Autism awareness training for law enforcement officers: a pilot study Open Access

With prevalence rates of autism spectrum disorder (ASD) increasing (National Center for Health Statistics, 2023) and victimization rates for individuals with disabilities becoming disproportionately high (Harrell, 2021), the likelihood that a law enforcement officer (LEO) will interact with an individual with autism in a professional capacity is high. Despite these odds, officers have reported a lack of training pertaining to neurodevelopmental disorders, a lack of confidence when interacting with individuals with developmental disabilities and a lack of knowledge regarding the characteristics of autism (Christiansen et al., 2021; Copenhaver et al., 2020; Gardner et al., 2018; Gardner and Campbell, 2020; Herbert et al., 2022; Love et al., 2020a; Rascon, 2022). Therefore, it is imperative that LEOs receive specific and appropriate training so that characteristics of autism are not misattributed to mental health crises (Copenhaver et al., 2020) or signs of criminality (Wallace et al., 2022). In this paper, we respond to the gap in officer training on neurodevelopment disorders and confidence in interacting with individuals with developmental disabilities by describing (1) a brief research-based professional learning session, (2) identification and testing of the Police Self-Efficacy for Autism (PSEA) scale with officers related to working with individuals with autism (Love et al., 2020b) and (3) pilot findings of a pre- and post-evaluation of the brief research-based professional learning session.

Though the estimated rate of autism is higher in California than worldwide (1 in 22 children and 1 in 100 individuals, respectively), worldwide rates are likely higher (Centers for Disease Control and Prevention, 2023). Difficulties in estimation are compounded by a lack of published data in over 80% of countries, differences in diagnostic criteria between countries, changing criteria across time, and a lack of screening tools that are culturally and linguistically relevant (Zeidan et al., 2022). In the United States, the diagnostic criteria for ASD were expanded in 2013. The current criteria of the Diagnostic and Statistical Manual, Fifth Edition define ASD as a developmental disorder characterized by social communication deficits and restricted and/or repetitive behaviors (American Psychiatric Association, 2013). As a spectrum disorder, the presentation of characteristics differs across individuals and to varying degrees. Social communication is a broad category, the specifics of which depend on the social and cultural context. Aspects of social communication may include difficulties interpreting or using body language, tone of voice, figurative language or in providing eye contact to the length expected by the given culture. Restricted and repetitive behaviors may refer to having strong, focused interests on a particular subject, a desire for routines, and repetitive speech or movements (e.g. rocking, pacing or hand flapping). While autism can be diagnosed in adulthood, characteristics of social communication deficits and restricted and/or repetitive behaviors are present since early childhood (American Psychiatric Association, 2013).

When not understood, characteristics of ASD may be misinterpreted as signs of guilt, non-compliance or drug-induced behavior (Diamond and Hogue, 2021; Gardner and Campbell, 2020; Hinkle and Lerman, 2021; Love et al., 2022; Railey et al., 2019; Wallace et al., 2020). The introduction of California Senate Bill 882 (SB 882) stated that “with the absence of a national database tracking [interactions between officers and individuals with developmental disabilities], this bill will require the tracking of such interactions in California” (Senate Committee on Public Safety, 2022). Prior to approval of SB 882 (2022), governmental code 12525.2 required that law enforcement agencies provide a monthly report to the Department of Justice with incidents involving the shooting of a civilian by a peace officer, the shooting of a peace officer by a civilian, and the use of force by a peace officer against a civilian or a civilian against a peace officer resulting in death or serious bodily injury. For each incident, the government code required that the report “include, but not be limited to,” demographic characteristics such as the gender, race, and age of each individual involved. The passage of SB 882 amended this list to add “whether the officer perceived the civilian had a developmental, physical, or mental disability” (California Senate Bill 882, 2022, § 12525.2). SB 882 clearly elucidates the need for increasing LEOs’ understanding of developmental disabilities such as autism.

In response to this need, and after a search for current local LEO awareness training did not yield results, a local police lieutenant and members of a children’s hospital and university partnership (i.e. a multidisciplinary team, or MDT, including a school psychologist, applied behavior analyst, speech-language pathologist and members of the autism community) began conversations on developing training for LEOs. Following a systematic review of the literature to ensure best practices, autism awareness training was developed. This training was named Interaction with Neurodiverse Citizens: Training on Effective Response and Communication (INTERAC) and was developed using an iterative feedback process between the police lieutenant and the MDT’s review of research-based materials to inform the training. The finalized training was condensed into a 40-min session that could be delivered during shift briefing.

The purpose of this study was to articulate a research-based training curriculum, test the PSEA scale as a measurement tool for that curriculum, and then examine the relationship between INTERAC and LEO self-efficacy ratings.

Individuals with autism, their caregivers and first responders have all reported a desire for departments to receive training specific to autism (Christiansen et al., 2021; Railey et al., 2019; Rascon, 2022). Greater knowledge of autism, whether gained from personal or professional experience, is associated with increased confidence in the ability to identify characteristics of autism, effectively interact and be prepared for a call (Copenhaver et al., 2020; Gardner et al., 2018; Love et al., 2020a, b). According to Bandura’s social cognitive theory (Bandura, 1977, 1997), self-efficacy refers to the belief in one’s capability to engage successfully in a task, which influences motivation, persistence and behavior. Due to the connection between knowledge of autism and self-efficacy in interacting with individuals with autism, Bandura’s social cognitive theory was adopted as a framework for this project. Consistent with Bandura’s theory, additional knowledge may inform behavior, though we acknowledge that the stress of different situations police officers encounter could lead to differential application.

Since Bandura first described self-efficacy in 1977, researchers in various fields have drawn upon this theory. Meta-analyses such as (1) Dibapile’s (2012) review of teacher efficacy and classroom management, (2) Tan et al.’s (2021) review of self-efficacy and hypertension management, and (3) Lochbaum et al.’s (2023) review of self-efficacy and sports performance show the breadth of self-efficacy research among different fields. Recently, self-efficacy theory has been applied in the field of law enforcement and in research specific to working with individuals with autism. In law enforcement, research has demonstrated a relationship between confidence in the ability to “handle police work” (Kim and Xu, 2020, p. 211) and plans to stay in the profession; the efficacy of mental health first aid training based on increases in self-efficacy (Arazon and Weich, 2024); and the relationship between self-efficacy, suicide intervention training, and usage of intervention behaviors (Osteen et al., 2020). Regarding professionals who work with individuals with autism, research has shown the following relationships: teacher self-efficacy and job satisfaction (Love et al., 2019), allied health worker experience with autism and self-efficacy in the ability to diagnose and treat (Atun-Einy and Ben-Sasson, 2018), and medical and psychiatric professional knowledge of autism and self-efficacy in providing care to people with autism (Corden et al., 2021). This present project combined the two aforementioned foci: self-efficacy of LEOs when interacting with individuals with autism.

Research has sought to estimate the frequency of interactions between LEOs and individuals with ASD. A systematic review of studies worldwide found that between 7.9% and 32.5% of individuals with autism reported contact with LEOs (studies published in the United States, United Kingdom, Netherlands, Australia, Denmark, Germany, New Zealand, Norway; Cooper et al., 2022). In the United States, research by Gardner et al. (2018), Gardner and Campbell (2020), and Gardner et al. (2022), found that 49%–60% of officers reported answering at least one call involving an individual with autism per year, 40% reported responding to at least three calls per year and 35% reported responding to six or more calls per year. However, depending on the study, between 5% and 40% of officers and cadets in the United States reported having received training with topics on ASD (Copenhaver et al., 2020; Gardner et al., 2018; Gardner and Campbell, 2020; Herbert et al., 2022). Additionally, according to Herbert et al. (2022), 84% of officers who were asked about prior training specific to autism stated they had received Crisis Intervention Training (CIT), a training initially designed to address mental health crises. Associating autism with mental health disorders can be problematic; as Copenhaver et al. (2020) stated, ASD is a developmental disability and not a mental illness or crisis state. Additionally, Gardner and Campbell (2020) found officers who had completed CIT training were more likely to use physical force in calls involving individuals with autism compared to officers without CIT training.

A systematic search of the literature was conducted via EBSCO using search terms: “first responder,” “EMT,” “police officer” or “law enforcement officer,” “training,” “autism,” “autism spectrum disorder” (ASD) and “intellectual disabilities” (IDDs). Filtering by peer-reviewed articles published in the United States in English within five years of training development yielded 166 articles. Articles were screened first by title, then abstract, and last by the results and discussion sections for relevant usage of search terms in context. Twenty articles were retained and read in full. Of these 20 articles, 12 studies surveyed officers on previous or desired training but did not provide training as part of the study. Three were review articles, and five were studies that measured the efficacy of a given training. In what follows, we first detail systematic reviews that laid the foundation for our own. We then turn to the results of our literature search.

Soares et al. (2019) described eight law enforcement education programs developed in Canada, the United States and the United Kingdom pertaining to mental or behavioral health. Of these, half were autism-specific education programs and half were general trainings regarding mental or behavioral health. Soares et al. stated that although the mental health training programs (e.g. CIT) have published evaluative data regarding training effectiveness, the autism-specific programs have not. To gather what data exist regarding the efficacy of autism-specific training for LEOs, Railey et al. (2020) conducted a systematic review. Across 13 databases and 28 journals, without date range or geographical restrictions, only two articles met the inclusion criteria (i.e. LEOs participated in the training, training topics related to autism and the study used an experimental design). The two studies were Murphy et al. (2017) and Teagardin et al. (2012), neither of which resulted from our literature search due to geographic (i.e. United States) and date range (i.e. after 2018) filters.

In Murphy et al.’s (2017) study, 11 officers of Ireland’s National Police Service participated in a 90-min training. A five-item survey was distributed prior to and at the conclusion of the training. Post-training, increases were observed in self-reported understanding of autism and the ability to interact and communicate with individuals with autism. In contrast, Teagardin et al. (2012) used a greater number of participants (i.e. 42 officers), but the training duration was shorter (i.e. 13-min training video). A 12-item nonstandardized pre- and post-test survey was administered, which included 10 questions pertaining to autism knowledge and two questions pertaining to confidence when interacting with individuals with autism. Officers demonstrated gains in autism knowledge from pre-to post-test, along with increases in self-reported confidence. Additionally, officers who had received the training outperformed the officers in the control group, who had not watched the 13-min video.

In 2022, Sreckovic et al. sought to build on Railey et al.’s (2020) work with a scoping review of all studies on autism-specific training for LEOs. A search of nine databases for peer-reviewed articles (any year of publication) yielded 655 articles, of which only five met inclusion criteria (i.e. training was conducted as part of the study, and participants were LEOs; Sreckovic et al., 2022). Two of the five articles were the same as those identified by Railey et al. (2020): Murphy et al. (2017) and Teagardin et al. (2012). The remaining three articles were from Gardner and Campbell (2020), Love et al. (2020a), and Hinkle and Lerman (2021), each of which was likewise found in our literature search. In Gardner and Campbell’s (2020) study, 157 LEOs participated in a 4-h autism-specific training presentation, followed by an optional 4-h role-play simulation. Post-test responses indicated significant increases in autism knowledge and self-confidence in responding to a call with an individual with autism. Love et al. (2020a) measured the efficacy of the training Emergency Network Autism Community Training, a 2-h training with lecture, discussion and video clips provided to 224 first responders. Post-training results demonstrated increased autism knowledge and self-rated confidence working with individuals with autism. Lastly, in Hinkle and Lerman’s (2021) study, training was provided to 24 cadets and three LEOs. Participants were split into groups: those receiving a lecture only and those receiving a lecture plus role-play sessions. Participants in both groups demonstrated statistically significant increases in the ability to promote compliance using the steps outlined in the lecture. Participants who had received the hands-on sessions outperformed participants who had received the lecture-only format.

Following the work of Sreckovic et al. (2022) by 2 years, our literature search found an additional two articles from Kent and Hughes (2022) and Rascon (2022), which had been published in the United States (see Table 1). In Kent and Hughes’s study, 33 officers participated in autism-specific de-escalation training in 1 of 3 ways: watching a video, participating in a virtual reality exercise, or both watching the video and participating in the virtual reality exercise. Although significant differences in training effectiveness were not found among the different training methods, Kent and Hughes found that using virtual reality or similar activities, which can be tailored to reflect the city or context in which the officer works, may increase officer buy-in. Last, Rascon detailed a 75-min training using a “Communication Complex” approach, which emphasized the integrated nature of communication, biological state and perception of safety. The training involved a video, a lecture-based presentation and reflection opportunities. Post-survey responses indicated most participants found the training to be both informative and valuable, with 98% of participants recommending attendance for others in their department (Rascon, 2022).

As outlined in Table 1, studies on the efficacy of autism-specific training for LEOs are limited in number, which generally include small sample sizes and use non-standardized surveys; however, these studies demonstrated positive effects of training in relation to officer-reported autism knowledge and interaction confidence.

The development of INTERAC was framed by social cognitive theory and the studies outlined in Table 1. The eight studies identified in our current search were then examined for the specific training program elements that have evidence of efficacy to guide INTERAC program development (see Table 2). These eight studies provided guidance for the training format, who should do the training and what the training content should include. First, the most desirable format for training was in person. Both Diamond and Hogue (2021) and Gardner and Campbell (2020) recommended in-person training over other modalities (i.e. video-based or online training modules). Second, the recommended training approach included collaborative team teaching with a disability specialist and police trainer (Diamond and Hogue, 2021; Gardner and Campbell, 2020) and members of the autism community (Love et al., 2022). Furthermore, training was recommended to be dynamic, interactive, and responsive to the needs and questions of the training participants (Hinkle and Lerman, 2021; Love et al., 2022; Railey et al., 2019). Done in this way, in vivo (in-person) training was considered to have greater validity and to provide greater opportunity for mastery of content compared to video-based or online modalities (Gardner and Campbell, 2020). Additionally, studies recommended that disability awareness training be mandatory for all officers (Love et al., 2022; Railey et al., 2019).

As far as content for training from the literature, the training should provide national, state, and local prevalence data and actionable strategies and resources (Diamond and Hogue, 2021; Gardner and Campbell, 2020; Hinkle and Lerman, 2021; Love et al., 2022; Railey et al., 2019; Rascon, 2022; Soares et al., 2019; Wallace et al., 2020). Content should include a discussion of disability categories, characteristics of autism, and associated behaviors to differentiate neurodevelopmental disability from mental health crisis or drug-induced behavior (Diamond and Hogue, 2021; Gardner and Campbell, 2020; Hinkle and Lerman, 2021; Love et al., 2022; Railey et al., 2019; Wallace et al., 2020). Though providing this information is essential to increasing knowledge and awareness of autism, training should make clear that officers are not responsible for diagnosing the person with whom they interact (Love et al., 2022; Rascon, 2022). The use of body camera footage and case studies provides an opportunity for further discussion of characteristics of disability, strategy usage, positive and negative outcomes of an interaction, and reflection upon additional ways in which an officer may respond (Diamond and Hogue, 2021; Railey et al., 2019).

Using the systematic literature review, the MDT developed a draft training on best practices and components from the literature (see Table 2). Permission was formally requested and obtained for body camera footage incorporated into the presentation. After an initial round of training development, members of the children’s hospital and university partnership met with a local police lieutenant via Zoom for review of the presentation and feedback. The presentation was revised iteratively after this meeting and two subsequent meetings. After gaining the lieutenant’s approval, an executive summary was sent to the police chief, and a meeting was scheduled. After reviewing the training, the police chief provided approval, and training dates were scheduled with the department’s training director. At this point, the training (about 2 h in length) was consolidated to meet the 40-min timeline of the shift briefing. The final approved training included a didactic presentation (about 13 min), a discussion of body camera footage and scenarios (about 15 min), presentation of resources (about 7 min), and an opportunity to ask questions (about 5 min).

Ultimately, INTERAC included the following elements: (1) The training was designed to be mandatory for all officers and conducted in person during department briefing. (2) The training was conducted via team teaching. While an autism specialist and officer from the MDT co-presented, co-presentation by MDT members with autism was not feasible during this pilot implementation. (3) INTERAC included the following content; disability categories (i.e. what disorders are included under the umbrella of neurodevelopmental disorders; e.g. ASD, intellectual or developmental disorder, attention-deficit/hyperactivity disorder), prevalence data (i.e. data on disability prevalence, neurodevelopmental disability prevalence, and autism prevalence in the United States and in California), the definition and diagnostic criteria of autism from the American Psychiatric Association, characteristics of autism and potentially misinterpreted behaviors (e.g. reduced eye contact, elopement, echolalia), strategies for working with individuals with autism (i.e. demonstration of augmentative and alternative communication tools: voice output applications and laminated visual boards), case study examples of LEOs interacting with individuals with autism drawn from real experience, body-worn camera footage with interactions between officers and individuals with autism, clarification that the first responder’s role is not to diagnosis, emphasis on empathy/compassion and resources to support individuals with autism (e.g. contact information for local service organizations, neighborhood alert letter templates, sensory bags with communication tools, noise cancelling headphones and sunglasses). (4) INTERAC was measured using pre- and post-test data utilizing the PSEA scale.

While Soares et al. (2019) described a lack of evaluative data for autism-specific training, Leydier (2022) stated Love et al.’s (2020b) PSEA scale can be a “potential starting point” (para. 3). As of February 2025, no studies were found using the PSEA, making this study the first of our knowledge to use this scale after its initial validation. Additionally, this project implements and examines the efficacy of the INTERAC training. For this study, we focused on the research questions:

1. What personal and professional experience do LEOs within a large department in California have specific to autism?

2. Does training regarding developmental disabilities, specifically autism, have a positive effect on officers’ reported self-efficacy scores?

   • •

     Is there a difference between pre- and post-test PSEA scores on each item for LEOs who participated in the INTERAC training?

   • •

     Is there a difference between pre- and post-test PSEA scores on created subscales (receptive knowledge, interaction with a third party, ground setting and application) for LEOs who participated in the INTERAC training?

INTERAC was presented to officers and detectives of a large police department in an urban/suburban county in California during seven shift briefings across a 6-day period. To measure the relationship between training and self-efficacy scores, participants were invited to complete a brief demographic survey (items informed by Christiansen et al., 2021; Gardner et al., 2018; Gardner et al., 2022; Gardner and Campbell, 2020; Love et al., 2020b; Wallace et al., 2022) and the 13-item PSEA scale developed by Love et al. (2020b) prior to and at the conclusion of the training. The PSEA scale asks participants to rate self-confidence along a four-point scale (1 = I cannot do that, 2 = I doubt I can do that, 3 = I’m fairly certain I can do that, 4 = I can do that). Statements included the ability to identify signs of autism and explain characteristics, distinguish autism from other conditions, establish rapport and adapt communication, gather information from a person with autism or their caregiver, find a missing person and de-escalate a situation involving a person with autism.

The PSEA draws upon Bandura’s (1989) concept of self-efficacy and was developed using multiple iterations of (1) expert review and focus groups with 28 experts (i.e. veteran officers with professional experience working with individuals with autism, family members of individuals with autism, and specialists with knowledge of self-efficacy measurement); (2) cognitive interviews to verify interpretation of items as intended (three officers); and (3) two phases of survey administration (i.e. 182 officers in Phase 1 followed by refinement of formatting and survey items, and 438 officers in Phase 2). Love et al. (2020b) explored the psychometric properties of the 48-item Autism Stigma and Knowledge Questionnaire (ASK-Q) and the PSEA. Results from the dimensionality assessments indicated that items on both the ASK-Q and the PSEA reflected unidimensional constructs. Additionally, a significant positive relationship was found between autism knowledge and self-efficacy (Love et al., 2020b).

After obtaining author permission via email to use the PSEA, the PSEA items were entered into Qualtrics. A link and QR code were generated and sent to officers via departmental email three days prior to the start of the first shift briefing. This link and QR code were additionally displayed as officers entered the room on the day of their shift briefing. The last presentation slide contained the link and QR code to the post-survey. In alignment with institutional review board approval, participation in the survey was optional and anonymous.

The INTERAC training was offered seven times, during a morning and/or evening shift briefing, across a 6-day period in Summer 2023. The pilot study included a convenience sample of officers who had attended one shift briefing during the window in which the INTERAC training was provided. No officers attended more than once. No restriction in selection criteria was made regarding educational or professional training, number of years in the profession, or personal or professional experience with autism (see Table 3).

Data were transferred from Qualtrics to IBM SPSS for Statistics 28. All respondents with incomplete surveys were removed. We began with descriptive analysis, looking at demographic characteristics and comparing survey item responses pre- and post-training (see Tables 3 and 4). Next, we conducted an independent sample Mann–Whitney U to assess differences in each of the 13 PSEA items. An independent sample Mann–Whitney is a nonparametric test used with ordinal data and data that do not meet the assumptions of homogeneity or normality (Field et al., 2012). Next, we developed scales, tested those scales, and subsequently conducted a two-tailed t-test. To develop scales, we first looked for commonalities among survey items and grouped them such that items pertaining to receptive knowledge (i.e. identifying and distinguishing signs of autism); interaction with a third party (i.e. explaining characteristics to another officer or a concerned bystander and gathering information from a caregiver); ground setting (gathering information from an individual who does not use spoken communication, establishing rapport, and modifying communication and the environment) and application (finding a missing person or de-escalating a situation) were grouped together. We tested these scales for reliability using Cronbach’s alpha. With an acceptable Cronbach’s alpha exceeding 0.7 (Creswell and Creswell, 2018), we examined the assumptions of homogeneity via Levene’s test and normality with a Shapiro–Wilk test (see Table 5). Although the assumption of normality was not met (p < 0.05), we continued with a two-tailed t-test because t tests are considered robust to violations of normality, skewness did not exceed +1.0 or fall below −1.0 and visual inspection of QQ plots demonstrated approximate normality (Leech et al., 2015).

From approximately 170 training participants, 57 pre-test and 55 post-test responses were received, with a response rate of about 30%. Incomplete pre- and post-test submittals were removed. The following analysis includes 50 pre-test and 42 post-test submittals. Findings regarding officers’ personal and professional experiences specific to autism and whether autism-specific training had a positive effect on their reported self-efficacy scores are reviewed.

Research Question 1 focused on the personal and professional experience of LEOs. Most respondents reported serving in the profession for more than 5 years (see Table 3). Although about half the officers reported they did not personally know someone with autism, 26% of post-test respondents reported personally knowing one person, 14% reported knowing two people and 5% reported knowing three or more people with autism. Of respondents who reported a personal relationship with at least one person with autism on the post-survey, 17% indicated the individual(s) they knew were extended family, 12% indicated the individual(s) they knew were immediate family and 10% indicated the individual(s) they knew were friends. Seventy percent of officers indicated they had previously received training specific to autism, though in speaking to one officer, responses may have referred to a 10-min briefing he gave months prior in which he notified officers of newly distributed sensory kits in department vehicles. In contrast, 12% of officers responded they had not received training specific to autism, and 18% were unsure. In a professional capacity, 21% of officers in the post-training survey estimated interacting with an individual with autism once a year, 38% estimated once a month and 12% estimated once a week. Nearly a quarter of officers indicated being unsure about their frequency of interactions with individuals with autism (see Table 3). These interaction rates were higher than those reported by Gardner et al. (2022). In contrast to the 35% of officers who reported responding to 6 or more calls per year in Gardner et al.’s study, approximately 52% of officers in our study reported responding to more than six calls per year (totaling across response categories once per month, once per week and once per day).

In response to Research Question 2a, descriptive results indicated a positive relationship between the brief, autism-specific training and officers’ reported self-efficacy scores. From pre-to post-training results, the frequency of respondents answering “I can do that” (answer of “4” on the 4-point scale) increased on all items of the PSEA. The smallest change (i.e. 4.5% growth) was observed on the ability to identify signs of autism, while the greatest change (i.e. 28.0% growth) was observed on the ability to modify the environment as needed. A positive change was likewise observed in the mean response pre- to post-test for all items (see Table 4). An independent sample Mann–Whitney U-test revealed these increases in pre- and post-test scores were significant (p < 0.05) on 7 of the 13 PSEA items: the perceived ability to de-escalate a situation in which a person with autism is harming themself, explain at least three characteristics of autism to another first responder who asks, distinguish between autism from other disabilities, use knowledge of autism to help find an individual with autism who has become a missing person, gather identifying information from someone with autism who does not use spoken language, inform a concerned citizen about autism and modify the environment to help an individual with autism feel calm (see Table 4). Additionally, officers reported the INTERAC training was clear (mean of 4.92), helpful (mean of 4.80) and increased comfort in interacting with individuals with autism (mean of 4.75) on a 5-point scale (1 = disagree, 5 = agree).

To answer Research Question 2b, the 13 items of the PSEA were grouped into four subscales of related target skills: receptive knowledge, interaction with a third party, ground setting and application. Items in these subscales were tested for reliability using Cronbach’s alpha, and results were as follows: receptive knowledge (⍺ = 0.807), interaction with a third party (⍺ = 0.776), ground setting (⍺ = 0.839) and application (⍺ = 0.798). All values exceeded 0.7, indicating that survey items hung well together in each scaled construct (Creswell and Creswell, 2018). To examine the assumption of homogeneity of variance across groups, Levene’s test for equality of variances (Field et al., 2012) was conducted. Homogeneity of variance was met across groups, p > 0.05 (see Table 5). Visual inspection of QQ plots demonstrated approximate normality. The mean for each scale increased from pre-to post-test (see Table 5). Results of the independent samples t-test showed a significant difference in pre- and post-training means in each of the four subscales (see Table 5): receptive knowledge (p = 0.047), third party (p = 0.004), ground setting (p = 0.012) and application (p = 0.002).

Past research has demonstrated an association between knowledge of autism and increased confidence in identifying characteristics of autism, effectively interacting with individuals with autism and feeling prepared for a call involving individuals with autism (Copenhaver et al., 2020; Gardner et al., 2018; Love et al., 2020a, b). Love et al. (2020b) demonstrated alignment of the PSEA with the ASK-Q and found a significant positive relationship between autism knowledge and self-efficacy. The findings of this project showed increases in self-efficacy following autism-specific training. Of note, post-training officers reported greater self-efficacy in distinguishing autism from other disabilities; informing a first responder or concerned citizen about autism; de-escalating a situation in which a person with autism was harming themself; and using knowledge of autism to find a missing person, gather information and modify the environment as needed.

Previous studies have noted a tendency to associate autism with mental illness (Copenhaver et al., 2020; Herbert et al., 2022), but doing so can lead to negative outcomes (Gardner and Campbell, 2020). Although officers trained in CIT are more likely to use physical force in calls involving individuals with autism compared to officers without CIT training (Gardner and Campbell, 2020), officers who participated in INTERAC demonstrated statistically significant increases in perceived ability to de-escalate a situation in which a person with autism was self-harming, along with perceived ability to modify the environment to help an individual with autism feel calm. In recognition that self-efficacy is impacted by situational demands that are both domain-specific (i.e. law enforcement) and population-specific (i.e. autism), as highlighted by Love et al. (2019), this study lends support for autism-specific training.

The rising prevalence of autism in California and the passage of SB 882 made it a unique area to explore. We acknowledge that similar training may be relevant to other parts of the world and advise consideration of social communication within the specific socio-cultural context. INTERAC was designed at the request of a local police department, and therefore, we faced certain limitations such as duration of training, training schedule, who participated and when. Due to concerns with anonymity with the Institutional Review Board, we are unable to confidently match our pre- and post-test. We have a reasonable assumption that the participants are largely the same, but there are differences, as seen in Table 3. This is a limitation of this current pilot study. Furthermore, although this pilot effort showed positive results, we recommend further testing of the training and PSEA.

With recent approval from the county and certification through the Commission on Peace Officer Standards and Training, we have begun to expand INTERAC to a 3-h training containing each of the recommended training components (see Table 2) with co-presentation by members of the autism community. A comparison of the 40-min INTERAC with the 3-h version may provide insight into the effect of increased training length and inclusion of all recommended training components on self-efficacy scores. Whereas the current post-training measure asked participants whether they believed the training was helpful (mean of 4.8 out of 5.0), future survey items may wish to ask whether participants believe the training to be applicable to practice.

Last, there is a need to study the behavioral impacts and long-term effects of training (Love et al., 2020a; Railey et al., 2019; Sreckovic et al., 2022). SB882, requiring officers to document perception of developmental disability in some instances, is an important step in understanding interactions between officers and individuals with developmental disabilities in the field. It may be several years, however, before this data is available. Additionally, documentation of a perceived developmental disability relies upon officer ability to recognize characteristics of developmental disabilities. Within this study, officers’ self-efficacy in distinguishing autism improved post-training. Therefore, we believe autism-specific training can support documentation efforts within SB882.

The likelihood that LEOs will interact with individuals with autism in a professional capacity is great. Though previous legislation in California grouped training on mental health and developmental disabilities together, it is important to note that developmental disabilities such as autism are not mental health disorders or crisis states. In response to SB 882, INTERAC was developed through a children’s hospital, community and university partnership that included a local police lieutenant, school psychologist, speech pathologist, applied behavior analyst and members of the autism community. Few studies have researched the efficacy of autism-specific training for LEOs, even fewer of which have used a measurement tool with established validity and reliability. The present study used the PSEA scale coupled with brief demographic items pre- and post-training to answer the research questions. Most of the study respondents reported serving in the profession for more than 5 years, and just under half the officers reported personally knowing someone with autism. Of note, rates of interaction between officers in a professional capacity and individuals with autism appear higher than indicated in previous research. Pre- and post-training responses on the PSEA indicated a positive relationship between the brief, autism-specific training session and officers’ reported self-efficacy scores. A positive change was observed in the mean response of all pre- and post-test PSEA items. Additionally, results of the independent samples t-test showed a significant difference in pre- and post-training means in each of the four subscales. Though the training was short in length, the findings of this study may support the feasibility of autism-specific training implementation where scheduling otherwise poses a challenge.