Fostering Rescued Dogs: An Exploratory Study of the Experiences of Foster Care Providers

Animals in the care of shelters and rescues frequently have health and behavioral challenges that must be addressed before they can be adopted. Volunteer foster providers perform an important service by bringing vulnerable animals into their homes to address their needs therefore making them better prepared for successful adoption. Despite their critical role in animal rescue, the experiences of volunteer foster providers have not been examined systematically. Given the difficulty recruiting and retaining these volunteers and the potential for them to experience negative outcomes such as compassion fatigue and secondary traumatic stress caused by performing care work with rescued animals, such research is necessary. Current and former foster providers (N = 85) completed an online survey of their perceptions of the positives and negatives of the experience and measures of psychological outcomes of care work. The findings show that fostering rescued dogs comes with a mix of positive and negative experiences. Further, the findings suggest the importance for rescues of being aware of the risk factors associated with compassion fatigue and protective factors predicting compassion satisfaction.

high (Hill et al., 2020). They also found that compassion fatigue increased with performing euthanasia and exposure to neglect and cruelty cases.
At present, no empirical research has been published which involved exploring the social and psychological impacts of the care work provided by people volunteering to foster rescued animals. In this research, I investigated the experiences of current and former foster care providers who volunteer with dog rescue organizations to answer the following qualitative questions: 1) What motivated them to become foster care providers for dogs? 2) What do they see as the most rewarding aspects of fostering dogs? 3) What do they see as the most challenging aspects of fostering dogs? 4) Why do they stop fostering? In addition, I explored the following quantitative questions: 1) How are outcome variables compassion satisfaction, burnout, secondary traumatic stress, and eudemonic wellbeing predicted by functions performed by volunteering?
2) How do current and former foster providers differ on both the outcome and predictor variables? Given the exploratory nature of the study, I did not have an a priori set of hypotheses. However, I did have expectations based on my own experiences fostering dogs and those of other people fostering in my organization. I employed Grounded Theory Method to analyze the qualitative data.

Method Participants
This research was reviewed and approved by the Human Subjects Committee which is part of the Institutional Review Board at the University of Mount Union. The research was conducted in full accordance with the policies of that committee including recruiting, informed consent, and debriefing. Participants were recruited by posting a link to the survey on social media sites related to animal welfare and rescue. Participants were also be recruited by directly contacting shelters and rescue organizations and asking them to share the link to the study to their volunteers. Finally, participants were recruited through the Human-Animal Interaction Section (Section 13), Division 17 of the American Psychological Association and the Companion Animal Psychology Facebook page. Informed consent and debriefing were conducted online. Participants were directed to contact me if they had questions or concerns. The research plan included providing information related to mental health resources and suicide prevention to participants if contacted. I was not contacted by any of the participants.
Ninety-two individuals began the survey but seven did not go beyond the consent form to finish responding leaving 85 participants. Of the participants, 57 reported that they were currently fostering (62%) while 24 indicated that they were former foster caregivers (26.1%). Two participants reported that they had never fostered, and one indicated that they had fostered previously but did not have a foster in their care currently. Most participants were female (n = 81, 88%) and white (n = 81, 88%). One participant reported being Hispanic or Latino/Latina, one reported being Asian or Asian American, and one reported being mixed. The bulk of the participants were between 25 and 64 years of age (91% of participants) with the most common age group being between 45 and 54 (n = 31, 33.7%). Two participants were between 18 and 24, 14 were between 35 and 44, 16 were between 55 and 64, and 5 were 65 and over. Sixty-one (66.3%) participants reported being employed and working full-time. Of the remaining participants, 7 (7.6%) were employed part-time, 8 (8.7%) were not employed and not looking for work, 8 (8.7%) were retired, and 1 (1.1%) reported being disabled and unable to work. Fifty-three (57.6%) participants reported serving in capacities in addition to fostering in their organizations while 32 did not (34.8%). Most of the participants reported volunteering with a non-profit rescue that is not breed-specific (n = 50, 54.3%). Twenty-four participants (26.1%) were volunteers with breedspecific organizations. The remaining participants volunteered with city or county pounds (n = 2, 2.2%), a Humane Society or ASPCA shelter (n = 4, 4.3%), or a breed and non-breed non-profit (n = 2, 2.2%). One participant indicated volunteering with multiple types of organizations while two indicated that they did not volunteer with any of the indicated organizations.
When asked how long they had been or were involved in fostering, 48 (52.2%) indicated 0 to 5 years, 21 (22.8%) chose 6 to 10 years, and 16 (17.4%) reported fostering for more than 10 years. The responses to the question concerning how many total dogs the participants had fostered were difficult to interpret. The number of dogs reported ranged from 0 to 900 (M = 55.14, SD = 123.126). It is unclear whether some participants who reported fostering very large numbers of dogs were referring to personally doing so in their homes or if they were reporting the number of dogs fostered by an organization in which they were leaders or coordinators. In retrospect, the question should have been worded differently to ensure that respondents answered the question only in terms of the number of dogs personally fostered in their own homes.

Design
The design of this exploratory study was descriptive and correlational. The Grounded Theory Method (GMT) was employed to conduct thematic content analysis of open-ended responses. As described by Bryant (2017), GMT is an inductive approach to research with qualitative data. In general, the process involves systematically grounding the data through the identification of thematic categories which arise from the words of participants.

Materials
The survey was constructed and delivered through SurveyMonkey. The following demographic information was collected: age, race/ethnicity, sex, and employment status. The purpose of collecting demographic information was only to describe the sample rather than to make between groups comparisons. Participants were asked to indicate if they were current foster providers or former foster providers. They were asked to indicate if they serve in other capacities with animal rescue organizations and to identify what type of organization they work with (e.g., publicly funded shelter, non-profit rescue, breed rescue, etc.). They were also asked how long they had been fostering dogs and how many total dogs they had fostered.
The survey included the following open-ended questions: Why did you become involved in fostering dogs? What are the most positive or rewarding aspects of your fostering experience? What are the negatives or most challenging aspects of your fostering experience? If you no longer foster dogs, why did you stop? Thematic coding schemes (see Appendix) were generated by the author and an undergraduate research assistant who, independently, read all the participants' responses to the open-ended questions. While reading, we noted each novel theme we encountered while reading on notecards. After doing so, we met to review and reduce the number of categories and discuss examples of thoughts belonging to each category. The next step was for us to independently code the responses given by five participants after which we met to discuss discrepancies in the application of the coding schemes. This was followed by each of us coding all the participants' responses. When we had both completed coding, I evaluated the hit rate (agreement between the two coders) after the first pass. We met a final time to discuss and resolve discrepancies prior to calculating the frequency of thoughts in each category. The initial hit rate for the first question assessing why participants chose to foster was only 76%. After discussion, we found that we were interpreting thoughts referring specifically to helping dogs differently. We were essentially placing these thoughts in two different categories. After discussion, we resolved the problem by refining both categories.
With the second open-ended question (positives of fostering), the initial hit rate was 84%. As with the previous question, the sources of the discrepancies were discussed and resolved. For the question concerning the negatives of fostering, the initial hit rate was 89%. Finally, for the question concerning why people stopped fostering, the initial hit rate was 97%. Discrepancies were discussed and resolved.
Compassion satisfaction, burnout, and secondary traumatic stress were assessed using the Professional Quality of Life Scale (ProQOL) (Stamm, 1995). Permission for use of the scale was granted through The Center for Victims of Torture (www.ProQOL.org). This scale is made up of 30 items responded to on a 5-point scale from 1 (never) to 5 (very often). The scale generates separate scores for compassion satisfaction, burnout, and secondary traumatic stress. Each subscale contains 10 items. Five items were reverse-scored before total subscale scores were calculated. High subscale scores indicate greater experience of the quality being measured. The only modification to the scale was in the instructions which were modified to refer specifically to fostering dogs. For each subscale, the highest possible score is 50 and the lowest 10.
The Volunteer Functions Inventory (VFI) (Clary et al., 1998) was used to assess perceptions of the functions performed by volunteering in the participants' lives. The scale is made up of 30 items rated on a 7-point response scale ranging from 1 (not at all important/accurate) to 7 (extremely important/accurate). The scale generates scores on 6 5-item subscales (protective, values, career, social, understanding, and enhancement). Total subscale scores were calculated with high scores indicating greater importance in each domain. For each subscale, the highest possible score is 35 and the lowest 5.
Wellbeing was assessed using the Questionnaire for Eudaimonic Well-Being (QEWB) (Waterman et al., 2010). Questions on this 21-item measure are responded to on a 5-point scale from 0 (Strongly Disagree) to 4 (Strongly Agree). Seven items were reverse-scored before calculating a total scale score. Higher scores reflect greater Eudaimonic well-being. The highest possible score is 105 and the lowest 21.
Self-care habits were assessed using a modified version of the Self-Care Assessment for Psychologists (SCAP) (Dorociak et al., 2017). The instructions and the items were modified to refer specifically to work and habits while fostering dogs. The scale includes 21 items divided into 5 subscales. The items are rated on a seven-point scale ranging from 1 (never) to 7 (almost always). The subscales are in the areas of professional support (5 items), professional development (5 items), life balance (4 items), cognitive strategies (4 items), and daily balance (3 items). For the professional support and professional development subscales, the highest possible score is 35 and the lowest 5. For the life balance and cognitive strategies subscales, the highest possible score is 28 and the lowest 4. For the daily balance subscale, the highest possible score is 21 and the lowest 3. Total subscale scores were calculated with higher scores indicating greater frequency of the experience indicated in the item.
Other materials included informed consent and debriefing forms which were delivered online.

Procedure
Participants followed the link to the survey in SurveyMonkey where they first saw the Informed Consent form. They then completed the demographic items followed by the questionnaires measuring the variables. They then saw the debriefing statement. Participation was expected to take no more than 40 minutes.

Why Did You Become Involved in Fostering Dogs?
The results of the content analysis for the question concerning why respondents became involved in fostering dogs are presented in Table 1 The next most common type of thought at approximately 9% of statements involved seeing that there was a need. Sometimes these expressions involved indicating that participants were responding to becoming aware of an organization's need for fosters. For example: Rescue made a general public request for fosters. Participant  As indicated in Table 1, the remaining categories comprised 4% or less of thoughts expressed in respondents' answers.

What Are the Most Positive or Rewarding Aspects of Your Fostering Experience?
The content analysis results for the question concerning the positive and rewarding aspects of fostering are presented in Table 2. Combined, four types of thoughts made up nearly 90% of expressions made by respondents to this question. References to seeing fostered dogs adopted into a good home or with a loving family were the most common thoughts (35% The second most common type of thought involved expressions about seeing dogs flourish, recover, grow, and/or thrive (33%).
The most rewarding is watching the fearful animals learn to trust again. Thematic Coding Category Frequency Percent Saving dogs' lives/Helping dogs/animals (e.g., euthanasia concern, concern for animal welfare, concern about bad shelter outcomes, help specific breed) 17 18.48%   Table 3, the remaining categories represented 4% or less of expressions.

If You No Longer Foster Dogs, Why Did You Stop?
Despite the wording of the question, the most common thought participants gave in response to the question concerning no longer fostering was that they were still doing so or that the question was not applicable (52% of expressions). Statements concerning no longer having room for additional dogs made up roughly 14% of expressions. Examples of these thoughts include: We adopted a dog we fostered and did not have room for another in our home.

Participant 2
We've fostered and failed (adopted   Table 5 contains the descriptive statistics for the full sample for the quantitative variables in the study. Unfortunately, of the 85 participants, an average of only 63 completed the quantitative scales. This represents approximately 74% of the initial sample. On the Professional Quality of Life Scale (PROQOL), average compassion satisfaction scores tended to be high while scores for burnout and post-traumatic secondary stress were typically low. Responses to the Volunteer Functions Inventory (VFI) revealed fairly high average scores for the value expressive and selfunderstanding functions of volunteering. Protective, social, and enhancement mean scores were moderate. The mean for seeing volunteering as a means of career development was low. While the mean for responses to the Questionnaire for Eudaimonic Well-Being (QEWB) was moderate, the reliability level of the scale was unacceptably low limiting the ability to interpret the results. As a result, hypothesis testing with this variable was not pursued. Responses to the modified version of the Self-Care Assessment for Psychologists (SCAP) revealed moderately high means for selfcare through professional support, professional development, life balance, and cognitive strategies while the mean for daily balance was moderate.

Correlations
The correlations among scores on the Professional Quality of Life Scale (ProQOL) subscales are presented in Table 6. There was a large, significant, negative correlation between compassion satisfaction and burnout and a large, significant, positive correlation between burnout and secondary traumatic stress. None of the correlations between ProQOL subscale scores and QEWB scores were significant.
Correlations between the subscales from the ProQOL (compassion satisfaction, burnout, and secondary traumatic stress) and subscales from the Volunteer Functions Inventory (VFI) and the modified Self-Care Assessment (SCAP) are presented in Table 7. Compassion satisfaction was predicted by three domains from the VFI (protective, understanding, and enhancement). In all three cases, the correlation was small and positive. Compassion satisfaction was also predicted by three domains from the SCAP (professional development, life balance, and cognitive strategies). In the case of self-care through professional development and through maintaining life balance, the correlations were large and positive while that for self-care using cognitive strategies was medium and positive. There were significant negative correlations between four of the self-care subscales (professional development, life balance, cognitive strategies, and daily balance) and burnout. These correlations were all medium in size. Finally, there were three significant negative correlations between secondary traumatic stress and forms of self-care (professional development, life balance, and daily balance). The correlation with professional development was small, while those with life balance and daily balance were medium.  Tables 8, 9, and 10 contain the means and standard deviations for the ProQOL, VFI, and SCAP respectively. While the goal was to compare individuals who were currently fostering dogs with those who had stopped fostering either permanently or temporarily, the majority of participants reported currently fostering and only one person was temporarily not fostering. As a result, between groups comparisons are difficult to interpret. The results of ANOVA tests revealed non-significant differences between groups in all cases.

Discussion
This project was born out of a combination of both professional and personal interests. As researcher, I have a general interest in human-animal interactions. As a social psychologist I have a particular interest in how people describe their experiences, beliefs, and values. I also foster dogs for a Labrador Retriever rescue in Northeast Ohio. While the study launched in late 2019 prior to the onset of restrictions due to the COVID-19 pandemic, data collection continued through spring semester of 2020 as lockdowns quickly went into place. While it was not typically referred to as a reason for becoming involved in fostering, the pandemic likely had unintended impacts on the study given the shortage of available dogs apparently caused by suddenly homebound people adopting pets in large numbers. The impact of the COVID-19 pandemic on the work of animal rescues is a topic that should be investigated systematically in the future given that information on the topic at present is anecdotal.    The research questions explored in this study came primarily from my experiences as a foster provider and my conversations with others performing various roles in my rescue organization. A search of the available literature revealed a paucity of research on people, particularly volunteers, doing foster work with animals. As noted in the introduction to this report, recruiting and retaining foster providers is a challenge for rescues (Edge Research, 2017). This exploratory study was the result. My hope is that the findings will spark further research and discussion of ways to improve the process of on-boarding, training, and supporting foster providers to improve outcomes for rescue organizations.
The first question explored in this research concerned participants' motivations for fostering dogs. Not surprisingly, the most common motives reported involved a desire to help dogs and save the lives of dogs followed by simply loving and being passionate about them. Other motives included giving back, contributing to the greater good, and meeting unmet needs identified by rescues. Some participants indicated a desire to have dogs on a short-term basis. It was rare for participants to say they were trying out fostering as an alternate route to dog ownership. As foster "failure" or choosing to adopt one's foster(s) can be a problem for rescues as it can limit the ability for foster providers to take on new dogs, this was a motive I was particularly interested in pursuing. I will return to this issue when discussing reasons participants gave for no longer fostering.
Participants reported many rewards of fostering dogs the most prominent of which was the joy of seeing dogs in permanent homes with loving people. Participants often referred making a difference for the dogs they foster and for the people who adopt them. In addition, it was common for participants to write about the transformation dogs go through during their time in foster care. Many referred to enjoying watching their fosters come out of their shells and learn to just be dogs. Some participants noted that they enjoy the novelty of meeting new dogs and the challenges presented by their unique personalities and backgrounds.
While participants saw clear benefits to fostering, they were also quite clear about the drawbacks. The most common difficulties noted by participants had to do with problem behaviors that are difficult to address. Many participants noted that dogs often come to them with little or no obedience training, poor socialization with humans and other animals, and/or with destructive behaviors. Many noted the difficulty of integrating a foster dog into their households which often include resident dogs and other pets. For some foster providers, particularly new ones with limited experience with dogs that have been abused, neglected, or that are poorly socialized, difficult dogs may be beyond their current abilities. This can be problematic for rescue organizations. In addition to potentially losing a foster provider due to bad experiences with dogs in their care, the dogs must be moved from foster to foster sometimes resulting in more trauma to the dogs and greater difficulty finding permanent homes.
In addition to behavioral challenges, many participants noted that it is painful to see the physical and psychological condition of many rescued dogs. I know the pain of this quite well as I have cared for fearful, unsocialized dogs that came from puppy mills. One of these dogs was so terrified that she would involuntarily defecate and urinate at the site of unfamiliar people, particularly men. Some participants expressed anger at the abuse and neglect inflicted by people resulting in problems that are sometimes very difficult if not impossible to correct. Many participants reflected on the preventable diseases and health problems faced by foster dogs sometimes necessitating euthanasia. For some foster providers, such experiences are incredibly painful.
As expected, many participants noted that it is frequently difficult to let dogs go when they have become attached to them. Given that the attraction to fostering for many volunteers is their love of dogs and their desire to make their lives better, this is not surprising. As argued by Hanrahan et al. (2018) and Bradshaw et al. (2012), caring for animals is shaped by the humananimal bond (HAB). At the same time, professionals in animal care are given very little education or training in the navigation of the impacts of the HAB (Hanrahan et al., 2018). The fact that volunteer fosters find becoming over-attached aversive, is a challenge for rescue organizations. In some cases, foster providers are unwilling to let go and decide to adopt their charges. While that is a good outcome for the individual dog, it means that dogs are not going to approved adopters and foster providers who "fail" may drop out because they no longer have room for additional dogs.
The final open-ended question concerned why participants were no longer fostering. The most common response was that they did not have room for any more dogs. Most participants citing this reason referred specifically to having adopted dogs they had fostered. In addition to "foster-failure", some participants stated that something had changed in their lives (e.g., housing, health, or relationship status) while others referred to their own dogs or other pets being intolerant of other dogs. In some cases, resident dogs had health issues or had become dog aggressive. For many participants who were no longer fostering, they stated they intend to foster in the future. These findings echo those reported by Maddie's Fund (2018).
Given the size of the sample and the fact that data collection took place in part during a global pandemic, the results related to the quantitative variables should not be interpreted as anything other than speculative. Many participants stopped responding after completing the open-ended questions. As a result, the sample size for the correlational analyses was quite small. In addition, the number of participants who were currently fostering far outpaced the number who were no longer doing so. With those caveats in mind, there are some insights that can be gleaned from the findings.
First, higher compassion satisfaction scores were found to be associated with lower burnout scores. In other words, participants who experienced higher levels of burnout as volunteers felt less satisfaction with their work as care providers. According to Stamm (1995), compassion satisfaction reflects pleasure taken in doing successful care work while burnout (one element of compassion fatigue) involves frustration and hopelessness associated with work.
Reflecting on the open-ended responses provides insight into this finding. The joys of fostering are centered on helping dogs become happy and finding them forever homes. That joy can be diminished by coping with the more frustrating and repetitive aspects of the work such as dog behavior problems or the uncertainty and ambiguity of finding quality adoptive homes.
In addition, burnout and secondary traumatic stress (the second component of compassion fatigue) were found to be positively correlated. In their open-ended responses, it was not uncommon for participants to mention frustration with seeing the same problems repeatedly or impatience with the foster/adoption process. Secondary traumatic stress refers to exposure to trauma experienced by others (Stamm, 1995). For people doing rescue work with dogs, avoiding exposure to traumatized, sick, and injured animals is virtually impossible. Many participants remarked on the pain they experienced seeing the horrible physical and emotional condition of some of the dogs they had cared for. Participants used words such as "heartbroken", "heartbreaking", "disgusting", and "horrifying" to describe the trauma experienced by their foster dogs. Several described the trauma they experience while caring for dogs going through heartworm treatments or struggling with other preventable diseases. Some described the death of animals in their care as particularly painful. It was also common for participants to express anger and revulsion directed at humans who had, through abuse and neglect, harmed animals in some cases irreparably. While the most common reasons for no longer fostering were not burnout and secondary traumatic stress, phrases associated with these concepts were clearly seen as downsides to the fostering experience that can have a cumulative negative effect.
As noted above, the results for the quantitative variables should be interpreted with care based on the sample size and the drop-out rate. Nevertheless, the findings provide some insights that may spur further research into risk and protective factors for foster care providers in animal rescue. Findings from research on secondary trauma and compassion fatigue in foster parents caring for human children echo those presented here. As noted by both Whitt-Woosley et al. (2020) and Hannah and Woolgar (2018), foster parents caring for children face routine exposures to trauma and that there are correlations between secondary traumatic stress and compassion fatigue, between those variables and intention to continue fostering, and between years doing the work and levels of secondary traumatic stress and compassion fatigue.
Compassion satisfaction was predicted by several aspects of both volunteer functions and self-care. Satisfaction with performing care work as a foster was positively associated with seeing volunteering as emotionally protective (e.g., working through one's own problems, escaping from one's troubles, relieving guilt, or feeing less lonely), as promoting understanding (e.g., learning by doing, learning to deal with different people, gaining new perspectives, or exploring one's strengths), and as a route to self-enhancement (e.g., feeling important, increasing self-esteem, feeling needed, and feeling connected). In terms of self-care, compassion satisfaction was positively associated with professional development efforts (e.g., staying current, maximizing time on enjoyable work, being involved in rescue activities, and engaging in activities that improve skills), maintaining life balance (e.g., spending time with friends and family, seeking comforting activities, and fostering social connections, and using cognitive strategies to cope effectively (e.g., taking proactive steps to manage challenges, monitoring feelings, and being mindful of triggers).
Burnout and secondary traumatic stress were found to be negatively associated with several aspects of self-care. In other words, these are behaviors that people scoring higher in compassion fatigue are doing less often. Burnout was negatively correlated with enhancing professional development, maintaining life balance, using cognitive strategies, and maintaining daily balance (e.g., avoiding over-commitments, taking time for relaxation, and taking breaks). Similarly, secondary traumatic stress was negatively correlated with professional development, life balance, and daily balance.
As noted by Figley and Roop (2006) and Hanrahan et al. (2018), care work with animals comes with both risk and protective factors. In their open-ended responses, particularly to the questions concerning the positive and negative aspects of fostering, respondents referred to the potentially damaging stressors involved in fostering as well as the positives of personal growth and social connections. For rescue leaders who are tasked with recruiting and maintaining a stable group of volunteer foster providers, it may be helpful to give special attention to both the risks inherent in fostering including burnout and secondary traumatic stress while training and interacting with volunteers. In addition, it may be beneficial to provide instruction in and opportunities for engaging in self-care strategies that are useful in mitigating the stress involved in care work.