Exploring Collaborative Patterns in Neurodiverse Teams: A Hidden Markov Model Approach Using Physiological Signals

Abstract

Autistic individuals face challenges in successful employment, emphasizing the need for targeted workplace support. This study explored collaborative dynamics within neurodiverse teams during a simulated remote work task by applying Hidden Markov Models (HMMs) to heart rate data. Eighteen participants formed nine dyads: six nonautistic (NA-NA) pairs and three autistic-non-autistic (ASD-NA) pairs. Dyads completed two trials of a collaborative programming task over Zoom, alternating roles between trials. Heart rate data were collected, segmented, and transformed to extract features reflecting participants’ interactions. The final HMM was fitted with seven hidden states, and transition probabilities were derived for each dyad type. Results showed that NA-NA dyads exhibited more frequent transitions among states compared to ASD-NA dyads, potentially suggesting more varied interaction patterns. These findings demonstrate the utility of HMMs in capturing collaborative behaviors through physiological signals and highlight their potential in helping develop effective support strategies for neurodiverse teams.

The prevalence of autism spectrum disorder (ASD) in the U.S. is increasing, with reports indicating that 1 in 36 children and 1 in 54 adults were identified with ASD in 2020 (Dietz et al., 2020; Maenner et al., 2023). Despite rising awareness, autistic individuals face significant challenges transitioning into the workforce, with unemployment rates ranging from 50% to 80%, significantly higher than those for individuals with other disabilities (Richards, 2012; Taylor & Seltzer, 2011).

Today’s work environment demands skills such as communication, complex problem-solving, and teamwork (Cascio & Aguinis, 2008; Sousa & Wilks, 2018), which often do not align with the characteristics associated with autism (Bury et al., 2021; Vanbergeijk et al., 2008). This misalignment contributes to the challenging employment outcomes for autistic individuals. However, autistic individuals possess unique strengths like exceptional attention to detail, reliability, and efficiency (Cope & Remington, 2022), highlighting the critical need for targeted workplace support.

Developing effective workplace support requires a deeper understanding of the collaborative dynamics between autistic and neurotypical individuals. This study explores the temporal dynamics of collaborative behaviors in neurodiverse teams during a simulated remote work task by applying Hidden Markov Models (HMMs) to heart rate data. The study aims to capture the different states and their probabilistic transitions during collaboration, informing the development of more effective and sustainable support strategies for collaborative tasks in various workplaces.

The study recruited a purposive sample of 18 participants, forming nine dyads: six non-autistic (NA-NA) pairs and three autistic-non-autistic (ASD-NA) pairs. The dyads completed two trials of a collaborative programming task over Zoom, with participants working from separate rooms for up to 15 min to accomplish the task goals. Heart rate data were collected using Empatica E4 devices and segmented into 30 s intervals to capture the temporal dynamics of collaborative behavior. Features reflecting participants’ interactions were extracted using a discrete wavelet transform.

The final HMM, fitted with seven hidden states, was used to estimate the sequence of hidden states for each dyad type using the hmmviterbi algorithm. Transition probabilities were derived and visualized to identify distinct interaction patterns between NA-NA and ASD-NA dyads. The analysis revealed that NA-NA dyads exhibited more frequent transitions among different states, suggesting varied interaction patterns, while ASD-NA dyads had a higher number of zero-probability transitions, indicating prolonged stability in specific states before transitioning. This pattern may reflect a preference for consistency and predictability among ASD-NA dyads.

The study demonstrates the utility of HMMs in capturing collaborative behaviors through physiological signals, highlighting the potential for developing effective support strategies for neurodiverse teams. The distinct interaction patterns suggest that NA-NA dyads engage in more exploratory behavior, frequently adjusting their strategies, while ASD-NA dyads prefer structured and predictable interactions.

Future research should aim to generalize findings across different datasets and conditions, incorporate additional physiological indicators, and tailor models to specific dyad characteristics. Such enhancements can facilitate the development of supportive strategies that improve collaborative experiences and outcomes in neurodiverse workplaces, contributing to a more inclusive and effective workforce.

By understanding the distinct collaborative behaviors of neurodiverse teams, employers can create more supportive and accommodating work environments that leverage the unique strengths of autistic individuals. This approach not only benefits autistic employees but also enhances overall team performance and workplace inclusivity. Targeted support strategies that address the unique needs and strengths of autistic individuals are essential for fostering successful and equitable employment outcomes in an increasingly neurodiverse workforce.