A Survey of Factors That Influence Affinity With Dogs for People With Autism

ABSTRACT

Affinity with dogs has been defined as the degree to which people value interactions with dogs; it may be influenced by individual preferences, experiences with animals, and characteristics of the dog. To date, however, there has been limited investigation regarding factors that influence dog affinity for people with autism. An anonymous online survey was used in this study; it included demographic questions; exploration of experiences with animals; attitudes, preferences, likes, and dislikes of dogs; and the perceived impact of dogs on the person with autism’s social and community participation. The survey was distributed via social media channels, assistance dog organizations, and a disability provider to people living in Australia (n = 131) who were (1) an adult with a diagnosis of autism or (2) a next-of-kin of a person of any age with autism (completing the survey on behalf of their family member). Data were analyzed using descriptive and inferential statistics and content analysis. There were no significant differences in levels of affinity between children and adults with autism. However, findings indicated significant differences in levels of affinity between participants with and without a dog. Text entry responses to questions provided additional insights into participants’ attitudes and sensory preferences toward dogs, suggesting that affinity with dogs can be influenced by personal, disability-related, environmental, and animal-related factors. An awareness and consideration of these factors is essential for individuals with autism, their families, and service providers – including animal-assisted therapy or assistance animal organizations – when considering individual participation goals and determining if animal supports are the most appropriate option. Without an awareness and understanding of these factors, this could limit the effectiveness of the human–animal interaction.

KEYWORDS:

• Affinity
• autism
• dogs
• human–animal bond
• human–animal interaction

Autism spectrum disorder, also known as autism, is a lifelong neurodevelopmental condition characterized by challenges in cognition, communication, and behavioral performance (AIHW, 2017; American Psychiatric Association, 2013). People with autism¹ can also experience heightened sensory preferences and aversions to light, sound, taste, touch, and smell (Marco et al., 2011). Experiences of autism are, however, unique to each individual, and as such, tailored approaches are required to support participation in home- and community-based activities (AIHW, 2017; Vasilevska Petrovska et al., 2019). For some people, this may include support from staff and/or other key supporters (e.g., family, caregivers) and animal supports. The supports provided by animals can be across a continuum of pet animals, through to animal-assisted therapies and assistance animals (Lisk et al., 2021). Pet animals have no formalized training – they are kept primarily for companionship and have been defined as animals that people share their domestic lives and spaces with (Young & Carr, 2018). Assistance animals may support people with autism in skill development, communication, and community participation (Assistance Dogs Australia, 2021), and therapy animals can be used in goal-directed therapy interventions (Hill et al., 2019, 2020; London et al., 2020).

The term affinity has been used to describe “the degree to which people value interactions with animals” (Bryan et al., 2014, p. 24). Affinity with dogs for people with autism is of particular interest as dogs are thought to “present a powerful multisensory stimulus” (Redefer & Goodman, 1989, p. 461). For example, a case study by Solomon (2010) described a child with autism who typically disliked loud noises but was accepting of a therapy dog barking when she commanded it to “speak.” In addition, high levels of attachment have been reported for children with autism and pet, assistance, and therapy dogs (Carlisle, 2015; Grandgeorge et al., 2012; Hill et al., 2020; Prothmann et al., 2009). It is proposed that people with autism relate to dogs more readily than humans because both share “sensory-based thinking” (Grandin et al., 2019, p. 229), and dogs may communicate in a way that is more easily understood by some individuals with autism than human communication (Prothmann et al., 2009).

To date, most of the research on human–animal interactions and affinity has focused on children with autism, and some consistent benefits have been identified for pet, assistance, and therapy dogs. These include the ability of dogs to facilitate emotional regulation and reduce feelings of stress (Bystrom & Lundqvist Persson, 2015; Carlisle, 2014; Ward et al., 2017); enhance community participation (Burrows et al., 2008; Hall et al., 2016; London et al., 2020); and result in gains in communication and behavioral skills (Ben-Itzchak & Zachor, 2021; Berry et al., 2013; London et al., 2020; Martin & Farnum, 2002; O'Haire, 2013; Sams et al., 2006). In addition, the presence of a dog in therapy has been shown to facilitate children with autism’s motivation to engage in the therapeutic process (London et al., 2020). These benefits are thought to be due to the non-judgmental (London et al., 2020) and accepting nature of dogs (Barnhart et al., 2023) and their ability to increase a person’s sense of safety (Burgoyne et al., 2014; Burrows et al., 2008; Solomon, 2010).

Recent research has, however, challenged the general assumption that children with autism have a natural affinity with dogs, highlighting the procedures for considering and assessing preference for human–animal interactions are important but at times overlooked (Clay et al., 2022). Attitudes toward animals can be influenced by personal experience and familiarity (Guerin et al., 2017; Kellert, 1985). Within animal-assisted therapy, some features of the dog, such as smell, size, and the presence of claws and teeth, can impact the therapeutic process (Hill et al., 2020; London et al., 2020). In other instances, children with autism show fear of dogs used in animal-assisted interventions (Barnhart et al., 2023). Carlisle et al. (2018) found that among families of children with autism who owned a pet dog, some children were unable to connect with the dog and caregivers expressed a range of concerns that may impact dog affinity. These included dogs moving too quickly, barking, licking, or jumping on the child, the size of the dog, and the animal exhibiting unpredictable and excitable behaviors that could upset the child. In some cases, children with autism have harmed the dog (Harwood et al., 2019). Similarly, Burrows et al. (2008) observed that some autism assistance dogs receive negative attention and maltreatment from children.

Affinity with dogs, or lack thereof, is likely to impact behaviors, skill development, and social participation of people with autism in the home, community, and therapy settings. Research highlights the potential for conflict between the sensory preferences of the person with autism and the sensory input offered by a dog (Carlisle et al., 2018; Hill et al., 2020; London et al., 2020). This information is important when considering the range of supports that may be suitable for a person with autism – including animal supports. However, further research is needed that includes adults with autism. Additionally, Hall et al. (2016) and Carlisle (2014) suggest that future studies should explore other factors, such as how the characteristics of the dog are important when considering a “match” for the person with autism and their family.

In undertaking this research, we endeavored to address this evidence gap: we aimed to explore affinity with dogs for both children and adults with autism. This information holds relevance for individuals with autism, their families, and service providers, including animal-assisted therapy or assistance animal organizations when considering individual participation goals, and planning associated supports, within clinical or community settings.

Methods

Design

The data reported in this paper are drawn from an online anonymous survey, administered using Qualtrics (https://www.qualtrics.com). Survey responses were collected between February and May 2021 from people with autism and their next-of-kin. Ethics approval was obtained from Monash University Human Research Ethics Committee (ID No. 26914) prior to research commencing.

Participants

Non-probability convenience sampling methods were used to recruit participants. The anonymous survey link was distributed through two Australian dog organizations that provide trained assistance dogs to people with autism: a disability organization offering home and community services to people with autism, and the social media platforms of the researchers (e.g., Facebook, LinkedIn, and Twitter). Eligibility criteria included the person living in Australia, having a diagnosis of autism, and being aged 18 years or over. However, if the person with autism could not give informed consent, and/or was aged 17 years or under, a next of kin (aged 18 years or over) could complete the survey on their behalf.

Instrument

An anonymous survey was designed based on: (1) a review of existing literature that reports sensory experiences (e.g., Carlisle, 2014; Carlisle et al., 2018; Harwood et al., 2019; London et al., 2020) and altered behavior (Hall et al., 2016) associated with dog interactions; (2) a review of existing measures of the human–animal bond via databases and a compendium of measures (Anderson, 2007); and (3) expert opinion from assistance-dog organizations and members of the research team with expertise on the use of animal supports for people with disabilities living in the community.

The survey was pre-tested by representatives from two assistance dog organizations and two personal contacts of the first author: a person with autism, and a next-of-kin of a person with autism. Only minor modifications (e.g., clarity of questions) were made to the survey based on this review. This pre-testing also helped provide an estimate of completion time, which ranged from 10 to 15 min. The final version of the survey was divided into four sections and consisted of 42 closed questions (including six with “other” category that allowed for open-ended responses) and two optional open-ended questions (see online supplemental file). Section one covered demographic information, including questions on the state or territory the person was located in, if they (or their next of kin) had autism, and the respondent’s age (to check eligibility). The online survey was closed if the person did not meet the eligibility criteria.

Section two focused on experiences with animals. Questions included pet-ownership status, type of pets owned (currently and previously), current and previous engagement with assistance animals and therapy animals, and the type of animal the person engaged with. Section three measured affinity with dogs and comprised 27 items (scored out of 135) that were rated on a 5-point Likert scale (anchored by Strongly Disagree (1) and Strongly Agree (5)). There were six items from the Coleman Dog Attitude Scale (C-DAS; Coleman et al., 2016), which were used with written permission from the C-DAS authors and deemed most relevant to the research questions (I love dogs; dogs comfort me; I avoid dogs; I talk to dogs; I think dogs are fun; I hate dogs); six items relating to touch (e.g., Patting a dog helps me relax), 11 items relating to sight (e.g., I get nervous when I see a dog running toward me), and four items relating to hearing and smell (e.g., I do not like it when a dog barks suddenly; I move away from a dog if it smells). It was not possible to assess test-retest agreement during this study, but the reliabilities of the summated scores were assessed using Cronbach’s alpha. Finally, section four of the survey included two optional open-ended questions for participants to provide comments about anything else they liked or disliked about dogs.

The survey was administered online and at the start of it participants were provided with a project explanatory statement. Participants were asked to tick a checkbox to indicate consent or decline participation. Those who consented were directed to the beginning of the survey, whereas the online survey was closed if the person declined consent.

Data Analysis

Data from the online survey were exported from Qualtrics into SPSS 28 statistical software, and each participant was given a unique code (P1–P170). Of the 170 responses received, 131 (77%) met the inclusion criteria and completed 100% of the survey and thus were included in the data analysis. Quantitative data were analyzed using descriptive and inferential statistics.

For each participant, a total score for the 27 items in part three of the survey was calculated by reverse scoring negatively phrased items. Potential scores could range from 27 to 135, with a higher score indicating a higher affinity with dogs. Cronbach’s alpha indicated excellent reliability (α = 0.94). In addition, sub-scores were calculated for the six items from the C-DAS (scored out of 30). Although not specifically designed for people with autism, the C-DAS has demonstrated the ability to predict behavioral intentions of dog interactions (Coleman et al., 2016). The Cronbach’s alpha indicated excellent reliability (α = 0.911), which was comparable to the reliability obtained by Coleman et al. (2016) for all 24 items (α = 0.98). Finally, sub-scores were calculated for the items relating to sensory experiences associated with dog interactions, with a maximum score of 20 for hearing and smell, 30 for touch, and 55 for sight. The Cronbach’s alpha indicated acceptable reliability (α = 0.7) for the items relating to hearing and smell, and good reliability for the items relating to touch (α = 0.808) and sight (α = 0.859). There were also high and significant correlations between the total score and sub-scores measuring attitudes toward dogs (r_s = 0.8, n = 131, p = 0.001) and the sensory preferences of hearing and smell (r_s = 0.7, n = 131, p = 0.001), touch (r_s = 0.871, n = 131, p = 0.001), and sight (r_s = 0.935, n = 131, p = 0.001). According to Cohen’s (1988) guidelines for interpreting coefficients (see Dunst & Hamby, 2012; Lipsey & Wilson, 2001), these are all large effect sizes.

For some analyses, participants were categorized into two age groups: children (aged 17 years and younger, n = 63); and adults (18 years and older, n = 84). Participants were also categorized according to their experiences with dogs. As shown in Table 1, of the participant sample who responded to the survey, only 16 (12%) had never owned a pet dog or worked with an assistance or therapy dog. It was therefore only possible to categorize participants by those currently with a pet, assistance, or therapy dog (D) (n = 94) and those without a dog (ND) (n = 37). Depending on level of measurement and whether other parametric assumptions were met, differences between participant groups were explored using t-tests, Mann–Whitney U, chi-square, and Fisher exact tests. To reduce the risk of Type II errors, Bonferroni adjustments were used, with significance reported at p < 0.002 (α_altered = 0.05/25). Cohen’s (1988) guidelines were used to interpret and report effect sizes (r), which were calculated using the methodology of Fritz et al. (2012): converting z scores into r using the formula r = the absolute value of z ÷ $\sqrt{N}$.

Table 1. Ownership of and/or work with a pet, assistance, or therapy dog for participants currently with a dog (D) and without a dog (ND).

	D (n = 94)	ND (n = 37)
	n (%)	n (%)
Currently have/work with
Pet dog	81 (87%)	0 (0%)
Assistance dog	27 (29%)	0 (0%)
Therapy dog	20 (21%)	0 (0%)
Previously had/worked with
Pet dog	26 (28%)	15 (41%)
Assistance dog	27 (28%)	1 (0.03%)
Therapy dog	35 (37%)	5 (14%)
Never had a pet dog or worked with an assistance or therapy dog	0 (0%)	16 (12%)

There was a total of 85 responses (n = 59 currently with a dog) for the two optional open-ended questions regarding what the person with autism liked or disliked about dogs. These responses were exported into Microsoft Excel. To ensure validity, two researchers (the 1st and 2nd authors) separately performed the content analysis following the abstraction process outlined by Elo and Kyngäs (2008). This involved reading the responses carefully and open coding, followed by combining codes with other codes to form categories or sub-themes, and the organization of analyzed content into overall themes. These themes were checked through consensus work between the two researchers, with a final set of codes and themes agreed upon (Creswell, 2012; Elo et al., 2014).

Results

Participant Characteristics

Participants were 54 (41.2%) adults (aged 18–51 years, M = 28, SD = 10) with autism who provided their own consent and 77 (58.8%) next-of-kin who completed the survey on behalf of their family member with autism (aged 2–62 years, M = 13, SD = 9). From here on, “participants” refers to people with autism as described by either themselves or their next-of-kin. Responses were received from all states and territories of Australia, with the majority (n = 81, 62%) from Victoria. Table 2 shows the characteristics of participants with a dog (D) and without a dog (ND). There were no significant differences in age, gender, or other disabilities experienced in addition to autism.

Table 2. Characteristics of participants currently with (D) and without a dog (ND).

	D (n = 94)	ND (n = 37)	p
	M (range)	M (range)
Age (years)	19 (6–62)	22 (2–50)	0.098
	n (%)	n (%)
Gender			0.776
Male	52 (55%)	23 (62%)
Female	36 (38%)	12 (32%)
Non-binary/gender diverse	6 (6%)	2 (5%)
Other disabilities in addition to autism
Yes	71 (76%)	21 (57%)	0.034
Sensory and speech	24 (26%)	8 (22%)
Intellectual disability	17 (18%)	7 (19%)
Psychosocial disability	14 (15%)	4 (11%)
Neurological disability	10 (11%)	1 (3%)
Physical disability	8 (9%)	3 (8%)
Acquired Brain Injury	2 (2%)	2 (5%)
Hearing impairment	1 (1%)	2 (5%)
Vision impairment	2 (2%)	1 (3%)
Cerebral Palsy	2 (2%)	1 (3%)
Down Syndrome	2 (2%)	0 (0%)
Multiple sclerosis	1 (1%)	0 (0%)
Stroke	0 (0%)	2 (5%)

Note: To reduce the risk of Type II errors, Bonferroni adjustments were used; significance reported at p < 0.002.

Affinity with Dogs

Presented in Table 3 are the scores by age group for the 27 items measuring overall affinity with dogs and the sub-scores for the items measuring attitudes and sensory preferences toward dogs. There were no significant differences in scores between children and adults with autism. Nevertheless, significant differences were shown between participants with a dog (D) and without a dog (ND) for the 27 items measuring overall affinity and each of the sub-scores (see Table 4). The effect sizes for these differences were all medium (range 0.31 to 0.35). As shown in Table 5, over eight in ten participants with a dog agreed or strongly agreed with the six items measuring attitudes toward dogs, compared with approximately 6 in 10 participants without a dog. Items measuring sensory preferences toward dogs were also more favorable for participants with a dog.

Table 3. Affinity with dogs – overall and sub-scores (mean and range) by age group.

	Children (n = 63)	Adults (n = 68)	p
	M (range)	M (range)
Total score (out of 135) across 27 items	99	98	0.503
	(32–127)	(51–128)
Sub-score (out of 30): Attitudes toward dogs	25	26	0.536
	(6–30)	(11–30)
Sub-score (out of 30): Touch	21	20	0.350
	(6–29)	(9–30)
Sub-score (out of 66): Sight	37	36	0.539
	(14–49)	(19–46)
Sub-score (out of 24): Hearing/smell	13	13	0.718
	(4–20)	(6–18)

Table 4. Affinity with dogs – overall and sub-scores (mean and range) for participants currently with (D) and without a dog (ND).

	D (n = 94)	ND (n = 37)	p
	M (range)	M (range)
Total score (out of 135) across 27 items	104	98	z = –4.041
	(58–128)	(32–125)	p = 0.0001
Sub-score (out of 30): Attitudes toward dogs	27	23	z = –3.645
	(12–30)	(6–30)	p = 0.0001
Sub-score (out of 30): Touch	22	18	z = –3.720
	(11–30)	(6–29)	p = 0.0001
Sub-score (out of 66): Sight	38	32	z = –3.989
	(23–49)	(14–47)	p = 0.0001
Sub-score (out of 24): Hearing/smell	13	11	z = –2.875
	(5–20)	(4–17)	p = 0.0001

Table 5. Agreement across the 27 items for participants currently with (D) and without a dog (ND).

	D (n = 94)		ND (n = 37)
	Disagree/Strongly Disagree	Agree/Strongly Agree	Disagree/Strongly Disagree	Agree/Strongly Agree
	n (%)	n (%)	n (%)	n (%)
Attitudes toward dogs
Dogs comfort me	6 (6%)	83 (88%)	7 (19%)	24 (65%)
I avoid dogs*	80 (85%)	8 (9%)	23 (62%)	11 (30%)
I love dogs	4 (4%)	82 (87%)	6 (16%)	25 (68%)
I like talking to dogs	9 (10%)	80 (85%)	4 (11%)	27 (73%)
I hate dogs*	90 (96%)	1 (1%)	30 (81%)	5 (14%)
I think dogs are fun	3 (3%)	83 (88%)	4 (11%)	27 (73%)
Sensory preferences: Touch
A dog pressing against me or sitting on me makes me happy	3 (3%)	91 (97%)	3 (8%)	25 (68%)
Patting a dog helps me relax	2 (2%)	86 (91%)	3 (8%)	24 (65%)
I do not like getting dog saliva on myself*	36 (38%)	43 (46%)	5 (14%)	22 (59%)
I do not mind patting a dog that is wet or dirty	33 (35%)	49 (52%)	19 (51%)	10 (27%)
I like it when a dog jumps up on me	39 (41%)	37 (39%)	20 (54%)	11 (30%)
I like a dog licking my face or body	29 (31%)	46 (49%)	21 (57%)	8 (22%)
Sensory preferences: Sight
I will allow a dog to get close to my face	14 (15%)	68 (72%)	16 (43%)	16 (43%)
My reaction to a dog can depend on its physical appearance (e.g,. size, breed)*	47 (50%)	38 (40%)	10 (27%)	19 (51%)
I get nervous when a dog runs towards me*	52 (55%)	25 (27%)	12 (32%)	16 (43%)
Dogs that move around or run quickly upset me*	75 (80%)	12 (13%)	19 (51%)	13 (35%)
Social events are more enjoyable when a dog is present	3 (3%)	82 (87%)	6 (16%)	22 (59%)
Interacting with family members and friends is easier when a dog is present	2 (2%)	79 (84%)	11 (30%)	14 (38%)
Dogs distract me from conversations*	42 (45%)	33 (35%)	11 (30%)	16 (43%)
When I go out into the community, I fear that I will have an unpleasant encounter with a dog*	82 (87%)	6 (6%)	21 (57%)	12 (32%)
I would try and avoid going somewhere in the community where there are dogs present*	86 (91%)	3 (3%)	23 (62%)	11 (30%)
I am happy to go to an off-lead dog park	12 (13%)	63 (67%)	14 (38%)	17 (46%)
Seeing assistance animals in the community is less stressful for me than seeing other dogs*	17 (18%)	33 (35%)	7 (19%)	15 (41%)
Sensory preferences: Hearing/smell
I do not like when a dog barks suddenly*	26 (28%)	49 (52%)	7 (19%)	24 (65%)
The sound of a dog chewing on food bothers me*	69 (73%)	13 (14%)	23 (62%)	4 (11%)
Dog wee or poo upsets me*	60 (64%)	16 (17%)	7 (19%)	13 (35%)
I move away from a dog if it smells*	33 (35%)	28 (30%)	9 (24%)	17 (46%)

*Total score was calculated by reverse scoring item.

Results from Open-Ended Survey Responses

The quantitative data showed some key differences in overall affinity with dogs between participants with a dog (D) and those without a dog (ND). The qualitative data provided additional insights into participants’ attitudes and sensory preferences toward dogs. There were three major themes: (1) companionship and sense of responsibility; (2) participating in community and social activities; and (3) experiences of sensory input offered by a dog. Two of the major themes comprised sub-themes. Each theme and sub-theme will now be briefly discussed, with verbatim data identified by each participant’s unique numerical code [P1–170] and dog status specific to participants with a dog (D) and without a dog [ND].

Companionship and Sense of Responsibility

This theme was unique to participants with a dog, who described the sense of responsibility and companionship their dog gave them. For example, “My son strokes the dogs’ coat and LOVES their fluffy ears. He likes that dogs are non-judgmental and feels a strong sense of responsibility toward the dog in our home” [P99, D]. Participants identified three key benefits. Firstly, reducing feelings of loneliness: “My dog is exceptionally intelligent and very in tune with me, so I don’t feel alone with her always beside me” [P76, D]. Secondly, improving mental health: “My mental health is significantly improved when I have a dog at home” [P88, D]. Finally, the care and attention required by their dog was a motivating factor for some participants to undertake physical activity: “Dogs need a lot of care, attention and training and that helps me go out and do things because it’s all for her benefit” [P28, D].

Participating in Community and Social Activities

As participant 28 highlighted above, for participants with a dog, the needs of their dog were a motivational factor to leave their home and access the community. Participant 45 wrote: “My daughter always wants to bring our dog everywhere she goes” [P45, D]. The support of their pet or assistance dog increased some participants’ ability to engage in social activities:

My daughter would barely leave her room. Now she has an assistance dog it has changed her life completely. She now goes to university and catches buses all with the support of her assistance dog and her best friend, and freely engages with other people by opening up the conversation about her dog. [P75, D]

Helps me have confidence to go out if I have my dog with me. Helps me have conversations with strangers. Makes me braver. [P153, D]

The presence of a dog in the community also offered a benefit to one participant without a dog of their own. They wrote: “My son has a strong tendency to sprint away when in open spaces (beach, park, backyard). However, there tend to be less absconding behaviors when a dog is present” [P56, ND]. For three other participants without a dog, they perceived increased community participation to be a potential benefit of having a dog. This included pet dogs: “I prefer dogs to people. If I had a dog, I would be able to go to more events where dogs are encouraged/welcome e.g., dog parks and would take more walks” [P154, ND]; and assistance dogs: “An assistance dog would enable me to participate in social and community events with more confidence” [P106, ND].

This theme included three sub-themes: communication benefits; dogs as a social buffer; and negative encounters. Communication benefits were identified by participants with a dog and two participants without a dog. Participants commented on the increased communication with other family members when a dog was present: “Sometimes she will see a dog and ask one of us what the dog’s name is” [P154, ND]. For participants with a dog, they also identified that talking to a dog was easier than talking to people:

He enjoys the succinct communication with the dog that also doesn’t have (human) emotion to interpret. He doesn’t pick up on sarcasm or tone so it’s easy for him to communicate. He also talks “through” his dogs and will say “I think the dog is tired now, we better go.” [P97, D]

When my daughter is upset, she will at times say she is only talking to [name of dog] (her dog) and spend time talking to her before she is ready to interact with the rest of the family. Usually the only way I can coax her to visit people is if they have a dog she can play with. [P153, D]

The second sub-theme was unique to participants with a dog. They saw their dog as “a social buffer and anxiety reducer” [P29, D], providing a reason to talk to other people and make acquaintances:

My dog helps me have conversations with people when I’m out in the community. When we talk about my dog, I can talk easily and it’s easier to start making acquaintances and even friendships. My dog is the one thing that keeps me from being entirely socially isolated and helps keep my agoraphobia under control. [P76, D]

Two of the few community activities she enjoys most include visiting the dog park and beach walks on a dog friendly beach. She loves meeting all the dogs and will ask to pat them. Dogs are a huge social facilitator for her. She wouldn’t interact with people without the dogs being present. [P48, D]

The third sub-theme was negative encounters with dogs in the community. These were identified by participants with and without a dog but were not seen as issues with the dog, rather as “people problems.” Participants wrote: “Things I dislike about dogs have less to do with the dog and more to do with the owner” [P96, D]. This included “dogs not under control of their owner” [P105, ND]; “untrained dogs, which sadly is most dogs” [P72, D]; and “owners not cleaning up after their dogs” [P162, D]. Some participants discussed how they avoid certain places where dogs are present:

I have severe anxiety after several dog attacks because dog owners do not keep their dogs on leads. I avoid some places, but there are now very few places I can go to walk away from dogs. I feel trapped. [P4, ND]

Some participants empathized the “problem behaviors” they see in dogs, as this excerpt illustrates:

Dogs are infinitely trainable when you can find their motivations and work with them in a way that encourages a positive attitude. It distresses me a lot that most people don’t have enough knowledge of how to interact with dogs, leading to all the problem behaviors. It reminds me of how people treat us autistics. We are not dogs, but we are being misunderstood and punished for our distress, which is a traumatic experience. [P76, D]

Experiences of Sensory Input Offered by a Dog

Participants with a dog reflected on positive changes observed by the sensory input offered by a dog. They wrote, for example:

The puppy helped calm him and he strove to improve his self-control because he didn't want to scare or upset the puppy. Before having the puppy, he would become violent during meltdowns. [P93, D]

This theme comprises three sub-themes: dogs helping to control or regulate emotions; sensory dysregulation; and attitudes and sensory preferences dependent on familiarity and appearance.

The first sub-theme was how dogs helped to regulate emotions of participants with and, to a lesser extent participants without, a dog. For three participants without a dog, being in the presence of a dog was identified as offering benefits:

My son enjoys playing fetch with a ball and he'll stay in the area of the dog. When he becomes dysregulated, he'll press up against a dog; instead of other behaviors such as absconding or hitting other people. [P56, ND]

Similarly, for participants with a dog, patting, pressure, and hugging were identified as sources of stress relief and emotional regulation:

My son loves the sensory feeling of running his hands through dogs’ fur (long haired dogs, e.g., golden retriever). This has a very calming effect on him. The deep pressure of a dog leaning/laying on him also helps him with emotional regulation. [P69, D]

The second sub-theme related to the sensory features of dogs identified to cause sensory dysregulation. Participants without dogs wrote for example, “I hate their smell, hate them licking me” [P46, ND]; “I don’t like animals that shed or bark a lot” [P6, ND]; and “Too much barking can be overloading” [P154, ND]. Shedding of hair, drooling, licking, and barking were also disliked by some participants with a dog:

I do not like having dog fur on me or my things. I don't like having dog mess (saliva, mud, wee/poo, etc.) on me or my things. I am much less comfortable around a dog that smells. I need the dog to respect my boundaries and read my cue. [P88, D]

Finally, participants with and without a dog described how their attitudes and preferences were dependent on familiarity with the dog. Participants described being more “comfortable around dogs that I know in safe spaces” [P4, ND], and some highlighted differences between familiar and unfamiliar dogs: “Our dogs can do whatever they like to my son – run, jump roll and tumble – however unfamiliar dogs are triggering, and can create distress” [P47, D]. Participants with and without dogs also described various features of a dog they preferred, including their physical size: “I only like small dogs like sausage dogs (all my answers about liking dog features relate to small dogs only)” [P167, ND]; as well as their appearance and personality:

Some dog breeds have constant stringy drool, and I don’t like that. Some breeds have coats that I don’t like to touch because it’s a sensory issue. Some dogs bark too much which is also a sensory problem. [P76, D]

My son points out cute dogs a lot. Likes to pat dogs and see their tail wag. He really enjoys lightly patting or stroking dogs’ main body/back, especially if they are soft. He enjoys smaller dogs over larger dogs and prefers calm quiet dogs that don’t molt or smell. [P1, ND]

Discussion

This study explored affinity with dogs for both children and adults with autism, with affinity defined as the degree to which people value interactions with these animals (Bryan et al., 2014). The findings add to the existing body of research primarily focused on children with autism, with affinity levels in the current study shown to be similar for participants across the two age groups of people: those aged under 18 years of age and those 18 years or older. In this study, however, affinity was higher for participants who currently had a pet, assistance, or therapy dog compared with those who did not. This finding may suggest that affinity with dogs can be influenced by people with autism having more prolonged exposure, experience, and opportunities for desensitization to, and interaction with, dogs (Lisk et al., 2021), or a greater familiarity with dogs (Guerin et al., 2017; Hill et al., 2020; Kellert, 1985). However, it is important to note that, in the current study, many participants currently without a dog also had high levels of affinity with dogs. Consistent with previous research, this suggests the potential for strong animal attachments (Carlisle, 2015; Grandgeorge et al., 2012; Hill et al., 2020) and that levels of affinity with dogs may be influenced by multiple factors, including personal preferences, experience of disability, past interactions with animals (and animal handlers, including other pet owners), and/or animal-related factors (Hill et al., 2020; O'Haire, 2013).

Some people with autism have sensory processing that may include hypo- or hyper-reactivity to stimuli – assessing preferences for stimuli is therefore valuable when determining how to make interventions (including animal-assisted interventions) most effective (Clay et al., 2022; Marco et al., 2011). Similar to previous research, the current results suggest there are sensory stimuli inherent in the interactions between people with autism and dogs which may influence affinity with dogs (Harwood et al., 2019; Hill et al., 2020; London et al., 2020). For some participants in the current study, the sensory inputs offered by the dog, such as patting or stroking a dog, were beneficial, and – aligned with previous research – were identified to help regulate emotions and reduce feelings of stress (Bystrom & Lundqvist Persson, 2015; Carlisle, 2014; London et al., 2020). Conversely, other sensory experiences posed challenges. For example, a dislike of barking was at times reported to be linked to sensitivity to auditory stimuli that can be experienced by individuals with autism (Carlisle et al., 2018; Harwood et al., 2019). This finding provides further evidence of the importance of developing an understanding of the sensory preferences and processes for a person with autism when considering animal interactions or animal-assisted activities (Clay et al., 2022). Specific to pet, therapy or assistance animals, this information may then be used to identify the characteristics of a dog that could contribute to a higher level of affinity with the dog for an individual versus those that may have more likelihood of leading to aversion. For example, considerations may then be made regarding the size (i.e., preference for smaller or larger dogs) and breed-related characteristics (i.e., preference for short or long hair, dogs that do not shed hair, or dogs that do not drool) of a dog. The information could also be used to consider training for the dog, such as the need for a dog to only bark on command.

There was also evidence in the current study that for both adults and children with autism the presence of a dog was a motivating factor to engage in activities in the community (i.e., dog walking) (Burrows et al., 2008; Hall et al., 2016; London et al., 2020) and acted as a catalyst for communication with family, friends, and strangers (Berry et al., 2013; London et al., 2020; Martin & Farnum, 2002; Sams et al., 2006). However, both positive and negative experiences in the community were reported. For some participants, this led to a fear of dogs and an avoidance of community places where dogs would be present. Echoing prior studies, the current study findings highlight the need to continue to educate the community on responsible dog ownership, etiquette, and management (Mills, 2017; Stace, 2016). In the absence of education for community members, people with autism may benefit from planned interventions or strategies to address or mitigate these negative experiences. It will also be important to consider these mixed experiences when determining the most appropriate support for a person with autism, particularly for goals relating to social and community participation.

Methods to explore and understand affinity with dogs, such as the 27-item measure utilized in the current study, offer a means to gather information on affinity with dogs and the particular attitudes and sensory preferences of people with autism. This information holds relevance to people with autism, their families, and providers of animal supports such as animal-assisted therapy and assistance animal organizations. As emphasized by Hill et al. (2020), a secure attachment between the person with autism and the dog is essential for effective intervention. Awareness and consideration of how the dog’s personality and temperament interact with the preferences and behaviors of the person with autism is therefore essential to ensure effective human–animal interactions.

Study Limitations and Future Research

There were limitations in this study which should be considered. Despite attempts to recruit a sample of participants with various experiences of dogs, only 16 of the study participants had never owned a pet dog or worked with an assistance or therapy dog. This could be due to self-selection bias, with a sample consisting of people more motivated to participate owing to having stronger positive or negative affinity with dogs. There was also a higher response rate from people living in one state of Australia, which further reduces the external validity. Thus, results may not be representative of the broader population of people with autism. Additionally, over half of survey responses were completed by a next-of-kin of an individual with autism. It is uncertain as to whether the responses provided by next-of-kin reflect the true attitudes and preferences of the person with autism. Future research will benefit from additional recruitment strategies to attempt to ensure greater representation of the viewpoints of people with autism. This should also include alternative response options and/or methodological designs for people unable to complete an online survey or for whom augmentative and alternative communication is required (Guerin et al., 2017).

Finally, although the 27 items utilized in this study had excellent reliability, it was not possible to determine the psychometric properties of the customized survey tool used. To determine if the 27 items are reliable and psychometrically sound to measure affinity with dogs, the homogeneity (inter-rater reliability) and stability (test-retest reliability) will need to be evaluated extensively. Furthermore, the utility of the scores will be limited by the absence of normative data. Future research may therefore include the collection of normative data representative of the broader population to assist to interpret the scores of people with autism in relation to those of other respondents.

Conclusion

The findings from this study begin to inform recommendations for individuals with autism, their families, and service providers – including animal-assisted therapy or assistance animal organizations – when considering individual participation goals and determining if animal supports are the most appropriate option. In particular, this study further highlighted that dog affinity may be influenced by multiple factors, including personal preferences, experience of disability, past interactions with animals and animal handlers (including other pet owners), and/or animal-related factors. Without an awareness and understanding of these factors it will not be possible to ensure an appropriate match between the dog and the person. This in turn risks conflict between the sensory preferences of the person with autism and the sensory input offered by a dog. This could also limit the effectiveness of the support provided by the pet, assistance, or therapy dog to assist a person to achieve their goals.

Acknowledgements

Thanks are extended to the people with autism and their next of kin for their involvement in this research. The authors also acknowledge the assistance animal organization that identified and proposed the need for this research, and staff from these organizations who drew on their lived and practice experience to provide input to the design of survey questions.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Notes

1 The Australian Federation of Disability Organisations’ (AFDO) language guide, which encourages the use of person-first language, has been followed in this manuscript; however, the authors acknowledge some people may preference “identity-first” language and respect individual choice of language that a person uses about themselves.