Designing for Strengths: Opportunities to Support Neurodiversity in the Workplace

Many supported employment technologies focus on pre-placement job skills such as interviewing [3, 4, 73, 74, 89]. Virtual reality (VR) is a common mode of training for workplace skills, and has been utilized in many recent studies as well. For example, Smith et al found simulated job interviews in VR can significantly improve job interview skills and interview self-confidence in autistic people [50, 73, 74]. These systems seek to correct or teach specific skills such as eye contact or appropriate responses with the ultimate goal of removing barriers to employment and addressing gaps in transition support for people with cognitive diversity. However, the systems could result in reduced user agency by focusing explicitly on a perceived deficiency in behavior based on neurotypical norms. Further, deficit-centric approaches to technological interventions intensify anxiety among ND job seekers [35], as their focus shifts towards their shortcomings rather than the development and application of their unique strengths and capabilities within workplace contexts.

Other researchers have investigated the use of personal devices for support in the workplace setting [36, 37, 44]. Gentry et al. found that introducing an assistive technology (AT) with task reminders, task lists, video-based task-sequencing prompts, behavioral self-management adaptations, way-finding tools, and communication with the job coach via WiFi can significantly reduce the need for in-person coaching support for ND employees without reducing functional performance. The intervention group in this study required 1.56 times less job coaching hours than the control group at twelve weeks. Similarly, Hill et al found that comparable AT such as iPads, iPods, and iPhones with applications specifically developed to support a structured work day (medication management, time management, emotional coping strategies) are capable of increasing independence, job placement, and job retention for ND individuals [44].

While the AT mentioned above offers tools for ND employees to self-manage their needs instead of teaching neurotypical skills, they still fail to provide personalized strategies for finding jobs, completing work tasks, and addressing workplace challenges that empower ND individuals to explore, cultivate, and apply their unique strengths. A 2018 review of 134 studies investigating employment interventions for people with autism found that none of the studies used a strengths-based approach to intervention [70], however, ND employees want their strengths and abilities highlighted in workplace settings [59]. Further, a 2022 literature review found that there is a significant gap in research that investigates the strengths of ND individuals [23].

The Tree of Life (ToL) collage activity was designed to allow participants to understand their self-reported strengths and values, and to synthesize these findings into artifacts that can be used as a reference in the rest of the co-design methodology. This activity lasted two study session periods. Participants were given a template of a tree on which each section–roots, trunk, and branches– represented a different part of their lives. They were asked to paste, draw, or write representations of themselves based on prompts for each section. Fig. 1 shows the prompts included in the template. The purpose of including roots is to understand one’s background and to engage the participants in thinking about how their background has shaped their personalities and characteristics. The trunk represented skills and strengths, and this portion of the tree was intended to initiate language and considerations of one’s strengths. The branches represented dreams, whose purpose was to encourage participants think of the future they envision for themselves and their motivations.

This activity was intended to surface the types of questions that our participants wanted to ask a virtual job coach, and lasted two research sessions. Researchers demonstrated a prototype of a ChatGPT-powered job coach chatbot (Fig. 2). The experience included an animated job coach avatar that the participants could verbally ask questions to and converse with, and the interface simulated a virtual call. This simulation was deployed on the researchers’ laptops, which participants could access. The virtual job coach employed ChatGPT 3’s davinci model for text generation. For example, the initial interaction began with the greeting, "Hi, I am your job coach. How are you doing today?" This opening statement, combined with the participant’s response, served as a prompt for ChatGPT to formulate subsequent dialogues. Each participant’s reply generated a comprehensive prompt for ChatGPT, which, in turn, influenced the virtual job coach’s spoken responses.

After a researcher demonstrated the virtual job coach’s abilities with an example conversation, participants were allowed to interact with the technology and ask questions that they might have for a virtual job coach. These interactions were video-recorded.

A crucial part of this co-design method is being able to effectively translate strengths and values into design requirements that will help our participants achieve their goals. From the ToL activity, we had data on our participant’s strengths and values, and the virtual job coach inquiry activity surfaced in what workplace situations they wanted support from a job coach. To create our target design requirements of strengths-based strategies to support our participants in the workplace, we engaged them in creating user stories that reconciled our current data. User stories are a user research method that uses a narrative to represent a user group’s goal and is designed to facilitate discussion of system requirements that will help them achieve that goal [19]. Our strength-contextualized user stories took the form of: "As <a user with certain strengths>, I want to <task/goal>." Using the user stories, discussions could focus on how a system can help that particular user achieve their goals. This activity lasted three study sessions.

Tasks were derived from questions that participants asked during the virtual job coach inquiry activity. For example, P6 asked the virtual job coach "My new co-workers have not been very nice to me... I’m not sure what to do right now?" which resulted in a user story template that included "As P6, I want to know what to do when I am not getting along with my coworkers." A list of the participant’s strengths from the ToL activity was used to motivate the discussion. For example, if one person’s strength was empathy, the group leader would prompt the discussion with "How could the system use your strength in empathy to help you achieve this task?" Each researcher recorded the key points, features, and strategies that participants discussed for each task. This activity revealed strategies that our participants find useful in different situations and would want available to them via a virtual job coach system. Further, the strengths-focused nature of the discussions allowed participants to consider what strategies would be best suited to their individual strengths.

This activity was designed to guide participants in combining all the previously made artifacts and user research into tangible paper prototypes. In particular, prototyping was intended for participants to design features that would support the strengths-focused strategies from the user stories. We then evaluated these prototypes for rationale and motivation in the design choices. In total, this activity lasted three study sessions.

Design: To create paper prototypes, researchers provided each participant with a packet that included their personal aggregated data from across all the previous design activities (ToL, Questions asked to the virtual job coach, User Stories), and a list of questions that guided the student to consider how all of these things might come together in a usable technology. In the packet, participants were given the questions they asked to the virtual job coach, and further prompted: How will the job coach help you do this using your strengths?, What are one or more strategies that the job coach can offer you?, What does your job coach look and sound like?

Participants were provided with paper, pens, and collage materials to create their designs based on the answers to the packet questions, or other inspirations. Collage materials included clip art of technology such as a blank computer screen, phone screen, VR headset, and apple watch, as well as different options for avatars (people, robots, characters), and features (chat interfaces, buttons, lists).

Reflection: The final activity of our strengths-based co-design method included interviews of each participant about their prototype and rationale for design choices. Participants were asked to explain in detail how they envisioned using the features in their prototypes, and why that feature or strategy was something they thought was useful. Researchers leading these interviews recorded all interviews.

When the backgrounds of the participants were discussed during the Tree of Life activity, all thirteen participants reported that their family and friends were a key source of support and encouragement in their lives. P7 said that her family "makes [her] happy and is there for [her] no matter what." Similarly, other participants indicated that their families are a significant source of support because they will be by their side during hard times and when problems arise (P2, P4, P12), take care of them if anything happens to them (P4, P11), and care for and love them (P2, P4). Additionally, relationships with friends were commonly brought up as sources of support for participants. P1 said that her friends are important because "they appreciate [her]". The act of valuing relationships indicates love as a core strength and powerful motivator [62].

When discussing friendships and social relationships, the most common characteristic that participants reported as making them good friends was kindness (P1, P4, P5, P7, P11), a strength that is characterized as helping others [62]. Further, two participants (P6, P11) claimed humor was a strength that supported their friendships. Humor is a strength that is distinguished by one’s ability to impact others in a positive way [62]. Love, kindness, and humor are all strengths that involve other people in order to exhibit them, making them highly relational. These findings support previous research regarding the impact of community, family, and friends on ND individuals and their employment outcomes [36, 53]. For example, ND individuals experience better employment satisfaction when they reside with their families [36], and structured programs that involve parents have been shown to improve the social experiences of autistic individuals [53]. These findings expand the scope of this concept to include love, kindness, humor, and other relationship-driven values as strengths that should be considered in workplace technological interventions.

The second half of the ToL activity was designed to elicit discussion on participants’ strengths and values specifically in a workplace context. Being a "hard worker" was the most commonly reported characteristic of being a good employee (P2, P7, P8). This can be understood as perseverance; a strength associated with persisting in a course of action despite obstacles [62]. One participant’s rationale for valuing his work ethic was that "hard workers get things done no matter what (P8)." Another participant said she was a good worker because she closely listened to her manager so that she could "do things right and be a better employee (P11)." These quotes indicate that our participants value their ability to meet their job requirements effectively and dependably.

Another common characteristic participants related to the workplace was teamwork (P4, P7, P12), with one participant claiming working with others is "better than doing it by yourself (P4)." Being professional during the teamwork was a characteristic that P12 found important, claiming that when you are faced with challenges at work, "you don’t have to be rude." Teamwork is a strength characterized by the ability to be a loyal member of a group [62], and not being rude to others can be thought of as an act of kindness. As with previously discussed strengths of love and humor, and kindness, teamwork is highly relational and exhibited in contexts with other people.

In line with the finding that our participants value their relationships with others highly, when we asked participants what their motivations are for their dream jobs, many of the answers indicated that they are motivated by interactions with others. For example, P11, whose dream job is to be a baker, wants to share her baked goods with others. P4 wants to be a Youtuber so that he can share videos and ideas that he enjoys with other people. P1 wants to help other people with disabilities by contributing to the design of physical spaces and technology. P2 wants to be a doctor or a vet so that she can help people and animals. Other participants chose dream jobs that related to their specific interests, like P6 who loves football and wants to be a NFL equipment manager, and (P8, P12) who both work in IT because they enjoy working with technology and felt they had strength in the field. In all examples, the participants were motivated by either their values–helping or sharing with others–or their personal interests.

Many participants (P1, P2, P5, P6, P7) indicated that a system that can support their emotional state in the workplace would be a useful resource. Strengths that support emotional support strategies used by our participants include self-regulation and hope.

Self-Regulation: Self-regulation is a strength that allows one to control their emotions and remain disciplined [62], and we found that regulating emotions was frequently the first strategy that participants turned to when discussing how to address stress at work. For example, P6 designed his prototype to support him when he is being asked to do a lot of tasks at work and feels overwhelmed. His job coach prototype included a feature to help him remain or regain calm through guided breathing exercises. During the evaluation, P6 stated that this is a strategy he frequently uses at work for self-regulation. He also included a wayfinding feature that led to the break room so he could remove himself from the stressful situation (Fig. 3). Gentry et al found that similar wayfinding and map features are effective at supporting ND employees in the workplace. P6’s prototype suggests that this wayfinding feature can further support his self-regulation by reducing the cognitive load required to navigate to a break room or outside when emotional regulation should be his main focus. Once removed from the situation that is causing overwhelm, P6 utilized features in his prototype that use his strength in humor to help him re-gain a calm composure, suggesting that the job coach could prompt him to think of a funny story to help "relax [his] brain."

Other participants designed both similar and different strategies that support self-regulation. Some strategies directly addressed the feelings of stress, such as job coach-led counting exercise, or a list of activities that could help manage stress such as taking a timed walk (P7). Other strategies allowed the participant to explore self-regulation options for themselves, such as resources for mental health support such as a list therapy options or a direct line to their current mental health counselor (P5). Interestingly, P7 included calling a parent as a strategy to help her re-regulate, which calls back to the love and relational strengths discussed in previous sections. Because our participants value their relationships with those they trust, this strength of love can be utilized to support their self-regulation in the workplace as well.

Hope: Beyond supporting self-regulation, participants discussed strategies that their job coaches could offer to help maintain a positive mindset in the workplace. In an open discussion during the user stories activity, P2 discussed the ability of a job coach to help her when she and her co-workers are not getting along. P2 said she would want the system to encourage her to try to resolve the tension by reminding her that "if [an attempt to resolve the tension] doesn’t work out, there are other friends you can make. But to remain professional with co-workers that you may not personally get along with (paraphrase, P2)." In the same discussion, P1 said that she "wants the system to encourage [her] to keep using [her] strength (paraphrase, P1)." Hope is a strength characterized by optimism about the future, belief in one’s actions, and feeling confident things will turn out well [62]. Both of these strategies suggest that participants want to be supported in feeling hopeful about the challenges they are facing, and encouraged that they are capable of overcoming them.

Similarly, many participants invoked the importance of encouragement in their virtual job coach designs. For example, P1 designed her prototype to support her in completing a new work task, and a strategy she included was the job coach offering praise because "the more reinforcement the better (P1)." P5 wanted her virtual job coach to encourage her with phrases such as "You’re doing a really good job... Keep on doing it." We found that such reassuring language can be thought of as bolstering perseverance, hope, and the confidence of the individual. For the workplace context, it can help increase feelings of agency by reminding the individual that they have the strengths and skills to address workplace challenges. Further, studies on positive psychology have shown that by framing language in a positive perspective, individuals have more confidence in their competence and a more positive outlook on their capabilities to affect change in a situation [66]. Therefore, our findings suggest the concept that positive language used for encouragement would benefit a strengths-supportive job coach technology by keeping the employee’s mental perspective positively framed.

Following emotional regulation, participants discussed how a virtual job coach could support them in approaching work-specific social interactions.

Judgement, Critical Thinking, & Empathy Challenges between participants and their co-workers or workplace customers were a frequent topic of discussion. During the virtual job coach inquiry, P6 asked the virtual job coach "My new co-workers have not been very nice to me... I’m not sure what to do right now?" Regarding customers, P7 said "[when] I have customers complaining, [it stresses] me out, and I don’t know what to tell them." During the discussion about how a virtual job coach could support the users in these situations, participants expressed their desired support in actively resolving tensions. For example, P2 specifically mentioned "ask[ing] the co-workers what the problems are, and then try[ing] to problem solve using... empathy. [You] can be curious and try to come up with ideas to problem solve (P2)." Here, the strategy of asking questions about the co-worker’s perspective aligns with judgment and critical thinking strengths as those indicate the ability to think through things and examine them from all sides [62]. Participants specifically mentioned strategies that a job coach system can offer such as recommending possible questions to ask co-workers about the tension (P2), ways to facilitate friendships such as inquiring about common interests via text message conversations (P2), or offering ice-breakers to open conversations (P1, P3).

Additionally, P7 said she would want a job coach to first remind her "to keep a positive mindset and behavior towards the customer," a strategy that calls back to previously discussed self-regulation methods. In order to resolve customer issues, similar to resolving tensions with co-workers, P5, P7, and P9 suggested effective strategies would be asking questions to learn more about the customer, and responding based on the customer’s concerns. P5 wants her prototype to "help [her] see what [she] can do to de-escalate the problem with a customer" and included features such as a list of questions she can ask to get to the root of the customer’s problem, and reminding her to consider the customer’s perspective and respond with empathy. While many technological interventions for ND individuals focus on providing solutions to social shortcomings of the ND population through assessing eye contact during interviews [89], or emotional recognition [9], this finding suggests that instead, our IDD participants want a supportive technology to offer actionable ways for the individual to interact with the specific situation and be encouraged to resolve it through more relational and empathetic strategies.

The need for work-related information was a common workplace experience that our study participants designed their job coach prototypes to address. For example, P1, P2, P6, and P12 designed their job coach prototype to help them learn new work-related skills while in the workplace.

Perseverance: We found that when our participants needed information or to complete a new task, they wanted resources to learn the task themselves. Similar to findings in Professional Values (5.1.2), we attribute this to the strength of perseverance, which is persisting in a course of action despite obstacles [62]. The prototypes of all four participants included their most effective strategy to learn new skills—-breaking down the new task into a list of steps. P1 stated that breaking things down into steps would help "so [she doesn’t] get overwhelmed," a strategy that also supports self-regulation, and she drew features to support this such as a visual list or checklist of actions she can take or next steps for her to follow. Further, all four prototypes included features that provide resources for learning the tasks such as videos and tutorials (Fig. 5). P8 stated he would like "links to watch videos to see what kind of steps you need to do... [and it could] give you some examples... to help you learn... better." Similarly, P3, P4, and P10 designed their prototypes to assist them in accomplishing the specific work tasks of building a playlist for a birthday party at a bowling alley, re-stocking shelves, and completing online documentation, respectively. Two of these designs included a list of steps as support for accomplishing the work task (Fig. 4). P6, P9, and P10 designed summary and read-aloud features of the steps and resources provided by the job coach (Fig. 5). P9 said the job coach would "share her screen with an article... that would help us [by] reading it step by step. [...] because that could help us understand what are we supposed to do next," indicating that the system can support them in accomplishing tasks correctly and helping them learn how to accomplish the task. P8 said a job coach technology providing resources can "help you learn [the skill] better" in order to be better at the job. These findings indicate that the use of step-by-step instructions and resources for learning as strategies for workplace support aligns with the desire to be a good, reliable worker, the most commonly cited value in the workplace during the ToL activity. This step-by-step feature has been found as useful in other workplace support technologies that utilize video- and audio-based modalities [71]. Video prompting allows the employee to independently navigate through the steps of the task, or repeat steps to successfully complete a job [10, 48], while video modeling can help employees learn and maintain vocational skills, and even generalize those skills over different tasks [5, 43, 68, 83].

An integral part of strengths-based psychology is to surface strengths by analyzing successful strategies that the individual has used in the past [2]. By doing so, individuals are able to make the connection between a past challenge and a current challenge, and are better able to understand their ability to address it. Similarly, we found that participants wanted their job coach systems to acknowledge their past success in a particular task through prompting questions and verbal affirmations of their past success. P4 designed his prototype to assist him with the specific work task of re-stocking shelves, a task he has accomplished before at work. He wanted his job coach to encourage him by reminding him that he has completed this task before, and can use the same strategies to do it again. "The job coach knows you’re intelligent... So the job coach wants to remind you... look, do you remember the last time this worked and then... the job coach will give the same suggestion (P4)." This participant recognized his abilities and designed his prototype to remind him of his strengths and encourage repeated use of successful strategies. When discussing this prompting concept, P12 said that she liked being asked questions that make her reflect on her own strengths and experiences, allowing her to come up with her own strategies to address the situation.

Many participants specifically discussed and designed features that allowed them to employ their judgment and decide the best course of action for themselves. For example, P1, who designed her job coach prototype to support her in learning a new work skill, included a feature where the job coach would suggest different courses of action for her, and she could choose which one she wanted to use: "So he’s giving you options, and it’s utilizing your strength of being able to evaluate which options are best to move forward (P1, paraphrase)."

Similarly, P2, who designed her prototype to help her stay focused when she is disengaged or bored at work, wanted her job coach to present two different strategies to help her re-engage. P2 said she wanted the opportunity to "pick which one would be best." During design discussions, P2 elaborated by sharing that she wants "suggestions based on strengths, but not to be told exactly what to do (paraphrase, P2)." Further, P5 designed her prototype to offer her "tips that [she] can approve or not approve" when addressing difficult customer interactions.

Our participants’ preferences for having self-management features in job coach technology such as scheduling, timekeeping, and reminders also represent the participant’s desire to be able to operate independently by managing their own schedules, time commitments, and responsibilities. For example, P8 and P11 both included calendars in their prototypes with reminder alerts for school and work responsibilities. The usefulness and importance of these self-management features in workplace interventions have been also emphasized in previous research, where researchers found that these types of features can increase functional independence in autistic individuals [38].

In line with previous HCI research that cautions against providing explicit technological solutions to a user’s or community’s challenges [34], we suggest that designing for independence requires providing strategies that can help lead the user to address the problem through their own strengths, instead of providing direct solutions. We find that this is particularly true in complex social interactions or workplace situations that require problem-solving. For example, in the case of challenging customer interactions, our participants wanted a list of questions that they could use to investigate the problem themselves, along with reminders to consider the customer’s perspective and respond with empathy. A technology that can provide a list of questions that facilitate problem-solving is supporting the user’s independence by allowing them to consider important aspects of the situation first and act accordingly, instead of being told exactly how to address the problem or to immediately involve a supervisor.

We found that another way to support the user and bolster their independence is through prompting the user to consider similar previous challenges and how they solved them. When the user has previously completed the task they are seeking support for, the system can remind them of this, therefore providing mental scaffolding for the user to recognize that they have the capability of doing it again. This strategy supports independence through bolstering self-confidence, and can be broadly applied. For example, it can be used for self-regulation, skill learning, and social interactions.

Lastly, our findings suggest that another way to support the independence of neurodivergent employees is to provide options for addressing problems, and allowing them to choose which, if any, are best for them to employ. For example, providing a list of therapy options or a series of learning resources for the user to choose and act on can provide the user the ability to choose how to move forward in the given situation, exercising their agency to make the final decision.

We found that relationships with loved ones and community were commonly referred to by participants when discussing values and related to both non-work and work contexts. While there are existing studies on the impacts of community and relations within technology for the ND population [46], little work has been done on integrating those findings into technologies for ’workplace’ support. Most vocational support interventions focus on skill developments to support social interactions such as developing social skills, making eye contact, or reading facial expressions [3, 9, 89]. However, few investigate the integration of supporting the individual through social relationships. Burt et al found that in modified work programs, lack of family support would lead to the intervention failing, and that success of the intervention was more likely if there was heavy family involvement [17]. Additionally, other studies have found that autistic individuals show great interest in sharing sensitive questions or challenging situations with their close social networks, and technology can facilitate this type of interaction with the goal of independence in everyday life [46].

When considering how to incorporate the support of community and loved ones into a workplace-supportive technology, designers and researchers can consider direct lines to known loved ones, such as the call a loved one feature from our participants’ prototype. Other researchers have examined the concept of utilizing community and close social networks to assist ND individuals. The SocialMirror was a design concept developed to support autistic individuals’ everyday decision-making by leveraging an online network of family, friends, or caregivers, to offer help in daily tasks such as professional attire [46]. However, this study raises concerns about conflicting input from different caregivers. Especially for workplace support systems, this can be a challenge facing technology that seeks to integrate community and relationships for support. For example, a loved one of an ND employee may not be aware of the job’s best practices or duties, and might unknowingly give advice to the user that goes against their work policy. Furthermore, for relationships and workplace support technology, there is a consideration to be made for the balance between support from a loved one and independence. To advance our understanding of how to effectively integrate loved ones into workplace technology, future research should explore design approaches for balancing the benefits that ND employees can derive from the strong support of their relationships [17] while ensuring that this support is received without leading to over-reliance.

Our participants valued their ability to be dependable in the workplace, and strengths they used to achieve this were self-regulation and self-management. For example, when a user needs immediate help in self-regulation to regain calm in the workplace, our participants suggested active strategies such as guided breathing and counting exercises. Optimal self-regulation abilities have been linked to positive social relationships, productivity, achievement, and a positive sense of self [12]. Our findings suggest that IDD employees find value in self-regulation, therefore technologies that support this population should incorporate strategies to support this.

Further, self-management strategies that do not require problem solving such as adhering to time commitments are effective direct solutions to support workplace independence. Similar to previous research, our study suggests that IDD participants want these features to support their desires to be dependable, hard-working employees who can operate independently [36]. Therefore, for technologies that are intended to support ND employees in SE settings, features that support their abilities to track and manage their responsibilities are crucial for facilitating agency and independence.