2. Materials and Methods
An online survey was created in Qualtrics to assess perceptions held by members of the United States general public regarding service dogs, emotional support dogs and therapy dogs. The survey was designed, reviewed, and tested by the co-investigators and their colleagues at Colorado State University (CSU) and North Carolina State University (NCSU) after seeking input from representatives from the community who either trained, owned, or interacted with either service, emotional support, or therapy dogs. They were asked to provide feedback on content, navigability, survey questions and choices, and overall questionnaire design. Revisions were made based on this feedback in an ongoing process that involved several reiterations. After this step was completed, the survey was tested by approximately ten individuals for any additional ambiguity and/or potentially missing or inappropriate response options. The survey originated from CSU and received approval from the Institutional Review Board at CSU. The online survey was posted as a human intelligence task to Amazon Mechanical Turk (MTurk). Participants were recruited online, from 21 March 2017 to 4 April 2017. A small payment was offered for completion of the task, and offered to a total of 505 participants. All data were collected anonymously. The survey began by asking participants to rate their ability to define the terms service dog, therapy dog, and emotional support dog. After participants responded to this question, definitions of each type of assistance dog were provided for use in the remainder of the survey.
The next section of the survey asked participants to indicate their perceptions of the value and acceptability of service, emotional support and therapy dogs, followed by a set of questions inviting respondents to estimate the percentage of people misrepresenting a need for a service or emotional support dog. Subsequent sections enquired about the questions that can legally be asked of someone with a dog that appears to be an assistance animal, as well as the types of settings in which these dogs should be allowed. A true/false group of questions then assessed participants’ understanding of the regulations surrounding each type of assistance dog. The next question set examined participants’ ability to discern between service, emotional support and therapy dogs. The last section of the survey contained demographic questions, such as age and educational level. It also asked about respondents’ exposure to assistance animals in general, and whether they owned a service or emotional support dog.
Survey responses were downloaded into SPSS (IBM Corp, Released 2012, IBM SPSS Statistics for Windows, Version 21.0, Armonk, NY, USA) for data analysis. The initial 505 submissions were screened in order to exclude respondents who personally owned an emotional support or service dog. The remaining data sets were screened for quality control, via a set of three questions with required responses. Participants who did not follow instructions for all three screening questions (e.g., enter the word “dog” in this blank) were removed from further analysis, yielding a total of 284 usable responses. This is roughly equivalent to a 56.3% response rate.
Descriptive statistics and frequency distribution (reported in percentages) were performed using commercially-available software Chi-square was used to test for any differences in perceived prevalence based on respondents’ gender, age, educational level, and personal pet ownership on misrepresentation of a need for an assistance dog (service or emotional support). Comparisons were also made between respondents with greater exposure to assistance dogs (owned by friends/family), and those who did not report such experiences. Differences were considered significant when p < 0.05.
3. Results
The online survey was posted from 21 March 2017 to 4 April 2017 with a total of 505 responses obtained. The data was then screened as described above, resulting in a total of 284 responses. However, the total number of responses for each question varied, since not all participants answered each question, therefore, the total numbers of responses are indicated where necessary.
The gender distribution of participants was approximately equal, with 135 males (47.5%) and 149 females (52.5%) represented in the sample. Most participants (131, 46.1%) were between the ages of 26 and 35 years old, and the vast majority had completed a college degree, with 97 (34.2%) reporting completion of a four-year degree, and an additional 115 (40.5%) indicating they held a graduate degree (
Table 1). Slightly more than one-half of the participants (153, 53.9%) reported owning a pet dog, while 108 (38.0%) indicated they had a friend or family member who owned a service dog, and 109 (38.4%) reported having a friend or family member who owned an emotional support dog.
When first asked to rate their ability to define three types of service dogs prior to any definitions given (
Table 2), the majority of participants rated themselves as very comfortable or somewhat comfortable in defining service dog, emotional support dog, and therapy dog, with service dog receiving the largest proportion of very comfortable ratings (151, 53.2%).
Once participants had rated their ability to define each type of assistance dog, they were presented with the definitions listed below, and asked to answer the rest of the survey using these definitions.
Service dog—a dog that is individually trained to do work or perform tasks for the benefit of an individual with a disability, including a physical, sensory, psychiatric, intellectual, or other mental disability.
Emotional support dog—a dog that provides companionship, relieves loneliness, and can help with depression, anxiety, and certain phobias, but does not have special training to perform tasks that assist people with disabilities.
Therapy dog—a dog that provides people with therapeutic contact, usually in a clinical setting, to improve their physical, social, emotional, and/or cognitive functioning.
3.1. Perceptions of Assistance Dogs
The majority of participants agreed or strongly agreed that it was acceptable for people to have a service dog if it was deemed helpful. Approximately half of the respondents disagreed (67, 23.6%) or strongly disagreed (90, 31.7%) with limiting the privileges granted to emotional support and therapy dogs. After completing these questions, 21.1% of participants were still not confident in their understanding of the differences between service dogs, emotional support dogs, and therapy dogs (
Table 3).
As seen in
Table 4, there was a substantial amount of confusion about the questions that can be legally asked to determine if a dog qualifies as an assistance dog. Most respondents (163, 57.4%) recognized that it is illegal to ask an individual to provide proof of a disability, and a slightly smaller number (138, 48.6%) correctly stated that assistance dog handlers may not be asked to identify their disability. In regards to the dogs themselves, the majority of respondents correctly recognized what is allowable to enquire about the tasks an assistance dog is trained to perform (159, 56.0%). However, the responses were approximately evenly distributed when it came to enquiring if the service dog is required because of a disability. Only 113 (39.8%) of participants correctly identified this question as permissible. Also of concern was the fact that just 81 participants (28.5%) selected the correct response regarding a request to see proof of the dog’s status, such as an identification card or certificate.
Overall, the public seems to agree that assistance dogs should have broad access to residential dwellings, transportation, and educational institutions. Service dogs appeared to have the greatest rate of acceptance, with 172 participants (60.6%) endorsing service dogs’ right to ride in airplane cabins, 170 (59.9%) agreeing they should be allowed to reside in school dormitories, and 163 (57.4%) agreeing that they should be allowed in classroom settings. Only 43 participants (15.1%) felt that landlords should be able to refuse service dogs. The response frequencies for emotional support dogs were slightly less favorable, with less than half of the participants feeling they should be allowed in airplane cabins, dormitories, or classroom settings, but only 54 respondents (19.0%) felt that landlords should be able to exclude emotional support dogs. The responses for therapy dogs were the least favorable regarding access to public spaces. Finally, 10–13% of participants expressed a lack of understanding of what is allowable for all types of assistance dogs (
Table 5).
3.2. Prior Interactions with Assistance Dogs
Most respondents had minimal prior contact with assistance dogs in public settings. Approximately half (121, 49.6%) of respondents had only minimal exposure to assistance dogs in public venues such as work, school, and airports in the year preceding the study. Fifty-five participants (19.4%) reported no exposure in public settings, with just 88 participants (31.0%) reporting 5 or more sightings or interactions with assistance animals.
3.3. Assistance Animal Misrepresentation
The relationship between respondents’ gender and rate of perceived prevalence of assistance animal misrepresentation was explored for both service dogs and emotional support dogs (
Table 6). The chi-square statistic was significant (
p = 0.007), with more women than men reporting lower incidence rates of service dog misrepresentation. A similar pattern (chi-square
p = 0.025) was found for the misrepresentation of a need for an emotional support dog. There was no relationship between participants’ highest level of education and their perceptions of misrepresentation for either type of assistance animal. Participants’ age was related to their perception of misrepresentation of a need for a service dog (chi-square = 0.001), but not for an emotional support dog (chi-square = 0.063).
The effects of canine companionship, as well as exposure to assistance animals on participants’ perception of misrepresentation was also examined (
Table 7). Participants with friends/family who owned a service dog were more likely to suspect misrepresentation of a need for a service dog (chi square < 0.000) or emotional support dog (chi square < 0.000), than those without contextual exposure to service dogs. A similar trend was found for suspected misrepresentation of need for an emotional support dog. Participants with friends/family who owned a service dog (chi square = 0.029) or emotional support dog (chi square = 0.001) were more suspicious of misrepresentation than those without such exposure.
4. Discussion
As assistance animals become more prevalent in society, there appears to be a parallel increase in the frequency of allegations of misrepresentation or fraudulent representation of animals as assistance animals. However, no objective studies have been conducted to characterize the public’s understanding of the varying roles and rights of assistance animals, the laws governing their access to public spaces, or their general perceptions.
The inconsistency in terminology used to describe assistance animals, as well as the lack of standardized certification requirements for these animals and overlapping sets of legal regulations make it difficult for the public to determine an animal’s role or job [
2]. This confusion paves the way for misrepresentation of pets as assistance animals in order to gain illegitimate benefits, thereby increasing skepticism and scrutiny of legitimate assistance animals. Even though 19 states have passed laws against fraudulent representation of assistance animals [
16], public confusion and lack of a standardized certification process makes these laws all but unenforceable from a practical point of view. The goal of this study was to address public perception regarding three types of assistance dogs: service dogs, emotional support dogs, and therapy dogs.
In this study, despite expressing a high level of confidence in defining the roles of service dogs, emotional support dogs, and therapy dogs, participants were still not able to fully apply these definitions and identify the legally-acceptable questions they can ask to identify a true assistance dog. It is interesting to note that a greater percentage of respondents expressed confidence in their ability to define a service dog than an emotional support dog or a therapy dog. This might be related to the longer history of working with assistance animals as guide dogs for the blind [
19], as well as the more readily observable tasks performed by a service dog assisting a physically handicapped user. In contrast, emotional support dogs alleviate an invisible disability, making it more challenging for casual observers to understand their role and how they are helping an owner with a mental illness. After reading the definitions of each type of assistance dog, study participants indicated their agreement with a series of statements about the perceived ability of an assistance animal to help users with a legitimate disability. Overall acceptance of assistance animal ownership was high, with the majority of participants agreeing that there is nothing wrong with people having an assistance animal if they feel it is useful. Once again, service dogs generated the highest rate of acceptance. But, it is interesting to note that a higher percentage of respondents strongly agreed with the statement that there “I see nothing wrong with providers using therapy dogs if they think they are helpful” than did with a similar statement regarding the use of emotional support dogs. This might be related to the perceived value/legitimacy of these animals, or to the clearer definition of a therapy dog’s role. Even though service dogs were more broadly accepted, participants still felt that other types of assistance dogs should be given special privileges. However, despite voicing these strong opinions, 21% of participants stated that they were not comfortable describing the differences among assistance animals, thus calling into question the knowledge base for their opinions.
This issue is further complicated by the difficulties associated with identifying a legitimate assistance dog in a public setting. In order to protect the rights of disabled individuals, the ADA limits the questions that may be asked when the type of service being provided by an assistance animal is not obvious. Under these laws, people may only ask if the dog (or miniature horse) is a service animal that is required because of a disability, and what work or task the dog has been trained to perform. Titles II and III of the ADA apply to state/government entities and areas of public accommodation [
3,
4]. Since access to residential dwellings is governed by the Fair Housing Act, providers are allowed to ask individuals seeking emotional support animals (of any species) as a reasonable accommodation to provide documentation from a health professional. This letter must establish that the individual in question has a disability, and the emotional support animal will provide some type of disability-related assistance or emotional support [
7]. While airlines may sometimes request this same type of letter, this type of proof is not required under the Air Carrier Act. Instead, airlines must accept identification cards, other written documentation, the presence of harnesses, tags, or the credible verbal assurances of a qualified individual with a disability using the animal [
13,
14].
Against this backdrop of overlapping regulations, the broad range of opinions regarding allowed access levels for each type of assistance dog isn’t surprising (
Table 5). The majority of participants felt service dogs should be allowed in airplane cabins, school dormitories and classrooms. These perceptions are in line with current laws regarding access to public areas for service dogs. A potential area for concern is that 15.1% of participants felt that landlords should be able to refuse service dogs. This practice is illegal under the FHA [
6,
7]. The mismatch with public perceptions may be due to a lack of understanding—either of the role of a service dog, or of the overlapping laws governing access to housing.
Public opinion regarding access for emotional support dogs was not as positive, with only 40.1% of respondents agreeing they should be able to ride in airplane cabins. Although this right is protected by the Air Carrier Act [
13,
14], the lower threshold of proof required for access may have resulted in a greater number of people having a negative perception of emotional support dogs. This could be due to an encounter with a poorly-trained emotional support animal, or one perceived to be fraudulent, in the confines of an airplane cabin. The FHA rules about emotional support animals also apply to residential dwellings, as well as school dormitories, guaranteeing these animals the right to reside with their owners [
7]. Thus, it is concerning to see that 19.0% (
n = 54) of the public feels that property owners should be able to refuse emotional support dogs within their properties, and begs the question of what might be driving this opinion. Moreover, while a sizeable percentage of respondents (46.1%, 131) agreed that emotional support dogs should be allowed to reside in school dormitories, the remaining 53.9% either opposed this accommodation or were unsure as to whether it should be available. While we might surmise why this perception is held (e.g., college students’ lack of time to properly care for assistance animals), the legal system has previously ruled that dormitories are dwellings [
20,
21]. Public endorsement of emotional support dogs in classrooms was poor, with only 34.5% (98) of respondents supporting their right to be present in classroom settings. This opinion matches current laws, as emotional support animals are not guaranteed the right to be present in classroom settings. Masinter [
22,
23] explains that even if an ESA is allowed to live in campus housing, an independent determination should be made as to whether the ESA is also needed for classroom access, based solely on the ADA and section 504 of the Rehabilitation Act standards.
Overall, participants’ responses regarding therapy dogs were more closely aligned with existing legislation, which does not grant this type of dog any special access rights. Respondents were more likely to agree that property owners should be able to refuse therapy dogs (22.5%, 64) than other types of assistance animals. This belief matches FHA law, which does not consider therapy dogs to be a reasonable accommodation [
6,
7] because they are not providing a service to a specific individual. Interestingly, the public was slightly more supportive of therapy dogs being allowed in classroom settings (38.0%, 108) than they were of emotional support dogs, and about equally supportive of therapy dogs riding in airplane cabins (40.5%, 115). Therapy dogs were the least likely to be viewed as appropriate for residing in school dormitories (39.4%, 112). The reasons behind this perception are an interesting area for future research, especially in light of the already-existing processes for evaluation and registration of therapy dogs. We speculate that public perceptions may be based upon the definition of therapy dogs as being part of a treatment team, and the assumption that professionals, such as therapists and counselors do not usually reside in dormitories.
Even though most participants in this study had minimal prior contact with service dogs and emotional support dogs, their overall perception of the legitimacy of these types of assistance dogs was quite high. As seen in
Table 6, most respondents felt that there were relatively low incidence rates of misrepresentation. The chi-square comparison for gender found that women reported lower perceived rates of misrepresentation for both service dogs and emotional support dogs than men. This might be because women are, by nature, more trusting, as research has found that increased levels of testosterone correlate with increased levels of suspicion [
24] and decreased interpersonal trust [
25]. Furthermore, women are more likely to regain trust after repeated transgressions [
26], which could explain why even multiple exposures to animals clearly misrepresented as assistance dogs would make less of an impression on female respondents.
Participants’ age was related to their perception of fraudulent use of service dogs, but not emotional support dogs, with participants over 45 years of age reporting lower perceived frequencies of misrepresentation of service dogs. A partial explanation for this perception can be found in prior research by Poulin [
27], which found that levels of interpersonal trust increase as people age. However, the fact that this trend was only statistically significant for service dogs warrants further exploration. It is possible that longer history and stringent accreditation requirements associated with service dogs, specifically guide dogs [
19], were determining factors in how these dogs are perceived. Additionally, the rather recent proliferation of ESAs in general [
28] might have affected the public’s perception of emotional support dogs.
An interesting finding was that participants with friends/family who owned either an emotional support or a service dog felt that there was a higher proportion of fraudulent use of both types of assistance dogs. Since we did not ask participants to explain the rationale for their responses, we can only speculate about the cause for this finding. Further research is necessary to determine if this result is related to participants observing their friends/family engaging in fraudulent representation of an assistance animal. Another possible explanation could be greater awareness of the duties performed by a true service animal, and consideration of how many people actually need this level of support. Finally, participants may also be aware of how the increased prevalence of misrepresented animals delegitimizes their friend/family member’s need for their assistance animal.
Despite these statistical differences, in general, most respondents to our survey perceived relatively low rates of assistance dog misrepresentation. While the exact numbers vary by participants’ demographic characteristics, on the whole, most people feel that fewer than 25% of people with a service dog or emotional support dog are “taking advantage of the system” and fraudulently representing a pet as an assistance dog. Assuming that the results of this survey are representative of true public opinion, it appears that the media may be over-representing the views of a small, but vocal, minority when it comes to allegations of assistance dog fraud. This is not surprising, given recent research [
29,
30] and the old adage that “bad news sells”.
While the results of the survey describe a generally favorable landscape for assistance dogs, our findings must be interpreted within the limitations of this study. This study relied on participants to self-report their true opinions and knowledge regarding assistance dogs, thus there is the possibility for social desirability bias. However, this was mitigated as much as possible by anonymous data collection, and by tapping a pool of participants with no prior connections to the researchers. The age distribution of participants led to the classification of all subjects over 45 years of age into a single group. In order to more fully investigate the relationship between age and perceived rate of assistance dog representation, future research should involve a greater number of older adults, representative of a broader age range.