Factors affecting social workers' inclusion of animals in practice.

Risley-Curtiss, Christina ; Rogge, Mary E. ; Kawam, Elisa 等

The popular media and the professional literature are replete with ways that animal-human relationships (AHR) can be beneficial or harmful to both the human and the animal. Over 63 percent of U.S. households have companion animals, and the vast majority considers them friends or family members (American Pet Products Manufacturers Association, 2007). These relationships are often characterized by reciprocity, with humans giving and receiving emotional support from their companion animals (Risley-Curtiss, Holley, et al., 2006).

Other areas in which research strongly supports positive benefits from interactions with companion animals include physical and mental health. Physiologically, researchers have found that having a companion animal can help reduce the cardiovascular effects of stress (Allen, Blascovich, & Mendes, 2002), and lower blood pressure and cholesterol levels (Walsh, 2009). With regard to mental health, a range of studies indicate positive AHR can assist children and adults in reducing anxiety, depression, and social isolation (Friedmann & Tsai, 2006). AHP. also can assist people with mental illnesses, such as schizophrenia, better cope with their disease (Beck, 2005). Studies also suggest that the presence of an animal can enhance the professional helping process of social workers and others. The presence of a dog may, for example, shorten the time needed for rapport building between counselors and clients; clients may be more willing to share personal information with counselors (Levinson, 1997; Schneider & Harley, 2006).

Animal abuse, another facet of AHR, is correlated with an increased risk of aggression against humans, including child abuse and domestic violence (DeGue & DiLillo, 2009) and other forms of deviant behavior (Gullone, 2011). Animal abuse is considered one symptom of conduct disorder in children. Animal hoarding, defined as failure to provide minimal standards of care for multiple animals, is also a phenomenon that may be on the increase and with which human service professionals are increasingly called on to intervene (Brown, 2011).

Despite the many ways that AHR infuse our lives, human service professions have been slow to embrace the relevance of animals in the lives of clients, and, hence, to include AHP, in research, education, and practice. Yet ecological-systems theory, family-centered practice, social support theory, and the strengths perspective are all central models of social work practice that support the inclusion of AHR (Arkow, 2007; Risley-Curtiss, 2009). Given that animals are part of many clients' ecologies, it would be appropriate for social workers to ask about AHR during assessments. Positive AHR can be considered protective factors for children and adults who experience any number of issues, such as family or other violence or isolation, especially due to illness or age (Castelli, Hart, & Zasloff, 2001). The potential for healing through relationships with animals can be incorporated through animal-assisted interventions (AAI).

The purpose of the current study was to shed light on what influences whether social workers do include animals in practice. We examined whether specific practitioner factors, including certain demographics, having a companion animal, contributing to animal welfare agencies, education level, and area of practice are associated with inclusion of AHR in social work assessments, interventions, and treatment of animal-related issues in clients.

INCLUSION OF AHR IN PRACTICE

Although there is ample documentation of the helpful and hurtful connections between humans and other animals, several studies have documented the lack of inclusion of AHR by various human service professions. Nelson (2002), in a study of 203 psychologists, found only 14 percent assessed for animal abuse, although 94 percent believed animal abuse to be connected to other human behavioral disturbances. Zilney and Zilney (2005), in a study of cross-reporting between child welfare workers and humane society workers, found that a number of child welfare workers thought cross-reporting of animal abuse was unimportant and were resistant to including animal welfare in their assessments. Although a number of these child welfare workers identified issues with animal wellbeing in their case notes, many failed to report their concerns to humane society workers.

Schaefer, Hays, and Steiner (2007) surveyed 174 therapists regarding animal abuse issues and found that therapists primarily asked about the abuse only after a client had mentioned it. Of those therapists working with clients who had witnessed or been threatened with animal abuse, only 16 percent regularly asked about such abuse in their initial assessments; only 14 percent of therapists working with actual animal abusers regularly asked about it. Reasons for their reluctance to address issues of animal abuse reflected a lack of familiarity with the topic, a need for more information, and attitudes demonstrating that human safety takes precedence over animal lives.

A recent survey of 46 state public child welfare agencies revealed that almost 75 percent of states do not provide training to new Child Protective Services workers with regard to asking whether families have animals. Only 20 percent provide information on assessing AHR, and only 17.4 percent train new workers to recognize and assess animal abuse (Risley-Curtiss, Zilney, & Homung, 2010). Barriers to incorporating AHR included lack of knowledge of AHR and administration-initiated direction as well as the use of standardized computer risk and safety assessments that take time and money to change. The lack of familiarity with AHR was also reflected in a study of attitudes and perceptions of AAIs among student nurses, and psychiatry and pediatric residents (Eaglin, 2008). Although 87 percent of respondents said they would prescribe AAIs for patients, 89 percent reported they had no formal training about AAIs.

Ascione (2005) asserted that "developmental psychology and related disciplines have virtually ignored the positive role that companion animals and other animals may play in the lives of children" (p. 5). Social work is one such related discipline in which research supports Ascione's assertion. Risley-Curtiss (2004) surveyed 511 accredited schools of social work regarding content on the human-animal bond in their courses. Of the 230 schools that responded, only six included any courses with 25 percent or more AHR content. What was offered was mostly regarding animal-assisted therapy (AAT). A national survey of social work practitioners found that only one-third asked questions about other animals in their intake assessments and even fewer (23.2 percent) included animals as part of their interventions (Risley-Curtiss, 2010).

Recently, Sato (2011) explored the effects of social workers' attachment to their companion animals, as well as organizational factors, such as procedural flexibility, on assessing the roles that animals play in clients' lives. She found that, of a regional sample of 427 agency social workers, only 7.5 percent always assessed AHR, 19.3 percent frequently did, and 21.2 percent sometimes did. Sato found that level of attachment to one's own animal and the presence of agency protocols about AHR increased the frequency of assessment of AHR with clients. This is the only study found to date that directly analyzes factors explaining social worker's inclusion of AHR.

The current study addresses this gap by examining data from a national social work practitioner study (Risley-Curtiss et al., 2010). The authors examined whether certain practitioner personal and professional demographics (for example, gender, area of practice), exposure to or knowledge about information or training on AHR (for example, having heard or read about treating animal abuse, knowing other social workers who include animals in their interventions), and personal experience with animals (for example, having a companion animal, involvement in animal welfare) are associated with four social work practice applications: (1) asking questions about animals in their assessments, (2) including animals as part of their interventions, (3) treating clients for animal abuse, or (4) treating clients for the loss of an animal.

METHOD

Sampling

This study used a subsample from a national random sample of U.S. NASW members who identified as BSW/MSW/PhD clinical/direct practitioners (Risley-Curtiss, 2010). This larger sample of 4,991 was drawn from the 2004-2005 NASW membership mailing list. Three survey mailings were completed between June and December 2005, with the first and third including a cover letter, hard copy survey, and self-addressed return envelope; and the second a postcard reminder. Participants could also complete the survey online using SurveyMonkey.com. A small lottery incentive of Amazon.com gift certificates was also included. The final sample consisted of 1,649 respondents, for a 33 percent response rate. There was no information on why people did not respond or whether nonresponders are significantly different from responders. From these 1,649, the current study used 1,262 participants; the others were excluded because of missing data on the tested variables. The missing variables were examined, and no trends were found that would suggest skewed data.

Measures

The current study used a subset of 19 questions from the original 48-question survey. The original survey was developed on the basis of theory, AHR literature, and previous research. Two international experts in the AHR field established content validity through a literature review and overall review of the survey. The survey was also pilot tested with a group of 22 MSW students. Factor and correlation analyses were used to reduce the number of variables entered into logistical regression models to insure enough power to detect effect sizes.

Dependent Variables. For this study, four dichotomous dependent variables that asked about inclusion of animals in various aspects of social work practice were used. They were operationalized as whether or not participants asked questions about animals in their assessments (no=0, yes= 1); included animals as part of their interventions (for example, AAT or AAI in their social work practice [no/yes]; treated clients for animal abuse [no/yes]; and treated clients for companion animal loss and grief [no/yes]). When these questions were not examined as the dependent variable, they were included as independent variables; the authors hypothesized, for example, that if someone assesses for animals they might be more likely to include them in interventions.

Independent Variables. The subset of independent personal and professional demographic variables used in this study included gender (male/ female) and age measured as total years. Ethnic identity was measured as the participant's "primary" ethnic identity using the nine categories from the 2004-2005 NASW membership mailing list. Because of the small number of people of color (for example, African Americans=2.2 percent), however, ethnic identity was recoded 0 = Caucasian and 1 = other. Response categories for "primary client population worked with" and "primary work focus" also replicated those categories on the NASW mailing list. Primary client population categories were children, adolescents, nonelderly adults, and elders. Primary work focus categories were family issues and individual/behavioral problems.

Five questions measured how much participants had heard about or read information on AHR, including the link between animal cruelty and child abuse, or domestic violence; the positive influence animals can have on various age groups; and treatment of clients who have abused animals or had experienced the loss of a companion animal. These variables were operationalized dichotomously as very little/none and some/a lot. Internal consistency for the original information exposure items was an acceptable .82 using Cronbach's alpha.

Three variables measured other aspects of participants' exposure to information on AHR: (1) whether they had any special training on including animals in social work practice (no/yes), (2) whether they knew any other social workers who included animals in their practice (no/yes), and (3) whether they would like to learn more about the human-animal bond (no/yes).

Three independent variables measured participants' personal experience with animals: (1) whether participants currently had companion animals (yes/ no), (2) whether they had made a financial contribution to an animal protection organization (yes/no), and (3) whether they had volunteered at an animal shelter or rescue group within the last year (yes/no).

Analysis

We used frequencies and measures of central tendency to describe the data. Stata, a data analysis and statistical software program, was used for analyzing the data. Logistic regression was used because the four dependent variables were all categorical and dichotomous. Because there is no other research on this topic to suggest any order, all variables were entered into each model simultaneously.

RESULTS

Sample Description

Seventy-nine percent of the 1,262 participants were female, 92.8 percent were Caucasian, and the mean age was 53 years (SD = 8.03). Mean income was $60,530 (SD = $38,562). This sample is very similar to the original study sample (80 percent female, 92 percent Caucasian, and mean age of 53). In addition, the sample is similar to the 2002 NASW Practice Research Network (2003) survey, which found that most NASW regular members were female (79 percent), Caucasian (87 percent), and had a median age of 50. A little more than 65 percent of our 1,262 participants had companion animals. In the year preceding their response to our survey, 7.5 percent had volunteered at a local animal shelter/rescue group and 45.6 percent had contributed financially to an animal protection organization.

In this sample, 95.7 percent had an MSW as their highest degree and an average of 21.7 years post-BSW/MSW experience (median = 21, SD = 7.93). Individual/behavioral problems were the major work focus for 44 percent of participants, and 14.7 percent focused on family issues were spread out among seven other primary work categories (for example, alcohol/drug abuse, employment related, and health), ranging from 1.2 percent to 8.6 percent. A majority (53.2 percent) worked primarily with nonelderly adults, whereas 16.2 percent worked primarily with families, 11.7 percent with children, and 11.6 percent with elders.

In relationship to exposure to information about AHR, 77.5 percent of participants had read or heard about the link between animal and child abuse and 69.7 percent had read or heard about the link between animal abuse and domestic violence. About 71 percent had read or heard about treatment of clients who have experienced animal loss, but only 28.7 percent had done so with regard to clients who abuse animals. One-third of participants said they knew other social workers who include animals as part of their interventions. Just over half (53.7 percent) said they would have taken a course on AHR if there had been one in their social work program. The vast majority (95.9 percent) reported that they had no training on including animals in practice.

One-third of participants reported that they included questions about companion or other animals in their intake assessments; 21.8 percent included animals as part of their interventions. Only 5.7 percent of participants said they treated clients for animal abuse, whereas 56.2 percent said they treated clients for companion animal loss and grief.

Inclusion of Animals in Social Work Practice

The logistic regression results for each of the four dependent variables: (1) inclusion of questions about animals in intake assessments, (2) inclusion of animals in interventions, (3) treatment of animal abuse, and (4) treatment of animal loss are provided in Table 1.

Model 1; Inclusion of Questions about Animals in Intake Assessments. The model was significant ([chi square] = 187.44, p < .001) with seven variables statistically significant in explaining whether social work practitioners included questions about animals in their intake assessments: (1) if their primary client populations were children (p < .01), (2) if their primary client populations were elders (p < .05), (3) the more they were exposed to information on the link between animal abuse and domestic violence (p < .05), (4) the more they were exposed to information on the on treatment of animal abuse (p < .05), (5) if they knew other social workers who included animals in their practice (p < .01), (6) if they wanted to know more about AHR (p < .05), and (7) if they treated clients for animal loss (p < .001).

Model 2: Inclusion of Animals in Interventions. The model was significant ([chi square] = 243.76, p < .001) with l I variables explaining whether social workers included animals in interventions with clients. Seven variables related to their exposure to AHR information. The more information they had on (1) the link between animal abuse and domestic violence (p < .001), (2) the positive influence of animals on humans (p < .001), and (3) treatment of clients who have experienced loss of a companion animal (p < .05), the more likely they were to include animals in interventions. Those who (4) knew other social workers who include animals in their practice (p < .001), (5) had special training on including animals in practice (p < .001), and (6) would like to know more about AHR (p < .01) were more likely to include animals in interventions. On the other hand, (7) the less information participants had on the link between child and animal abuse, the more likely they were to include animals in interventions (p < .001).

In addition to these seven variables, animals were more likely to be included in interventions by participants who worked primarily with elders (p < .05) and had companion animals themselves (p < .001). Participants were also more likely to include animals in interventions if they also treated clients for animal abuse (p < .00l) or companion animal loss and grief (p < .001).

Model 3: Treatment for Animal Abuse. The model was significant ([chi square]= 159.57, p < .001), with five variables explaining whether participants treated clients for animal abuse. Participants who (1) did not work primarily with nonelderly adults (p< .01), had more information on (2) the treatment of those who abuse animals (p < .001), and (3) who knew other social workers who included animals in their practice (p < .01) were more likely to treat clients for animal abuse. They were also more likely to treat for animal abuse if they (4) included animals in interventions (p < .001), and (5) treated clients for loss of companion animals (p < .05).

Model 4: Treatment of Animal Loss. The model was significant ([chi square]= 431.57, p < .001). Eleven variables explained whether participants treated clients for companion animal grief and loss. Participants who (1) were Caucasians (p < .05); (2) worked primarily with nonelderly adults (p < .001); and (3) on individual behavior (p < .001) or (4) family issues (p < .001) were more likely to treat clients for animal loss. Participants were less likely to treat clients for companion animal loss (5) the more exposure to AHR information they had on the positive influence of animals on humans (p < .05) and (6) the less information they had on treatment of individuals who abuse animals (p < .01). They were more likely to treat clients for companion animal loss if they (7) had more information on the treatment of clients who had experienced animal loss (p < .001) and (8) knew other social workers who included animals in practice (p < .05); practitioners who (9) had animals of their own were also more likely to treat for loss (p < .0l) than those without animals. Participants who (10) asked about animals in assessment (p < .001) and (11) included them in interventions (p < .001) were more likely to treat for animal loss.

LIMITATIONS

Several limitations should be considered when interpreting these results: The generalizability may be limited due to a lower response rate in the initial study than was desired, despite using multiple methods. The current sample demographics do appear to be similar to NASW's own and the original larger sample. The sample also did not appear to be biased in favor of those most interested in animals, as might be anticipated. About 65 percent of this sample had companion animals, which is very close to the 63 percent of American households who have companion animals (American Pet Products Manufacturers Association, 2007). Less than 50 percent of the sample had donated to an animal protection organization in the last year, and only 7.5 percent had volunteered at a shelter. Finally, the use of logistical regression has limitations in that a model is selected to fit a particular sample. Thus, there is no assurance that the same model would be selected if another sample from the same population were used. In addition, the model will always fit the sample better than the population from which it was selected (Norusis, 1989).

DISCUSSION AND IMPLICATIONS

Research suggests that social workers are only slowly beginning to embrace the importance of AHR for client well-being. This study begins to identify factors that can help the profession integrate AHR further into different facets of social work practice.

With regard to personal and professional demographic factors, populations served were significant in each model; however, the specific populations varied and depended on the type of practice. That is, participants who worked with children, elders, and nonelderly adults asked questions about animals during assessments, whereas only those who worked with elders were most likely to include animals in subsequent interventions. Participants who worked with nonelderly adults were most likely to treat for animal loss, yet less likely to treat for animal abuse. These findings suggest that although practitioners working across populations are asking about animals, assessment does not translate directly and consistently into animal-assisted intervention (AAI) and treatment. In addition, it has been the experience of one of the authors that many social workers "don't want to hear the details" with regard to animal abuse. Grief over the loss of an animal may be easier to understand and treat than animal abuse. This may explain why, in this sample of over 1,000 social work practitioners, only 5.7 percent treated clients for animal abuse and 56.2 percent treated clients for companion animal loss and grief. A fruitful, future exploration may be a comparison between these outcomes with the percentage of social workers who treat clients for more commonly recognized types of abuse (for example, children, domestic violence) and loss and grief. Treatment of loss may be a more ubiquitous and less specialized aspect of treatment than abuse for social workers generally, and in AHR.

The relationship between having nonelderly adults as a primary population and treatment for animal abuse was one of five negative relationships in the models. Two independent variables in the negative relationships were demographics (ethnicity and nonelderly population served), and three were AHR information exposure variables (participants had heard or read about the link between animal and child abuse, treatment of those who abuse animals, and positive influences of animals on humans). All negative relationships occurred in the AAI and treatment models. It is interesting that practitioners who knew less about the link between animal and child abuse were more likely than their counterparts to include animals in interventions. This finding is a concern because children who are abused, especially physically or sexually, may be more likely to abuse animals (Ascione, 2005). In addition, AAI, such as humane education and animal-assisted activities, are commonly provided to children (Fine, 2010). IfAAI is used with children who have been abused and the worker is insufficiently knowledgeable about animal abuse, both animals and children may be put in harm's way, and an important opportunity for intervention is missed.

A second negative demographic relationship occurred in the treatment of animal abuse model with having nonelderly adults as the primary population. This may reflect our finding that practitioners were more likely to assess for AHR in relationship with children and elders than with nonelderly adults. Social workers who do not assess such relationships may be less likely to uncover animal abuse. Adults who abuse animals, and are thus at greater risk for engaging in other forms of criminal behavior (Gullone, 2011), may end up in the criminal justice system, where charges for other crimes, such as assault or robbery, may take precedence over discovery of co-occurrence and treatment of animal abuse. Furthermore, individuals charged with animal abuse are often put on probation or referred to non-animal-related diversion programs. Evidence suggests the majority will not be referred for appropriate treatment; to date, Ani-Care (Jory & Randour, 1999) is the only nationally known adult treatment model.

The third negative demographic relationship, workers were less likely to treat people of color for animal loss, is consistent with research that indicates a lack of attention to AHR and people of color (Kaufman, 1999; Risley-Curtiss, Holley, & Wolf, 2006). This lack of attention may be due to the misperception that people of color do not care about their animals in the same way white people do (Richards, 2004) and, therefore, may not experience their loss as strongly.

The two final negative relationships occurred between treatment of animal loss and the information exposure variables of participants having heard or read about treatment for those who abuse animals and having heard or read about animals' positive influence on humans. These findings are puzzling and require scrutiny in future studies.

Outcomes for other information exposure variables were interesting in that participants who reported that they include animals in assessment were more likely to have heard or read about the link between animal abuse and domestic violence and about treatment of animal abuse but not about the link between child and animal abuse. Perhaps this finding reflects the current state of research, which appears to be more developed for the link between domestic violence and animal abuse than for child and animal abuse. Research on child maltreatment and animal abuse is growing, yet it does not appear to have reached many social work practitioners or education programs.

One of the most important findings in the study overall was that social workers who know other workers who include animals in their practice are more likely to do the same. This was the singular independent variable that attained significance in all four models. As a form of exposure to information about the emerging practice area of AHR, this outcome suggests that learning about AHR through some form of direct relationship with other practitioners was more salient than otherwise hearing and reading about AHR. This finding is consistent with diffusion of innovation theory, which holds that "interpersonal communication is usually more effective when there is a high degree of professional resemblance between the individual attempting to introduce the innovation and the recipient" (Sanson-Fisher, 2004, p. S56).

The relational nature of this AHR information exposure variable is particularly relevant because less than 5 percent of the sample had special training to use AHR. Because training may not be readily available to workers, the worker-to-worker relationships become even more important. Networking with other AHR-experienced social workers is a way for workers to learn from each other and augment their own professional skill set. There are a few examples of formal efforts at networking, such as the New York City NASW chapter's Social Workers Advancing the Human--Animal Bond Interest Group (http://www.naswnyc.org) and the scheduled Animals and Social Workers special issues meeting at the Society for Social Work and Research. Future research in this area should include more detailed assessment of the nature of the relationships and types of information sharing, mentoring, or other exchanges that are most effective in further distribution of AHR practice applications.

The relationship between specialized training and only the dependent variable of AAI makes sense, given the current status of AHR training. That is, a great majority of AHR training is AAI-focused and does not address assessment or treatment of animal abuse. In the current study, the majority of participants who had received specialized training identified some form of AAI training in an open-ended follow-up question.

Of the three personal experience variables, only having a companion animal was significant, in two models. First, practitioners' having a companion animal increased the likelihood of their using AAI. This finding makes sense in that people who use AAI often have their own therapy animal and there is more AAI training available than for other AHR practice applications, including treatment of animal abuse (Chandler, 2005; Fine, 2010). The University of Denver School of Social Work has an animal-assisted therapy certificate. The Humane Society University is developing its own AAI certificate program. Second, practitioners who had companion animals were more likely to treat clients for animal loss. Grief over animal loss is often dismissed as unimportant or seen as not normal (Barton-Ross & Baron-Sorensen, 2007). Practitioners who have their own companion animals may be more attuned to clients who experience animal loss and, thus, more likely to treat for loss.

A surprising finding was that having one's own companion animal, financially contributing to animal protection organizations, and volunteering in a shelter did not affect whether participants asked about animals during assessment. It was expected that those who are "more into" animals would be more likely to ask about animals. Although the lack of a relationship between having a companion animal and assessment differs from Sato's (2011) finding, it is perhaps a positive finding that workers do not have to be actively involved in animal welfare to include questions about animals in assessment.

CONCLUSION

The growing body of interdisciplinary evidence about the importance of positive AHR and the correlation between animal abuse and other antisocial behaviors should not continue to be largely ignored by the social work profession. More research is needed to discover other factors--such as agency environments, computerized case record systems, lack of education and training, and social work's humancentric focus--that can influence the inclusion of AHR in social work practice. The current study suggests that a number of practitioners are seeking out additional information on AHR, through training on AAI and networking, to supplement the limited availability of content in academic courses and formal trainings. It behooves training bodies, accrediting organizations, universities, state agencies, and private nonprofit organizations to provide access to accurate sources of information, trainings, and Webinars. Doing so will expand the clinical knowledge and skill base while greatly enhancing the quality and breadth of services that social work practitioners can offer to clients. Doing so meets social workers' professional and ethical obligations; offering the best services for clients is a core concept of any helping profession and of the individuals who take on the mantle of those professions.

doi: 10.1093/sw/swt009

Original manuscript received January 18, 2012 Final revision received May 24, 2012 Accepted June 19, 2012 Advance Access Publication March 29, 2013

REFERENCES

Allen, K. M., Blascovich J., & Mendes, W. B. (2002). Cardiovascular reactivity in the presence of pets, friends, and spouses: The truth about cats and dogs. Psychosomatic Medicine, 64, 727-739.

American Pet Products Manufacturers Association. (2007). New national pet owners survey details two decades of evolving pet ownership. Retrieved from http://www.appma.org/press_releasedetail.aspPid-109

Arkow, P. (2007). Animal maltreatment in the ecology of abused children: Compelling research and responses for prevention, assessment, and intervention. Protecting Children, 22(3-4), 66-79.

Ascione, F. R. (2005). Children and animals: Exploring the roots of kindness and cruelty. West Lafayette, IN: Purdue University Press.

Barton-Ross, C., & Baron-Sorensen, J. (2007). Pet loss and human emotion (2nd ed.). New York: Routledge.

Beck, A. M. (2005). Review of pets and our mental health: The why, the what and the how. Anthrozoos, 18, 441-443.

Brown, S. E. (2011). Theoretical concepts from self psychology applied to animal hoarding. Society & Animals, 19, 175-193.

Castelli, P., Hart, L. A., & Zasloff, R. L. (2001). Companion cats and the social support systems of men with AIDS. Psychological Reports, 89, 177-187.

Chandler, C. K. (2005). Animal assisted therapy in counseling. New York: Routledge.

DeGue, S., & DiLillo, D. (2009). Is animal cruelty a "red flag" for family violence? Investigating co-occurring violence toward children, partners, and pets. Journal of Interpersonal Violence, 24, 1036-1056.

Eaglin, V. H. (2008). Attitudes and perceptions of nurses-in-training and psychiatry and pediatric residents towards animal-assisted interventions. Hawaii Medical Journal, 67, 45-47.

Fine, A. (Ed.). (2010). Handbook on animal-assisted therapy: Theoretical foundations and practice guidelines (3rd ed.). San Diego: Academic Press.

Friedmann, E., & Tsai, C. C. (2006). The animal-human bond: Health and wellness. In A. Fine (Ed.), Handbook on animal-assisted therapy: Theoretical foundations and practice guidelines (2nd ed., pp. 95-117). San Diego: Academic Press.

Gullone, E. (2011). Conceptualizing animal abuse with an antisocial behavior framework. Animals, 1, 144-160.

Jory, B., & Randour, M. L. (1999). The AniCare Model of Treatment for Animal Abuse. Ann Arbor, MI: Animals and Society Institute.

Kaufman, M. E. (1999). The relevance of cultural competence to the link between violence to animals and people. In F. R. Ascione & P. Arkow (Eds.), Child abuse, domestic violence and animal abuse (pp. 260-270). West Lafayette, IN: Purdue University Press.

Levinson, B. M. (1997). Pet-oriented child psychotherapy. Springfield, IL: Charles C Thomas.

National Association of Social Workers Practice Research Network. (2003). Demographics. Retrieved from http://wwwsocialworkers.org/naswpm/survey/Two/Datagram2.pdf

Nelson, P. (2002). The survey of psychologists' attitudes, opinions, and clinical experiences with animal abuse (Doctoral dissertation, The Wright Institute, 2002). Dissertation Abstracts International, 63, 1040.

Norusis, M.J. (1989). SPSS/PC+ update for V3.0 and V3.1 [Computer software]. Chicago: SPSS, Inc.

Richards, J. (2004, January/February). Minority report: Part one: African Americans. Best Friends, 14-16.

Risley-Curtiss, C. (2004). Animal content in schools of social work, Unpublished study.

Risley-Curtiss, C. (2009). The role of animals in public child welfare work. In A. Linzey (Ed.), The link between animal abuse and human violence (pp. 126-141). Portland, OR: Sussex Academic Press.

Risley-Curtiss, C. (2010). Social work practitioners and the human-companion animal bond: A national study. Social Work, 55, 38-46.

Risley-Curtiss, C., Holley, L. C., Cruickshank, T., Porcelli, J., Rhoads, C., Bacchus, D., et al. (2006). "She was family": Women of color and their animal-human connections. Afflia, 21,433-447.

Risley-Curtiss, C., Holley, L. C., & Wolf, S. (2006). The animal-human bond and ethnic diversity. Social Work, 51, 257-268.

Risley-Curtiss, C., Zilney, L. A., & Homung, R. (2010). Animal-human relationships in child protective services: Getting a baseline. Child Welfare, 89, 67-82.

Sanson-Fisher, R. W. (2004). Diffusion innovation theory for clinical change. Medical Journal of Australia, 180, $55-$56.

Sato, A. (2011). Social workers' attachment to their pets, organizational structures, and their impact on professional assessment regarding roles pets play in clients' lives (Unpublished doctoral dissertation). University of Connecticut, Hartford.

Schaefer, K. D., Hays, K. A., & Steiner, R. L. (2007). Animal abuse issues in therapy: A survey of therapists attitudes. Professional Psychology: Research and Practice, 38, 530-537.

Schneider, M. S., & Harley, L. P. (2006). How dogs influence the evaluation of psychotherapists. Anthrozoos, 19, 128-142.

Walsh, F. (2009). Human-animal bonds 1: The relational significance of companion animals. Family Process, 48, 462-480.

Zilney, L. A., & Zilney, M. (2005). Reunification of child and animal welfare agencies: Cross-reporting of abuse in Wellington County, Ontario. Child Welfare, 84, 47-66.

Christina Risley-Curtiss, PhD, MSSW, is associate professor, School of Social Work, Arizona State University, 411 N. Central Avenue, Suite 800, MC 3920, Phoenix, AZ 85004; e-mail: risley.curtiss@asu.edu. Mary E. Rogge, PhD, is associate professor, College of Social Work, University of Tennessee, Knoxville. Elisa Kawam, MSFV, is a doctoral student, School of Social Work, Arizona State University, Phoenix. Table 1: Summary of Logistic Regression Analyzes for Variables Predicting Inclusion of Animals in Practice (N=1,262) Model 1: Assess Demographic b(SE) OR Age (in years) -.01(.01) 0.99 Gender (male) -.18(.17) 0.84 Primary ethnic identity (Caucasian) -.28(.27) 0.75 Population served: Children .80(.26) ** 2.23 Population served: Nonelderly adults -.17(.21) 0.85 Population served: Elders .57(.27) * 1.77 Work focus: Individual behavior -.04(.15) 0.96 Work focus: Family issues .08(.22) 1.08 Information exposure and knowledge Link between animal abuse and child -.90(.13) 0.91 abuse Link between animal abuse and domestic .29(.13) * 1.33 violence Positive influence of animals on humans .07(.05) 1.07 Treatment of those who abuse animals .29(.12) * 1.34 Treatment of those who experienced pet .13(.11) 1.14 loss Special training--inclusion of animals .62(.33) 1.86 Knows other social workers who include .41(.14) ** 1.51 animals Wants to know more about AHR .40(.18) 1.49 Personal experience Currently has companion animal .26(.15) 1.3 Volunteered at an animal shelter .17(.24) 1.19 Financially contributed to an animal .15(.14) 1.16 organization Practice application Includes animals in assessment Includes animals in interventions .17(.16) 1.18 Treats for animal abuse .52(.29) 1.69 Treats for animal loss .66(.15) *** 1.93 Constant 2.92(.64) *** pseudo [R.sup.2] 0.18 Chi-square [chi square] =187.44 Model 2: Intervene Demographic b(SE) OR Age (in years) .02(.01) 1.02 Gender (male) -.14(.20) 0.87 Primary ethnic identity (Caucasian) .18(.32) 1.2 Population served: Children -.03(.34) 0.97 Population served: Nonelderly adults .29(.26) 1.34 Population served: Elders .79(.32) 2.2 Work focus: Individual behavior .09(.18) 1.09 Work focus: Family issues .13(.26) 1.14 Information exposure and knowledge Link between animal abuse and child -.53(.16) ** 0.59 abuse Link between animal abuse and domestic .49(.16) ** 1.63 violence Positive influence of animals on humans .22(.07) *** 1.25 Treatment of those who abuse animals -.14(.14) 0.87 Treatment of those who experienced pet .27(.13) * 1.3 loss Special training--inclusion of animals 1.22(.33) *** 3.39 Knows other social workers who include .66(.16) *** 1.94 animals Wants to know more about AHR .84(.25) *** 2.31 Personal experience Currently has companion animal .53(.19) ** 1.69 Volunteered at an animal shelter .21(.27) 1.23 Financially contributed to an animal .16(.16) 1.17 organization Practice application Includes animals in assessment .19(.16) 1.21 Includes animals in interventions Treats for animal abuse 1.18(.30) *** 3.27 Treats for animal loss .85(.19) *** 2.35 Constant -6.87(.85) *** pseudo [R.sup.2] 0.12 Chi-square [chi square] =243.76 Model 3: Treat Abuse Demographic b(SE) OR Age (in years) -.01(.02) 0.99 Gender (male) .65(.34) 1.19 Primary ethnic identity (Caucasian) .06(.54) 1.07 Population served: Children .44(.45) 1.55 Population served: Nonelderly adults -1.10(.42) ** 0.33 Population served: Elders -.96(.65) 0.38 Work focus: Individual behavior .11(.34) 1.11 Work focus: Family issues .33(.43) 1.4 Information exposure and knowledge Link between animal abuse and child .50(.30) 1.65 abuse Link between animal abuse and domestic -.16(:29) 0.85 violence Positive influence of animals on humans -.15(.12) 0.86 Treatment of those who abuse animals 1.38(.22) 3.96 Treatment of those who experienced pet -.41(.24) 0.66 loss Special training--inclusion of animals .72(.46) 2.05 Knows other social workers who include .91(.29) ** 2.49 animals Wants to know more about AHR -.69(.36) 0.5 Personal experience Currently has companion animal .07(.34) 1.07 Volunteered at an animal shelter -.08(.50) 0.92 Financially contributed to an animal .28(.31) 1.32 organization Practice application Includes animals in assessment .39(.31) 1.48 Includes animals in interventions 1.15(.31) *** 3.16 Treats for animal abuse Treats for animal loss .86(.36) * 2.35 Constant 4.49(1.37) ** pseudo [R.sup.2] 0.29 Chi-square [chi square] =157.57 Model 4: Treat/Loss Demographic b(SE) OR Age (in years) .01(.01) 1.01 Gender (male) -.28(.17) 0.76 Primary ethnic identity (Caucasian) -.61(.27) * 0.54 Population served: Children .07(.28) 1.08 Population served: Nonelderly adults 1.17(.22) *** 3.21 Population served: Elders .05(.28) 1.05 Work focus: Individual behavior .70(.16) *** 2.02 Work focus: Family issues .90(.23) *** 2.46 Information exposure and knowledge Link between animal abuse and child -.08(.13) 0.92 abuse Link between animal abuse and domestic .15(.14) 1.17 violence Positive influence of animals on humans -.12(.05) * 0.89 Treatment of those who abuse animals -.41(.13) ** 0.67 Treatment of those who experienced pet 1.34(.12) *** 3.81 loss Special training--inclusion of animals .15(.43) 1.16 Knows other social workers who include .34(.15) * 1.4 animals Wants to know more about AHR .25(.17) 1.29 Personal experience Currently has companion animal .39(.15) ** 1.47 Volunteered at an animal shelter .02(.28) 1.02 Financially contributed to an animal .24(.15) 1.27 organization Practice application Includes animals in assessment .69(.16) *** 1.99 Includes animals in interventions .89(.20) *** 2.43 Treats for animal abuse .69(.35) 1.99 Treats for animal loss Constant -3.46(.67) *** pseudo [R.sup.2] 0.25 Chi-square [chi square] =431.57 Note: AHR=animal-human relationships; OR = odds ratio. * p<.05. ** p<.01. *** p<.001.