young people with intellectual disabilities in their online lives. Method Using an inductive research design 33 semi-structured interviews were conducted with care workers. Results Care workers report challenges that revolve around three themes: (1) the perceived intangibility of the online lives of clients and (2) Risk perception: feeling stuck between control and freedom. (3) Differences in guiding approach between care workers. Based on these results a conceptual model is presented that aims to show the connection between these challenges and guiding approaches. Conclusion Perceptional processes may influence care workers’ guiding approaches. This raises important questions about the implications of these guiding approaches on the level of support clients with intellectual disabilities receive and the effectuation of their rights and liberties under the 2006 United Nations (UN) Convention on the Rights of Persons with Disabilities.

Intellectual disability; mild intellectual disability; care workers; online lives; guiding approaches In our digitalising society, there is a growing body of knowledge on support and inclusion of people with intellectual disabilities in the digital world (Chadwick, 2019). From a disability and human rights perspective this is significant: the 2006 United Nations (UN) Convention on the Rights of Persons with Disabilities (CRPD) clearly states the importance of promoting the use of Internet technologies for persons with disabilities, so they can live free and independent online lives. The term "online lives" (OL) in this study refers to the set of actions people perform and the set of relationships they have when connected to the Internet (Benvenuti et al., 2019). Thus, the online activities and social interactions of young people with intellectual disabilities are part of their OL.
In line with Chadwick (2019) and Borgström et al. (2019), it can be argued that professional care givers have a responsibility to promote digital inclusion for people with intellectual disabilities without controlling or monitoring their OL, although this may apply more strongly for adults than children and young people. Löfgren-Mårtenson et al. (2018) suggest a participatory approach allowing people with intellectual disabilities and professionals to reflect upon themselves and their relationship in a continuous dialogue. This is in accordance with art 12.3 of the CRPD that all persons with disabilities have the right to the support they may require to make their own decisions. Several authors (e.g., Seale, 2014) advocate that the positive risk approach might help to create a support environment that adheres more to the aspirations of the CRPD. This approach sees risk as an opportunity to start a dialogue on how caregivers and people with intellectual disabilities perceive certain online acitivities and what support might help both to better manage risk instead of avoiding it (Sorbring et al., 2017). Although the theory of positive risk taking has gained more attention in the last decade, Chiner et al. (2017) point out that there is still little evidence of its applicability to online risk-management of people with intellectual disabilities.
Despite these suggestions, many care organisations struggle to determine what online support they should provide. This is definitely the case for care organisations in the Netherlands that work with higher-risk young people with mild intellectual disabilities and borderline intellectual functioning. 1 Although evidence is scarce and sample sizes are small, some studies conclude that Dutch professional healthcare workers often give online guidance without proper guidelines and education (De Groot et al., 2022;Nikken et al., 2018). From the international literature a similar picture can be drawn (e.g., Chiner et al., 2017;Ramsten et al., 2019). Furthermore, support strategies often seem to be based on individual perceptions (Löfgren-Mårtenson et al., 2018;Ramsten et al., 2019). Moreover, negative perceptions overshadow the positive ones. In practice, people with intellectual disabilities are seen as more vulnerable online and thus in greater need of protection and safeguarding (e.g., De Groot et al., 2022;Ramsten et al., 2019). These negative perceptions may stem from preconceptions about people with intellectual disabilities in regard to their coping skills or inability to indentify or recognise risk online (Chadwick et al., 2016;Löfgren-Mårtenson, 2008), but negative media representation of the Internet might also be of importance (De Groot et al., 2022).
There may be other challenges that care workers face in their work. There is some evidence that a lack of education, formal training and clear organisational guidelines may create inaccurate risk perceptions that could influence their guiding approaches (Chiner et al., 2017;Seale, 2014). Furthermore, both De Groot et al. (2022), and Ramsten et al. (2019) illustrate that care workers strongly believe that the Internet is private, and that interference can only take place on request of the people they support. Although in general, these beliefs may underscore the importance of supporting people with intellectual disabilities to be free and independent online, it may also create a conflict of interest when care workers suspect that clients are at risk online or when they deem it pedagogically important to interfere in the OL of young people with intellectual disabilities. Seale (2007) describes a similar challenge when support workers try to facilitate online independency of adults with learning disabilities who have their own homepage online, but also want to ensure that clients create safe and responsible content. This places clients with intellectual disabilities in a passive position where they are denied the opportunity to make their own choices online. Along a similar line, Cobigo et al. (2016) point out that professionals experience a paradox between believing that clients should be free to make their own choices and at the same time wanting to protect them from harm.
These barriers and challenges may help explain what guiding approaches care workers adopt. Although some studies show a more heterogeneous view of online guiding approaches and a strong support for a more dialogic approach (Chiner et al., 2017;Löfgren-Mårtenson et al., 2018), many care givers still adopt controlling, protective and risk-averse guiding strategies, and act as online gatekeepers (Chadwick, 2019). For example, Sorbring et al. (2017) found that parents of youths with intellectual disabilities reported giving their children both practical and emotional support on social codes, netiquette and technolological problems. Several parents also befriend their children on social media as a way to discuss distressing things they see online. This is in line with Seale (2001), who found that parents were active on the social media platforms that their children used so they could monitor and influence the content their children watched and created, or ban content with a violent or pornographic nature. Bannon et al. (2015) also found that parents restricted young people with intellectual disabilities by monitoring their online activities and blocking certain websites. Seale and Chadwick (2017) argue that it is difficult to assess the extent to which digital inclusion and online independence is possible for people with intellectual disabilities as long as their online world is overly mediated by parents and professional caregivers.
Although it is possible to draw a broad picture from the literate on what challenges care workers see in guiding young people with intellectual disabilities online and what guiding approaches they adopt, three knowledge gaps are addressed in this study. First, there is little research structuring which approaches professionals adopt in their daily work (e.g., Alfredsson Ågren et al., 2020;Chadwick, 2019). Secondly, no studies are known to the authors that explicitly try to describe how online guiding approaches are connected to the challenges care workers face in their work. Lastly, many people with intellectual disabilities that live in a Dutch care setting have borderline intellectual functioning and receive support to people with mild intellectual disability. As this is different from the practice in most other countries (Woittiez et al., 2018), it is unlikely that people with borderline intellectual functioning are often included in scientific research that is conducted in other countries than the Netherlands. In this light, it is important to note that all further references in this paper to young people with intellectual disabilities include both people with mild intellectual disabilities and borderline intellectual functioning. Given these knowledge gaps, the aim of this paper is to make a contribution to the evidence base by presenting the results of a qualitative study that describes what challenges care workers see in guiding young clients online, what guiding approaches care workers adopt in dealing with these challenges, and if those challenges and approaches are connected. Answering these questions is important as it gives care workers a better understanding of their online guiding strategies. This may help to improve the quality of support care workers provide their clients online.

Study design
This study is based on the same dataset as presented in De Groot et al. (2022). A qualitative, inductive research design was constructed that used semi-structured interviews with 33 care workers of Dutch care organisation Middin.

Ethical considerations
This study was approved by the research committee of Middin. Additionally, the researchers followed the general guidelines as provided by the Netherlands Code of Conduct for Research Integrity (Association of Universities in the Netherlands, 2018). All respondents gave their full consent for the recording and use of their interviews for the purpose of this study. Names of clients, dates, and places were altered or deleted to ensure the privacy of all respondents and their clients.

Recruitment of participants
Care workers were recruited from four group homes of Middin that operate in the South-West region of the Netherlands. These group homes all provide 24-hour in home care for people with intellectual disabilities, but differ in their purpose. All locations serve as living facilities, but one also provides short-term crisis care, while another specifically aims to prepare young people for a more independent life with ambulant care. Locations also differ in their clientele. While three locations work with young people with intellectual disabilities, one location specialises in clients with comorbidity (intellectual disabilities in combination with other mental health problems). Also, only at one location clients were under the age of 18 years, while the other three provided care for young people that are aged from 17 years and up with predominantly young adults.
Location managers of these four facilities were asked to identify 10 care workers from their team. Inclusion criteria were that participants had to have at least oneyear work experience with young people with intellectual disabilities in the age range of 12-27 years. This wide range was chosen after asking respresentatives of the care organisation to consider which group homes should be included in the sample. Although on paper Middin has separate group homes for young clients between 12 and 18, and 18+ years, in practice clients over 18 years can be allocated to a group home for young people based on their developmental age. Addittionally, the 18+ years group homes had clients up to 27 years with a developmental age of a teenage person, so their care workers were also included in the research sample. Further, participants needed to be willing to talk about their experiences in guiding clients in their online world.

Participants
The location managers provided the names of 40 care workers, of which seven later declined because of their workload. Six of these seven care workers worked at the location that provided care for young people under the age of 18 years. Of the remaining 33 participants, 8 were male and 25 were female. Twelve care workers who participated were distributed in the age range between 20-30 years, 10 between 30-40 years, six between 40-50 years, and five between 50-65 years. Most care workers predominantly work with young people with intellectual disabilities in one of the four group homes. However, one participant worked as a behavioural therapist in one of these homes and three participants also provided ambulant care for young adults who live independently. Participants received at least post-secondary vocational education and training in social work at a full professional level.

Data collection
The interview data was collected as described in De Groot et al. (2022). In brief, interviews between 60 and 90 minutes were conducted by graduating applied psychology students of Leiden University of Applied Sciences in the Netherlands at the group homes were care staff worked. Methodological rigour was established by training the students in interview techniques and how to be reflexive researchers (e.g., raising awareness about their own experiences, beliefs and perceptions), and by providing them with a practical interview protocol containing sections on research ethics, interview topics plus additional open-ended questions, a list of open-ended prompts, and a practical reminders to ensure a smooth interview process. The students also attended weekly feedback sessions with the lead researcher to discuss questions or troubles that arose prior, during and after their data collection. The interviews were recorded on a voice-recorder and transcribed verbatim by the student researchers.

Data analysis
The analysis of the data was conducted through an inductive thematic analysis (Braun & Clarke, 2006). By choosing this type of analysis it is to be expected that some depth and complexity is lost, but a rich overall description is maintained. Therefore, it can be considered a useful method for the investigation of under-researched topics (Braun & Clarke, 2006).
Quirkos, a qualitative analysis software tool, was used to facilitate the analysing process. The interview transcripts were analysed using open coding, after which codes were aggregated into themes. Subsequently, these themes were checked against all codes and the entire data set while generating a thematic map of the analysis. The authors refined the thematic map by discussing the themes, how they related to each other, and what overarching story they were telling.

Results
The results are divided in two parts. First, three themes were derived from the data: (1) The perceived intangibility of the OL of people with intellectual disabilities, (2) Risk perception: feeling stuck between exercising control and supporting online freedom, and (3) Differences in guiding approaches. Second, a conceptual model is presented that aims to show how care workers' perceptions on tangibility and risk are connected to their guiding approaches.

Theme 1: Perceived intangibility of the OL of people with intellectual disabilities
In general, care workers find it difficult to get a grip on the OL of young clients, because they think they lack the experience and know-how to e.g., help clients improve their communication and social skills, increase resilience, or how to resolve conflicts. Because clients sometimes refuse help and conceal their online activitiesdeleting chat histories or turning off location sharing guiding them becomes even more difficult and amplifies their insecurities to give good support online. Care worker Tom shares an example of how care workers feel unable to get a grip on the OL of clients: If a client is calling people names in this house or harasses someone, we talk to them about it and try to explain why their behaviour is not acceptable. This is something we can control to a certain degree. But in the online world their behaviour is less visible to us, so things get out of hand quickly. If one of them sends a negative text and his friends reinforce his negative behaviour by liking his texts and adding negative comments as well, the situation can explode quickly.
The perception that the OL of clients are intangible also seems to be reinforced because care workers often believe that getting older hinders them to understand what clients do online. Although care workers see that they have things in common with clients online, like having a Facebook or WhatsApp account and watching Netflix, they also experience many differences (e.g., clients playing videogames, using different social media platforms, watching influencers on YouTube or Twitch) that they sometimes attribute to a knowledge-age gap. For some care workers this causes a disconnect between them and their clients and as a result, they either loose interest in what clients do online, or it increases insecurities about their abilities to give good guidance online. Asking 36-year-old care worker Emma about her ability to get a grip on online activities of clients one thing she said mirrored what other care workers reported about their fear of gradually losing grip over clients online: These young people are so well versed in social media and all its possibilities. Looking at older colleagues, what is probably what I am, you can see that they can't keep up with them. This is when we lose a little grip over them. Same goes for street slang they use, sometimes you think: what are they saying to each other? Both this and social media develop so fast that is creates a gap with older colleagues.
Even care workers in their twenties, who say that they feel more in touch with what young clients do online, give examples like the one described by Emma. They also think that it is difficult to keep up with all the latest online trends and therefore stick to what they already know.
Some care workers, however, do not seem to regard the OL of their clients as part of real life. Consequently, care workers sometimes assess online issues differently than they would offline. Care worker Emilia for example explains that she is inclined to refrain from intervening when a client is sexting online, while she would not hesitate to act if for example a client is picked up by a stranger wearing a sexy dress. For some care workers, the online world is so far removed from their own life that investing in it feels pointless, especially when they believe that there are more pressing matters in the offline world.
Theme 2: Risk perception: feeling stuck between exercising control and supporting online freedom Although the OL of clients often feel intangible to care workers, they do get glimpses of clients participating in many of the same online activities as their non-disabled peers, like being active on social media, playing online games or using dating apps like Tinder. Care workers often question the extent to which clients can behave in a safe and responsible way, which makes it difficult for them to assess how much risk the online world poses to clients. Besides their own insecurities about their know-how of the OL of clients, care workers also often think that clients have limited learning skills or are not ready to use the Internet safely and independently. When talking to care worker Oliver about a male client who was scammed by someone who presented herself as a potential love interest on a site for webcam girls, he gives an example of a frequently mentioned challenge by care workers in this context: There are websites with content that should be appropriate considering their legal age. Especially when they turn 18, they are allowed to access anything they want online. But mentally they are sometimes not older than seven or eight years old. Shouldn't there be more legal protection for them?
This quote is exemplary for care workers' fear that the disbalance between clients' legal age and their developmental age has negative consequences for them when they go online. Clients who turn 18 in the Netherland are legally adults and therefore may access the Internet like their non-disabled adult peers. Although they often have acquired the technical skills to use the Internet to their advantage, care workers feel that because of their developmental age they can be more vulnerable to be exploited online.
Although most care workers believe that clients have the right to be private and free online, at the same time they have strong feelings of responsibility for the wellbeing of clients. They express a need to give clients more information and advice on Internet risks, help them with job applications, and show them how to contact organisations online. While doing this, it is difficult for them to find a balance between exercising some form of control on the OL of clients and supporting their digital freedoms. Care workers can feel uncomfortable in pushing clients to be open about their online activities, but without having some insight in clients' activities they are afraid that things can spiral out of control quickly. These fears are exacerbated by clients who do not want care workers meddling in their online affairs, or parents and legal guardians who sometimes oppose and even undermine guiding approaches of care workers. Because of this, the result of their efforts can be dissatisfying and frustrating.
When talking about the struggle between control and freedom, care workers point to the absence of legal or organisational guidelines, practical tools, and formal education on how to deal with online risks. They consider the Internet as a moral grey area for them as well as their clients. They find it difficult to know what digital content is risky for clients with intellectual disabilities and to decide what kind of risk warrants infringements on privacy and freedom is tricky without clear rules and guidelines.

Theme 3: Differences in guiding approach
Care workers were also asked what guiding approaches they adopt in dealing with the challenges regarding to the OL of clients. After analysing the data, four core guiding approaches are presented here: (1) the passive approach, (2) the reactive approach, (3) the proactive approach, and (4) the restrictive approach.

Passive approach
In practice, care workers often adopt a passive approach, especially when their know-how on the OL of clients is limited. Care workers stress that the OL of clients are seldom part of the care planning process as this is in large part a problem-oriented process. Because in practice clients usually do not raise any issues related to their OL and care workers often know little about it, care workers often do not feel the need to address the OL of clients. Moreover, protocolised care plans do not prompt care workers to ask about the OL of clients when forming goals. Additionally, many of these goals are heavily influenced by the municipalities who fund the care that clients receive which leaves care workers less room to add or change goals. As a result, care workers like Jennifer often give priority to "real world" issues when weighing and balancing their time working with clients: Interviewer: "How often do you talk to your clients about their OL?" Jennifer: "Less often than I would like to, but there are so many other issues I need to talk to them about (…) Usually they are in their rooms or away and when you do see them, you want to address the things that have more priority regarding their futures, like where they have to live for example. Those things seem more important than talking about their OL." Additionally, some care workers even say that if clients do not seem to be at risk online, giving clients online guidance is not part of their professional domain.

Restrictive approach
Staying passive is not always an option for care workers as restricting clients in their OL sometimes feels unavoidable to protect clients from doing harm to themselves or others. According to care workers the need to restrict clients increases when they experience difficulties in talking to clients about their OL, when they suspect that clients are in real danger online, when online activities might compromise the goals as set out in care plans, or when earlier attempts to give advice failed. Care workers say that frustrations about the inabilities of clients in learning from mistakes build up to the point that it becomes exceedingly difficult for them to see things from the client's perspective as care worker Tom explains: I think that most of us see that we must give them advice and coach them, but we can also feel frustrated because clients don't get it (how to behave online). (…) And at a certain point you stop listening to a client and stop trying to see things from his perspective and just tell him how to behave online.
However, care workers are often not allowed to restrict clients without a legal foundation. In most cases, clients cannot be forced to hand in their phones or have their Internet access denied. Care workers work around this by gently persuading clients that it is in their own best interest to agree with certain restrictions.
Care workers sometimes also pro-actively decide to take restrictive measures without the permission of clients or a legal foundation. Care workers give several reasons for this. First, because their own privacy is at stake, for example when they discover that clients made videos of them during work and uploaded these videos on YouTube. Care workers address this to clients, but some feel the need to protect themselves by making clients hand in their phones while they are in the communal area of the house.
Second, when clients are unwilling to talk about their OL and shield their online activities, care workers often feel unequipped to give proper guidance. Care workers who perceive clients to be at risk online or think that online behaviour has a negative effect on their offline lives (e.g., clients not getting out of bed after all-night gaming sessions) may turn to unorthodox measures like turning off the Wi-Fi after 10 pm or taking power cables from their rooms so gaming consoles cannot be used.
Third, sometimes care workers stumble upon conversations if clients forget to log out of their social media accounts on communal tablets or laptops. Although care workers do not see this as an opportunity to monitor the OL of clients, the temptation to take a quick look at a text message can be hard to resist. Things become muddier when care workers look at online content of clients when they have befriended them online, use WhatsApp to communicate with them, or when they know that clients have open profiles online. Although care workers do not always regard this as snooping around, it sometimes results in giving clients advice to do things differently online or, when care workers believe that clients are at risk, restrict or monitor their time online.

Reactive approach
When care workers know more about the OL of clients, but do not perceive clients are in danger online, they often tend to adopt a reactive wait-and-see approach, especially with clients who seem knowledgeable or even tech savvy. They are constructive in their approach, as they are willing to help clients with questions about their online activities but are not keen on meddling in the online affairs of clients. Care workers give three reasons for this. First, care workers deem it important that clients have freedom online and prying into their OL without indications that a client is in danger feels like an infringement on their freedom.
Second, care workers often refer to the pedagogical aspects of their work. They try to appeal to their client's self-responsibility in making the right choices online but also promise to be non-judgmental and supportive if a client makes mistakes. Care workers want to protect clients from online harm, but from a pedagogical perspective they deem learning through trial-and-error as an important method that helps clients grow. Moreover, care workers often believe that exercising more control would be detrimental to the development of responsibility and independence online.
Third, care workers explain that knowing too much about the OL of clients is risky because it may give clients the impression there is a shared responsibility online. While exploring this topic with care worker Susan she shares something that is illustrative for how her colleagues think in this respect: If we see too much and have more control over it (their online activities), it becomes more personal which increases the chance they'll get really angry with us and blame us for things that went wrong. If we leave it with them and things go wrong, they'll know they can come to us, but I'll be their mistakes and not ours. This way we maintain good relationships with them.
As the quote above shows, dealing more proactively with online issues of clients might result in blameand responsibility-shifting and may push clients away. Therefore, care workers can be reluctant to help clients store their online passwords, give advice on how to respond to a hateful comment on social media or discuss if they should date someone online. Guiding reactively, however, gives care workers the opportunity to give support without giving clients the impression that they are pushing them for information or that they are willing to share the consequences of what clients do online.

Proactive approach
The proactive approach seems to be the middle ground for care workers who want to prevent online harm to clients whose OL are perceived risky, but also want to be constructive and assist clients in leading a positive and inclusive digital life. Care workers explain that a proactive approach requires having an open and continuous dialogue with clients on their OL. These investments help care workers build rapport which increases willingness of clients to share online experiences and receptiveness to feedback and advice. Care workers, like Olivia, say guiding proactively requires patience and keeping an open mind: They really like it when you show some interest. And when I ask them stuff like: tell me, how do you do this, they are more willing to share things with me. But I think that it's important not to be judgmental and to keep an open mind. If you do give them feedback, try to be careful with this. (…) It is better to ask them questions so they can start thinking about what they do. This way, I don't need to get all strict on them, which doesn't work anyway.
By being more involved care workers realise that some loss of online freedom and privacy of clients is inevitable but also beneficial to their independence and resilience online. Despite their own insecurities care workers try assisting clients with online activities that can be difficult for clients, like filling in job applications, applying for governmental support, contacting insurance companies, communicating with others on social media, or how to enter a contract.
Care workers sometimes deem it necessary educating clients online because they often lack social support from family or friends. Care workers try teaching clients netiquette rules and how to avoid online risk. Some care workers only point clients to certain apps or websites for information, or involve a behavioural health professional, while others take matters into their own hands by giving clients advice while they are chatting online. One care worker went even further by taking part in a WhatsApp group with a client and his ex-girlfriend who were unable to maintain a healthy relationship with each other. This enabled her to educate her client on effective and positive communicate styles.

Connection between care workers' perceptions on tangibility and risk in the OL of clients and their guiding approaches
The link between care workers' perceptions and their guiding approaches is illustrated in the conceptual model that is presented in Figure 1. The model stresses that guiding processes are often subject to change (hence the dotted lines) as new developments may present themselves over time (e.g., a care worker or a client may learn new insights, or might discover risky online activities).
The model shows that the four guiding approaches have two distinctive elements. The first element is the type of action care workers are willing to take when considering getting involved in the OL of a client. Care workers who perceive clients to be at low risk and have little confidence in their abilities to learn more about the OL of clients often choose to stay passive. They would rather shift their focus on more tangible issues in the offline world. Care workers become more actively involved when the OL of clients become more tangible or when they believe that clients are at higher risk online. The actions they take seem to be more constructive when they want clients to learn something. However, care workers may also take more restrictive actions when their perception of online risk changes. These changes can be caused by their failed efforts to educate clients, or when they deem clients at risk online and feel that taking restrictive actions is the only way to neutralise these perceived risks.
The second distinctive element is that care workers differ in the timing of their actions. Constructive care workers can either be more reactive and adopt a waitand-see approach as they deem it best for clients to learn from experience and mistakes. When they think clients run more risk online, care workers often choose to time their actions differently. Sometimes they believe that taking constructive actions proactively they might strengthen resilience that may help to prevent online harm. However, many care workers reactively, or even proactively try to restrict clients when they believe that certain online activities are bad for clients, or when they are confronted with the consequences of certain online activities (e.g., having a police officer at the front door because a client has been used as a money mule).

Conclusion and discussion
This study set out to answer two questions: (1) What challenges do care workers see in guiding young clients with intellectual disabilities in their OL? (2) What guiding approaches do care workers adopt in dealing with the respective challenges and are they connected?
The findings in this study expand on previous studies that investigated online barriers and challenges of professionals who work with people with intellectual disabilities by showing first the importance of care workers' perceptions of the OL of clients and subsequently how these perceptions may affect their guiding approaches. First, this study highlights that care workers feel they have limited understanding of the OL of clients and think they are unable to change this. Their insecurities are prompted by clients who are unwilling to share parts of their digital lives, but mostly their belief that they are either too old or ill-equipped to connect to clients online. Nevertheless, these beliefs are shared by care workers of all ages and do not seem to be limited to older care workers. A similar picture emerges from studies among police professionals which conclude that age does not seem to play a key role in determining which professionals have more affinity with the digital world (Jansen et al., 2020;Zuurveen & Stol, 2020).
Second, the results also reveal that risk perceptions of care workers have a profound impact on the choices they make in giving online guidance to clients. These perceptions may be influenced by preconceptions of clients, their own fears of the online world, but also by the moral standards that care workers have (see also Ramsten et al., 2019). Care workers seem to move on a moral continuum between online freedom and safety (comparable to Löfgren-Mårtenson et. al., 2018). On the one hand many care workers seem to have a firm belief that clients should make their own, private, and free choices online, but on the other hand they sometimes have to deal with online behaviour that in the care workers' perception warrants protection or guidance. This creates a paradox between promoting online freedom and protecting clients from abuse and harm in their OL through education or restriction (in line with Cobigo, et. al., 2016). As care workers often have to make decisions without clear guidelines and organisational support (De Groot et al., 2022;Nikken et al., 2018), their perceptions might heavily influence the individual choices they often make in dealing with this paradox. This is reminiscent of the Thomas theorem as coined by Thomas & Thomas (1928, p. 572): "If men define situations as real, they are real in their consequences". This is not a problem per se, but becomes problematic when their assessments of the OL of clients are based on fear about the online world or negative preconceptions of clients and their OL.
Although findings in other studies suggest that guiding approaches are largely based on individual choices (e.g., Löfgren-Mårtenson et al., 2018), the model introduced in this study illustrates there are at least four types of guiding approaches that care workers adopt in their work. Choosing between these approaches depends on three things: (1) how tangible and risky the OL of clients are to care workers; (2) What type of action care workers take; 3. Timing of these actions. In practice, this means that care workers operate on two continuums: First from being passively to being actively involved in the OL of clients, and from taking constructive to taking restrictive actions.
The model raises an important question about the consequences of these guiding approaches for the rights and liberties of people with intellectual disabilities under the CRPD. Taking a constructive-active approach can be good, but what if it results in doing what care workers deem best for clients? Will clients still be able to make free choices online? This line of reasoning also applies to care workers who take a passive, or constructive-reactive approach. Are clients' rights and liberties guaranteed by giving them online freedom without guidance or by letting them experiment until they get into trouble? Furthermore, restricting the OL of clients may sometimes be necessary to keep them safe, but what if these restrictions are the result of a care worker's own negative perception of the OL of clients or the Internet in general? In addition, the lines between giving advice and support and restricting clients may be blurred as care workers use (problably not always deliberately) some form of soft coercion when they try to persuade clients to do things differently online. This may diminish the rights and liberties of clients under the CRPD.
Another important question that follows from the above is what moral justifications warrant placing restrictions on the OL of clients or (softly) coercing them in a direction that care workers deem best for them? It might be interesting to compare these justifications to Feinberg's liberty-limiting moral principles that help to justify infringements on the liberties of people, in particular the harm principle (preventing harm to others), the offence principle (preventing offensive conduct to others), and legal paternalism (preventing selfharm). Feinberg, nor many other legal philosophers, seem to oppose the idea of the harm principle and the offence principle as valid reasons to restrict liberties of people (Strikwerda, 2014). Although care workers are alert to and aware of clients who harm or offend others, they are mostly concerned with clients harming themselves online. This form of paternalism is often thought to be wrong (Birks, 2018). Feinberg himself regards paternalism as a preposterous doctrine that deprives people of the right to choose for themselves (Feinberg, 1986, p. 3.) However, Feinberg also realises that in reality paternalism may be valid under the condition that self-regarding harmful conduct is substanially nonvoluntarily or when temporary intervention is necessary to establish voluntariness (Feinberg, 1986, p. 9). According to Feinberg, making a voluntary choice requires time, information, a clear head, and highly developed rational faculties.
Care workers who restrict or coerce clients to protect them from self-harm often make a case for paternalism by questionning the extent to which clients with intellectual disabilities are able to behave safely and responsibly online because of their intellectual challenges. If care workers question the ability of clients to make voluntary choices, this could morally permit them to protect clients from making risky choices. However, this line of reasoning denies people with intellectual disabilities the right to supported-decision making under art. 12 of the CRPD which lies in in assisting and enabling decision-making capacity rather than using diagnosis and capacity thresholds to deny such enablement (Stavert, 2021). Translated to the practice of care workers, it might be difficult to provide supported decision-making when one must deal with high-risk online behaviour without proper knowledge, tools, and guidelines. Next to implementing the positive risk model, the literature provides alternative ideas on how to provide support-based decision-making and prevent coercive measures even where (mental) health and safety are at risk (Zinkler, 2019;Zinkler & Von Peter, 2019).
The above underscores that care organisations need to stimulate care teams to talk about the the good and bad of the OL of clients, and how they can support clients online without diminishing the rigths and liberties of clients under the CRPD. To this end, the conceptual model as presented in this study could be used to reflect on the guiding approaches care workers adopt and how their perceptions shape these approaches. Addittionally, care organisations could adopt a positive risk taking approach to find a middle ground between promoting freedom and providing safety, as it helps both care workers and clients to identify the benefits and risks of being a digital citizen. Through a shared-decision making process clients gain more control over their digital lives, while care workers may develop better strategies that aim to mitigate risk and promote online opportunities for clients.

Limitations
This study has three important limitations. First, accurately reproducing experiences in the past is difficult. Asking care workers to recall and evaluate certain choices they made in the past might not produce accurate depictions of these experiences. Furthermore, it might well be that care workers evaluate their choices of the past with knowledge and experience of the present.
Secondly, although the results did not draw out any specific differences in age or gender (see also Löfgren-Mårtenson et al., 2018), it might be that these factors, when looked at more carefully, affect risk perceptions of care workers and consequently their guiding approaches. This might be worth measuring in future studies. Additionally, although the conceptual model may help to reflect on online guiding approaches, it needs more empirical research and might need to be refined in the future.
Finally, generalisability of the findings is difficult because of the small respondent sample that was drawn from one care organisation. The described challenges and guiding approaches may not represent those from other care organisations. Also, most respondents in this study worked with young people in the age between 17 and 27. The findings may therefore not be representative for care workers who work with younger or older people. Finally, all professionals in this study work with young people with mild intellectual disabilities. Their observations may not reflect the experiences of professionals who work with young people with other types of intellectual disabilities. Note 1. In the Netherlands, a mild intellectual disability is defined by having an IQ between 50 and 70, while borderline intellectual functioning means having an IQ between 70 and 85 with diminished adaptive functioning.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was supported by the Dutch National Research Council (NWO): 023.004.084.