Introduction

Over the past 20 years animal-assisted interventions (AAI), particularly those involving dogs, have become more common in classrooms. Evidence continues to accumulate, showing AAI to be an effective tool benefiting children’s mental health (Crossman et al., 2020), physical health (Beetz et al., 2011), cognitive performance (Gee et al., 2009, 2012), and literacy (le Roux et al., 2014; Levinson et al., 2017; Syrnyk et al., 2022). With these findings in mind, dog-assisted (aka canine-assisted) literacy supports continue to gain popularity in the classroom. When the global coronavirus moved teaching into homes, we wondered if such supports might also be applied in the home.

Interest in dog-assisted AAI for children’s literacy can be tracked back to the Reading Education Assistance Dogs (R.E.A.D.) program (Friesen & Delisle, 2012). An early examination of this program showed that 10 struggling early readers who read to a dog once a week over the course of the school year had improved reading scores (Martin, 2001). Many similar studies have since followed also showing promise but lacking methodological rigour as evidenced by Hall et al. (2016) who reviewed the corpus of literature in this area and found most research to be of low quality. However, findings from more recent empirically sound studies using, for example, control or comparison groups, randomization, and standardized assessments, supports the results of earlier, less objective studies and find that - AAIs with dogs do indeed have a positive impact on children’s reading performance. For example, le Roux et al. (2014) randomly assigned grade 3 children to read to a dog, an inanimate object, an adult, or no one for 10-weeks. They found that children in the dog-led condition outperformed children in the other conditions in terms of their reading comprehension, rate, and accuracy. In 2017 Levinson et al. compared the oral reading fluency of children in grades 2–5 before and after reading to a dog or a peer for 5 weeks, then swapped conditions for the same amount of time. They found that when reading to a dog the youngest children’s reading skills benefited the most. Most recently, research by Syrnyk et al. (2022) examined 24 struggling readers in grade 3 who were exposed to a 15-minute-long dog-assisted literacy intervention once a week for 8 weeks. Following completion, children were then assigned to an adult-assisted intervention or vice versa (i.e., half of the participants experienced the adult intervention first and half experienced the dog intervention first). Compared to the adult intervention, the AAI intervention yielded bigger gains in reading comprehension and oral reading. However, as the first study to examine both reading performance and behaviour in the same study, it also found that the order of type of support made a difference to social functioning, as children’s behaviour only improved in the dog intervention once they had first been exposed to the adult intervention.

Dog-assisted AAIs positive influence on literacy is often attributed to its ability to improve children’s motivation for reading by decreasing their stress about doing so. Current evidence-based theoretical models suggest that such AAIs trigger positive valence, decreasing the stress response, making room for intrinsic motivation (Hall et al., 2016). Indeed, evidence shows children’s intrinsic motivation for reading increases with AAI (see Hall et al., 2016). AAI has also been found to improve mood (McArthur & Syrnyk, 2018), reduce perceived stress (Pendry et al., 2020), increase our sense of security (Levine et al., 2013), and offer physiological benefits by reducing the stress response (Gacsi et al., 2013).

Although dog-assisted literacy supports are on the rise (Fine et al., 2019), primary and traditional literacy supports continue to occur in the close quarters of the classroom under the supervision of a teacher with expertise. This changed in March 2020 when COVID-19 moved school learning online. While the impact of this unprecedented event continues to be assessed, children’s literacy abilities were seen to have stagnated. For example, research from the United States reports that oral reading fluency was significantly set back particularly among younger (grade 2 and 3) American lower income readers (Domingue et al., 2021). Indeed, other reports suggest children who struggle with reading are now a year behind their peers and that online teaching methods may not be effective for struggling readers (Alphonso, 2020), while more recent research reports the opposite (Cruz et al., 2022). To help recoup literacy losses, it has been suggested that more formal at-home literacy activities (e.g., focusing on printed words and letters, linking sounds, etc.) may help close the literacy gap of struggling readers compounded by COVID-19 (Iswara & Iswara, 2021). Although this is promising, it is unknown whether many parents have the necessary skills, knowledge, or time to assist children with these activities.

The sudden shift to online learning due to the pandemic meant that parents had to learn overnight how to accommodate and support their children’s learning at home. For many parents this meant they had to adopt teaching roles and responsibilities while simultaneously adapting to changes within their own work lives - all while caring for their families. It is therefore unsurprising that parental stress increased significantly over the course of the pandemic (Adams et al., 2021). Indeed, research by Sonnenschein et al. (2021) found that parents who actively engaged in their children’s online learning experienced more stress than parents who did not. They also found that 75% of the stress parents experienced when engaged in online learning was due to feeling uncertain about how to help their child. Furthermore, studies examining children’s learning during the pandemic found many struggled to complete assigned tasks on their own (Lau & Lee, 2021; Stites et al., 2021). At the same time, school-age children’s frequency of use of digital learning devices (tablets, laptops, etc.) due to their association with at-home literacy activities exceeded professional recommendations (Sonnenschein & Stitest, 2021). Read et al. (2021) found that while shared reading between children and caregivers did not change during the height of the pandemic, they did find that children’s use of digital devices for reading purposes had increased while the number of adult reading partners decreased.

Given the challenges presented by the pandemic to parent’s well-being and children’s literacy, we wondered if (a) it would be possible to adapt an in-school dog-assisted literacy support AAI for in-home application, and (b) if such an adaptation might benefit children’s reading and relieve parental stress? As such, this pilot project aimed to provide a preliminary investigation into whether an AAI literacy support, that is typically seen in schools, could be adapted to the home, improve children’s reading, and decrease parental stress. To our knowledge this is the first study to investigate such a literacy support in the home.

Methods

We feel it is necessary to explain that this study was originally timetabled to intersect with COVID-19 at the beginning of the school year (Fall) in 2020 with the hopes that it might appeal to parents as a possible means of support. However, despite a robust and lengthened advertising period, recruitment proved unfruitful. As a result, the start date was pushed back to the summer of 2021. While this is historically the time of the ‘summer slide’ (i.e., there is a drop in academic performance over the summer holidays; Broakman et al., 2021) we aimed to attract more parents hoping they might simply have more availability at this time. In the end, the parents of 10 children consented to participate in this research (which had institutional research ethics board approval). This necessitated the removal of a control condition where we had planned to have a separate group where children read to their parents, not their dog. Of these dyads, 2 were omitted from the final analysis due to lack of participation. Participant dyads were recruited via a social media campaign that targeted parents seeking additional reading supports for their children. In Canada, each provincial government sets their own educational directives so only parents residing in Alberta were invited to participate. Furthermore, all participants were required to have access to the internet, a tablet, and a dog 2 years old or older. This age requirement was used in order to be consistent with the local animal-assisted not-for-profit (Pet Access League Society) that participates in animal-assisted initiatives throughout the community. PALS notes that younger dogs are deemed too immature, lacking in socialization and training, to participate in such animal-assisted activities (Fine, 2010).

Participants

Children (6 girls, 2 boys) had an average age of 7 years and 2 months (all were going into grade 3) and were from monolingual homes where English was the native language. Participating parents were all mothers, with an average age of 37 years. When asked to report their highest educational achievement, two mothers had obtained high school diplomas, four had an undergraduate degree, one had a master’s degree, and four had a college-level diploma. Regarding their employment status, three mothers worked full-time, three worked part-time, and four were unemployed. Eight of the mothers were married and two divorced. Eight parents completed various surveys before and after the study, with two parents not responding to the call for input at follow-up.

Measures

Using the online survey provider Qualtrics©, demographic information was collected from parents before and after the study about themselves and their child. Using a Likert scale (1 = strongly agree, 7 = strongly disagree) parents were asked to respond to a series of statements about their child’s reading abilities, their own perceptions of schooling during the pandemic, and their participation experience.

Parental stress was assessed before and after the study using the Perceived Stress Scale − 10 (PSS-10; Cohen et al., 1983). This self-report instrument consists of 10 Likert-rated statements (0 = never, 4 = very often) that assess the general degree to which an individual has perceived stress over the past month. With good internal reliability (Cronbach’s α ≥ 0.80; Lee, 2012), higher total scores on the PSS-10 are indicative of higher perceived stress.

Assessment of children’s reading proficiency is nuanced; often involving measures of, for example, oral reading accuracy, word comprehension, word recognition, phoneme segmentation, and phonological awareness. This is typically achieved through multifaceted measures and processes that can include children’s classroom teacher, parents, and sometimes others (e.g., school psychologists, speech and language pathologists, etc.). For this study, we were wary of overburdening parents and sought to gain a ‘soft’ estimate of children’s current reading proficiency by asking parents to assess their children’s sight word recognition using the Fry Sight Words list (FSW; Fry et al., 2003) and to provide us with their child’s last report card.

The FSW list consists of 1000 frequently occurring words that most commonly appear in reading materials for young readers who are native English speakers. The words are presented in ten groups with the first ‘100’ grouping being words typically seen in Grade 1. Sight word reading performance has been found to be an accurate means of identifying and monitoring struggling readers (Compton et al., 2010), although accuracy is diminished from grade 4 onwards (Tilstra et al., 2009). With this in mind, copies of the 2nd, 3rd, and 4th FSW lists were provided to each parent-child dyad. Dyads were blinded as to the ‘levels’ of the lists. Parents were instructed to have their child start with the first list reading aloud each word while parents followed along placing a tick next to each word their child was able to say easily, without hesitation. Parents were informed that they should not attempt to assist their child. Words were not ticked if the child could not say the word, mispronounced the word, or hesitated for more than 3 seconds before attempting to say the word. After attempting, in order, all three lists, parents tallied up the total amount of words ticked for each list and returned these to the researchers who converted this information into a percentage of words read by that child. Later, a teacher with expertise in literacy support used these percentages in conjunction with information (provided by parents) from the child’s latest report card regarding their literacy achievement to informally ascertain the child’s current reading level and assign this level to a user-unique Raz-Kids© profile. Raz-Kids is a for-profit website that caters to young readers by providing online access to, amongst other things, leveled readers and assessments with the intention of improving reading proficiency. While this provider did not contribute financially to this research, they did waive the subscription fee. Raz-Kids was chosen for its broad range of titles, and easy access. Each child’s reading level was manually assigned in Raz-Kids, however, before and during the study parents were encouraged to contact the researchers if they believed their child should be reassigned to a more or less difficult reading level. Once assigned to the appropriate level, Raz-Kids automatically populated a series of titles for parents and children to choose from. To gain a measure reading proficiency before and after dog-assisted reading, children’s starting/assigned reading level in Raz-Kids was later compared to the level of reader that children were accessing at the end of the study. Additionally, after reading an e-book, Raz-Kids automatically asked readers a set of questions testing their reading comprehension in areas such as cause and effect, main ideas and details, inferences and details, sequence events, story elements, and vocabulary, among others. Since the category of questions asked varied, presumably according to each book, we only analysed scores in those categories in which all children had been tested.

Procedure

After providing informed consent and completing the online data collection, participants were emailed instructions along with a link to an online video tutorial demonstrating how to engage with the support. These instructions also contained information about how to log-in and use the Raz-Kids website, to access reading-level appropriate assigned materials and track reading progress.

Every week over an eight-week-long period, parents were asked to use Raz-Kids to choose a book with their child. They then had their child listen to the book following along on their electronic device (i.e., tablet). Parents had their child read the book aloud to their family dog, with the parent present, but unobtrusive, to assist, if necessary. For example, if the child could not figure out a word on their own and encouragement was ineffective, the parent may direct the child to click on the word and Raz-Kids will read it out to them. Child and dog were sat comfortably near each other, in a manner that allowed the child to best engage with the dog (e.g., they could touch the dog, “show” the dog pictures). This procedure was demonstrated in a 4-minute-long video that illustrated how the sessions would occur. The video started with a text screen that said, “Before beginning the session have your good fit book on your tablet. Ensure your dog is ready to sit with your child. Settle into a comfortable and distraction free place. Your child is free to pet the dog while listening”. Following this, a woman, her son, and dog enter a room and sit on the floor. The woman narrates. She first explains to the listener to find a quiet room. Then she sits next to the child, with the dog seated on the other side of the child. She explains to the child and listener that first they will listen to the tablet read the book. Text then appears on the screen that says “After listening to the story your child will read aloud to the dog. Encourage your child but allow them to figure out the words. Your child is free to pet the dog while reading and to show pictures to the dog.” The video resumes and shows the woman now seated away from the child. The woman explains that the child will now read to their dog (child and dog are sat next to each other) and the child proceeds to do so. This is followed by another screen of text that says, “When your child finishes the story provide positive and encouraging feedback. Explain that the child will read this same book two or more times this week. After the practice reading the child will complete the comprehensive quiz”. The video resumes and the woman congratulates the child for reading and explains next steps to them and the listener. Note that the children were encouraged to read the chosen book two more times over the course of the week on their own. Based on the students’ progress with the book a new book would be chosen for the following week. Also, at the end of a dog-assisted reading session children were encouraged to thank the dog for being a helpful reading partner by interacting with the dog (giving it a treat, doing a trick, have a pet/cuddle). At the same time, parents were asked to provide a verbal, positive response to the child’s effort regardless of performance along with a hug/cuddle, etc. It was made clear to parents that there should be no tangible reward (e.g., money, treat) given to the child for completing the session. Overall, each reading to dog session was estimated to take approximately 20 min/week to conduct. At the end of the 8-week-long period parents were again asked to complete another similar online questionnaire that included the PSS-10.

Findings

Parent Surveys & Stress

Prior to engaging with the reading support, parents were asked to respond to a series of statements using Likert ratings (strongly disagree = 1, strongly agree = 7). When presented with the statement “My child is a confident reader” parents, on average (M = 4.9), somewhat agreed. They also somewhat agreed (M = 5) that reading made their child happy. Parents strongly agreed (M = 6.7) that they were comfortable helping their child with homework and agreed (M = 6.2) that there was a lot they could do to help their child here. They also agreed that there was a lot they could do to help their child understand the value of school (M = 5.7). When asked to reflect on their family dog, parents agreed (M = 6) that they were easy going and that their children were attached to their dog (M = 6.2). All eight parents reported themselves to read regularly with their children, reading an average of 2.3 books a day for approximately 20 min. Eight parents visited the library with their child for an average of 4 visits per month. When asked if they would describe themselves as active readers, 8 parents agreed while one disagreed and another was unsure - reporting reading for their own pleasure, an average of two books per month.

When asked about their family dog, families reported them to have an average age of 6.2 years with the majority being of mixed breed (n = 6) and the remainder belonging to sporting and working dog breeds. Eight parents reported that their dog had been part of the family for more than 2 years, with one family having a dog for less than a year. Families reported an average of 1.6 pets in their household.

When parents were asked about their child’s schooling experience in the past year (2020-21), 8 reported that their children continued to receive their education at school, when possible, and one was home-schooled. Of those that continued at school, four reported having to move to online learning at least once and three reported no online learning. Those who did experience online school reported that this occurred because their child had been identified as a close contact with a confirmed case of COVID-19. When asked if they were happy with online schooling, parents somewhat disagreed on average (M = 4.85). When asked if they thought their child enjoyed online school they also somewhat disagreed (M = 5). Parents were also asked whether they felt that their child had fallen behind this past year and if they should have received more learning resources. Responses to both statements indicated that parents were ambivalent here (M = 4.3 and M = 4, respectively). Five out of 7 parents stated that they planned on sending their child to school (i.e., not home schooling) in the upcoming school year.

After completing the study parents reported that they found it somewhat difficult (M = 3.4) to get their child to engage with the reading program. Parents somewhat agreed (M = 2.85) that their children enjoyed reading to their dog. They were unsure if they found it easy to engage with the reading support weekly (M = 4.4), although they did agree that using Raz books was easy to do (M = 1.85). Overall, these parents appeared to find the dog-assisted AAI literacy support somewhat challenging to engage with.

A paired t-test of the PSS data collected from parents found a significant difference (t(6) = 26.95, p = 0.000) in stress scores from the start (M = 28.57, SD = 2.76) to the end (M = 56.42, SD = 3.86) of the study showing that there was a large increase in parental stress over the duration of their engagement with this pilot project.

Reading

The eight dyads that engaged with this online reading platform logged in an average of 13 times over the course of the project, totalling an average time of 3 h and 55 min. In that time dyads listened to and read an average of 9 and 6.6 books, respectively. Five out of eight participants went up at least one reading level (in Raz) over the course of the study, while two were static and one went back a level. Although on average this group went up 0.875 levels over the 8-weeks of the study, a paired-test comparing their start to end reading levels only approached significance (p = 0.087).

Regarding Raz-kids reading comprehension scores, in the cause and effect category children were accurate, on average, 73% of the time. Similarly, they answered correctly when queried about the main ideas and details of stories 74% of the time. When asked to make inferences and draw conclusions about what they read children were correct 90% of the time. When asked to detail the sequence of events, children were accurate 80.5% of the time. Children correctly reported story elements 80% of the time and their vocabulary recall was accurate 83.5% of the time. Overall, this can be taken to indicate that over the duration of the study children’s overall reading comprehension was good to strong.

Discussion

This pilot study set out to see if dog-assisted reading supports could be tenably provided in the home. Our initial observations suggest that such dog-assisted reading supports appear to have potential to be administered in the home provided these are carefully designed and thoughtfully applied. This may be especially pertinent when it comes to parents who are supporting the application of such dog-assisted supports since they were found to experience greater stress over the duration of this pilot project.

Other research has found AAI in schools to be more beneficial than an adult-led literacy support in terms of improving reading abilities (Syrnyk et al., 2022), however, while our pilot research suggests that it is possible to adapt AAI literacy supports to the home, the impact of doing so remains unknown as children’s reading levels appeared to increase, but not significantly. Furthermore, engagement with this at-home literacy support did not, as hoped, decrease parent stress, but instead seemed to do the opposite. How might we explain these findings? Compared to teachers, parents are less likely to be detailed reporters of their children’s literacy (Syrnyk et al., 2022); part of the impetus for this project was the hope that in-home AAI literacy supports could take some of the burden off parents without expecting them to have greater attention to or knowledge of literacy development. Instructions were carefully designed and resources (e.g., videos) were plentiful but parents reported being uneasy about applying this support. Tellingly, rather than help to reduce parents stress levels it is plausible that the detailed nature and required commitment of the project only served to burden them more. As described in the Methods section, we attempted to provide a very thorough description of how the support should be used (including a video demonstration), however, it is possible that parents were unsure about exactly how to conduct the sessions, and this could have added to their stress. Future research should ensure to query parents about their confidence in using the support. Alternatively, and possibly an even more plausible explanation, increases in parental stress may have been due to timing. Data was collected over the summer, and the approach of the new school year (final assessments of stress were gathered just prior to the start of the new school year), while historically stressful to children, may also be a period of stress for parents. There is, to our knowledge, no evidence of such although we posit that this is likely given that many parents spend the summer months balancing employment with children who are out of school, along with the stress of family holidays, etc. An increase in stress leading up to the school year is likely to have been even higher than typical due to the pandemic. In the summer the cases of COVID-19 and public health restrictions declined, however, leading into the new school year, an impending new wave of cases were being reported at the time (e.g., Herring, 2021). The uncertainty of the pandemic will have contributed here with parents likely growing concerned about the possible risks of children returning to school with new protocols (e.g., mask wearing, etc.). Put simply, we surmise that parents were likely to have been burned out during the time of our study and rather than help ease their concerns, participation in this project likely became just another “thing to add to the list”. This would also explain our poor participant recruitment. As such, replication of this research with a control comparison during the school year would be useful to get a clearer picture of the plausibility and efficacy of an adapted at-home AAI. Pandemic aside, other limitations include this being a small pilot study only consisting of self-selecting participants. Although they experienced more stress, these parents signed up with the intention of bettering their children’s literacy despite their children not being struggling readers. As some research has shown AAI to be of greatest benefit to struggling readers (Syrnyk et al., 2022), this may have impacted the efficacy of the intervention. We believe that our participants may have been the students/families who actually needed supports the least. In fact, based on our survey, these families already read to their children quite often (reading an average of 2.3 books with their children per day) and their Raz-books reading comprehension scores were not at chance but were consistently strong overall. This leads us to believe that it is likely that these children were not necessarily struggling but perhaps were not meeting their parents’ expectations with reading.

Furthermore, we recognize that this case study had additional limitations not yet mentioned. One of the main hurdles for us was establishing a ‘soft’ baseline in terms of children’s reading proficiency. We believe our use of the Fry sight words in combination with latest student report cards, when used in conjunction by an experienced literacy teacher to make an estimation of reading ability, was effective for this study. Future studies may wish to apply additional measures to assess reading. In addition, the study was also limited by the functionality of the online book provider. Future research should include a comparative condition, such as reading to the parent alone. Finally, a better means of tracking reading progress at home is also warranted since we were reliant on parents and children’s ability to determine the appropriateness of the books, it is possible they may not have challenged themselves enough. While we recognize the limited data collected, we suggest that the real strength of this pilot study is not in its results, but in how it can guide future, more rigorous research. For example, very careful consideration of timing, attempts to target families that could most benefit from supports in the home, ensuring parents are clear on instructions and how children should engage with the animal, etc. Overall, we believe at-home AAI literacy supports should continue to be explored as a means of furthering children’s literacy development with careful consideration should be given to exactly when and how data is collected.