Experimental Trial Demonstrates Positive Effects of Equine Facilitated Learning on Child Social Competence

Human-Animal Interaction Bulletin
2013, Vol. 1, No. 1, 1-19

       Experimental Trial Demonstrates Positive Effects of Equine Facilitated
                      Learning on Child Social Competence

                                         Patricia Pendry and Stephanie Roeter
                                         Washington State University, Pullman, Washington

                Although equine facilitated programs have gained in popularity over the last decade, virtually
                nothing is known about the causal effects of equine facilitated interventions on human
                development and wellbeing. To address this gap in the literature, researchers conducted a
                randomized controlled trial to determine if an 11-week equine facilitated learning program
                enhanced 5th-8th grade children’s social competence. Children were recruited for program
                participation through referral by school counselors and recruitment in schools and community
                agencies. Researchers then randomly assigned 64 physically and mentally able children to an
                experimental group or waitlisted control group. Children in the experimental group participated in
                an 11-week equine facilitated learning program designed to increase social competence through a
                series of once-weekly, 90-minute sessions of individual, team, and group-focused equine
                facilitated activities, whereas children in the control group did not until 16 weeks later. Parents of
                children in both groups provided ratings of child social competence at the beginning and again at
                the end of the 11-week program. Results indicated significant group differences in mean levels of
                child social competence at posttest (p = .020), suggesting a moderate positive effect of program
                participation (d = .61). Waitlisted children in the control group who completed the program at a
                later date demonstrated significantly higher posttest levels of social competence after program
                completion (p = .000), compared to their own pretest scores. Using a lagged dependent variable
                approach, program effects were robust (p = .026) when simultaneously considering children’s
                pretest levels of social competence, age, gender, and referral status.

                Keywords: child development, equine facilitated intervention, experimental trial, HAI, social

                Patricia Pendry, Department of Human Development, Washington State University; Stephanie Roeter,
                Department of Human Development, Washington State University.
                The project described was supported by Grant Number 1R03HD066590-01, awarded to the first author, Dr.
                Patricia Pendry, from the Eunice Kennedy Shriver National Institute of Child Health & Human Development
                and Mars-Waltham. The content is solely the responsibility of the authors and does not necessarily represent
                the official views of the Eunice Kennedy Shriver National Institute of Child Health & Human Development,
                the National Institutes of Health, or Mars-Waltham. We acknowledge Sue Jacobson, program coordinator of
                People-Pet Partnership at the College of Veterinary Medicine and Dr. Phyllis Erdman, Associate Dean of the
                College of Education at Washington State University, who together designed the PATH to Success program.
                We also acknowledge Dan Bayly and Dr. Leanne Parker for providing counseling expertise and facilitation
                supervision, and PATH to Success Study volunteers, research assistants, participants, their parents and
                teachers. Correspondence concerning this article should be addressed to Dr. Patricia Pendry, 525 Johnson
                Tower, Department of Human Development, Washington State University, Pullman, WA, 99164. E-mail:


     Although there is growing evidence that             2004), speech and language ability (Macauley &
promotion of behavioral and social-emotional             Gutierrez, 2004), as well as more global
competence in children and youth is an effective         measures of functioning and adjustment
strategy to prevent mental, emotional, and               (Schultz, Remick-Barlow & Robbins, 2007).
behavioral disorders in adulthood (O’Connell,            Also, compared to more traditional classroom-
Boat & Warner, 2009), identifying and                    based counseling activities, equine assisted
expanding evidence-based prevention programs             interventions were more strongly associated with
that promote these competencies in children and          lower levels of internalizing and externalizing
youth constitutes a significant challenge                behavioral      problem     (Trotter, Chandler,
(Greenberg, Domitrovich & Bumbarger, 2000;               Goodwin-Bond, & Casey, 2008). Other
Huang et al., 2005). One type of approach that           researchers, in a study on therapeutic riding
has caught the attention of researchers and              targeting boys with emotional disturbances in a
mental health professionals is equine facilitated        residential treatment program, found no
learning, which combines the experience of               improvement was seen in depression or anxiety
equine human interaction with counseling-based           (Greenwald, 2001). While current evidence
processing skills to increase participants’              regarding the efficacy of equine assisted
awareness and control of their emotions,                 approaches is promising indeed, more causal
cognitions, and behaviors. Although equine               evidence is needed to demonstrate the effects of
facilitated programs have gained popularity              equine facilities programs.
over the last decade, virtually nothing is known
about their causal effects on human                      Causal Pathways
development. To help alleviate this gap in the               Although there is not one specific human
literature, researchers conducted a randomized           animal interaction theory that postulates how
controlled trial to determine the effects of an 11-      equine facilitated learning programs may affect
week equine facilitated learning program on the          child development, Bandura’s social cognitive
social competence of 5th-8th grade children.             theory (1986) provides a nice framework
     Although research on the broader field of           (Kruger & Serpell, 2010; Serpell, 2000).
equine assisted interventions is in its infancy, a       According to this perspective, children learn by
small body of pioneering work has shown                  observing others, with the environment,
promising results. Findings suggest significant          behavior, and cognition all as important factors
associations between involvement in equine               in influencing development. These three factors
assisted intervention and adjustment in various          are not static or independent; rather, they are all
socio-emotional, cognitive, and behavioral               reciprocal. For example, each behavior
domains, including depression (Bowers &                  witnessed can change a person's way of thinking
MacDonald, 2001), positive behavior (Chandler,           (cognition). Similarly, the environment one is
2005), behavioral and mood disorders (Mann &             raised in may influence later behavior.
Williams, 2002), self-esteem (Brown &                        This theory has implications for equine
Alexander, 1991; Gatty, 2001; Katcher &                  facilitated learning because equines possess
Wilkins, 1994), feelings of social acceptance            several intrinsic and physical attributes that
and peer popularity (MacDonald & Cappo,                  strongly influence the cognitions and
2003),         interpersonal       communication         environments of those interacting with them
(MacDonald, 2004), sensitivity towards others            (McGreevy & McLean, 2010). For example,
(Vidrine, Owen-Smith & Faulkner, 2002), anger            horses and mules are prey animals, characterized
(Kaiser, Spence, Lavergne & Vanden Bosch,                by a highly attuned awareness of their physical


surroundings and a strong, instinctual inclination       This relates nicely to equine facilitated activities,
for a fight-or-flight response. This characteristic      as participants are encouraged to take on a
necessitates that program participants become            leadership role to facilitate positive interactions
more self-aware of their behavior, thoughts,             with equines, which rely heavily on the equine’s
emotional intent, and verbal and non-verbal              instinctual expectations about the natural
communication to ensure their personal safety,           hierarchy of herd dynamics, leading them to
and to enhance their ability to communicate with         oscillate between seeking and/ or assuming
the equine. By interacting with equines in the           leadership, or accepting it from others
context of horsemanship activities, children             (Budiansky, 1997).
learn to provide and recognize non-verbal                     During in-hand horsemanship activities,
communication cues to their equine partner (e.g.,        children are encouraged to engage in behaviors
pressure, release, posture, intent, physiological        that assert leadership to the equine by moving
arousal), which may provide frequent                     the horse’s body in a purposeful way through the
opportunity to reflect on the meaning of their           use of progressive shaping techniques, or
own behavior and intent. Since equines provide           “…gradually increasing a behavior/pressure to
immediate, meaningful feedback - pinning ears,           elicit a specific behavior from the equine”
swishing tail, licking lips, blinking - children         (McGreevy & McLean, 2007, p. 114), and
simultaneously gain experience in observing and          negative reinforcement to reward compliant
interpreting non-verbal cues. As children – with         responses from the horse. During these
the help from program facilitators - attempt to          interactions children are frequently reminded by
manage and direct their own behavior to evoke            program facilitators who self-management,
desirable responses from the horse (e.g., licking        fostering a give-and-take relationship, quick
lips, blinking, resting foot), rather than less          decision making, and optimistic thinking are
desirable responses (e.g., swishing tail, pinning        important skills that facilitate successful
ears, tuning out), reciprocal interactions between       communication and thus completion of each
the child’s behavior and cognitions take on an           task. Through these interactions with the horse,
important role. Using human equine interactions          children thus learn to provide and interpret clear,
as an analogy, program facilitators encourage            non-verbal cues. In sum, equine characteristics
children to reflect on their behavior, thoughts,         may have a significant influence on the nature of
feelings and communication skills and apply              interactions between participants and horses in
these to communication and collaboration with a          ways that promote children’s social and
human partner. Practicing of pro-social                  behavioral competencies through the practice of
behaviors, reflecting on thoughts and feelings,          pro-social behavior, as well as reflection on
and feedback – from the equine and facilitator -         participants’ cognitions, behaviors, emotion and
may thus improve self and social awareness,              intent.
personal responsibility, goal-directed behavior,
and communication skills.                                Description of Current Study and Hypotheses
     Social role-activity theory (Brickel, 1982)             The overall objective of the current study
also provides a useful perspective to explain            was to conduct a randomized controlled trial to
effects of horse-human interactions on child             determine the efficacy of an 11 week, equine
development. Role theory proposes that when an           facilitated learning program in improving the
individual assumes a new positive role (e.g.             social competence of 5th-8th grade children.
team leader), positive behavioral change may             Social competence is the ability of the child to
result (Lemon, Bengston & Peterson, 1972).               successfully interact with other children and


adults in a way that demonstrates an awareness          competence are also more likely to experience
of and ability to manage emotions in an age- and        difficulty regulating negative emotion and
contextually appropriate manner. Social                 arousal, which can interfere with initiating and
competence is comprised of social, emotional,           maintaining positive peer interactions, leading to
and cognitive skills and behaviors that enhance         peer rejection (Dodge & Somberg, 1987).
social adaptation such as social awareness, self-       Relationship problems with peers and adults,
awareness, goal-directed behavior, relationship         difficulty with emotion regulation, attention,
skills, personal responsibility, and control of         behavior, and challenges with learning and
behavior. The main research question was: Does          achievement contribute to the development of
participation in an 11-week equine facilitated          mental health issues and poor child wellbeing.
learning program increase children’s social             As such, enhancing child social competencies
competence and/or components thereof?                   that support the development of self-regulation,
     The following hypotheses guided our                academic skills, and peer social skills is
analyses. The first hypothesis stated that              considered a suitable target for preventive
children randomly assigned to participate in an         intervention programs (Greenberg, Domitrovich,
11-week equine facilitated learning program             & Bumbarger, 2000).
would have higher levels of social competence
at posttest, compared to children assigned to a         Method
waitlisted condition. The second hypothesis
stated that children in the waitlisted group would      The Program
demonstrate significant within-person gains in               The program under study was PATH to
social competence after program participation           Success1 conducted at a Premier Accredited
(i.e., after the experimental group completed the       Center of the Professional Association of
program). The third hypothesis stated that the          Therapeutic       Horsemanship,     International2
positive effects of program participation on            (PATH, Intl.) in a university setting. The stated
posttest levels of child social competence would        goal of the program under study is to enhance
be independent of children’s pretest levels of          child social competence through an 11-week
social competence, gender, age, and referral            session of once-weekly, 90-minute lessons of
status.                                                 individual, team, and group-focused equine
     Social competence was the main outcome of          facilitated activities. The program was designed,
interest to program designers and researchers,          implemented, and consistently supervised by a
because it is considered a central domain of            PATH, Intl. certified instructor in equine
child development that plays a critical role in         assisted mental health, and a doctoral level
later academic achievement, mental health, and          therapist. The curriculum contained standardized
overall wellbeing (Shonkoff & Phillips, 2000).          lessons with weekly goals and step-by-step
Lower social competence is positively                   instructions and procedures for each activity,
associated with the development of anti-social          which were piloted for 3 years before this study
behavior (Sørlie, Hagen, & Ogden, 2008),                on its effects was conducted. The program was
conduct problems and depression (Rockhill,              implemented by a team of volunteer horse
Vander Stoep, McCauley, & Katon, 2009),
substance use (Griffin, Epstein, Botvin, &                PATH to Success was designed by Sue Jacobson, coordinator of
                                                        People-Pet Partnership at the College of Veterinary Medicine and
Spoth, 2001) and lower academic competence              Dr. Phyllis Erdman, Associate Dean of the College of Education,
and performance (Wentzel, 1991; Sørlie &                both at Washington State University.
                                                          The professional association was known as the North American
Nordhal, 1998). Children with lower social              Riding for the Handicapped Association (NARHA) until its
                                                        renaming in the Spring of 2011.


specialists and facilitators, that included PATH                                 principles of equitation science and natural
certified instructors, undergraduate students in                                 horsemanship (McGreevy, 2007; McGreevy &
child-development, education, and animal                                         McLean, 2007; 2010) - featured a combination
science, professional counseling psychologists,                                  of mounted and un-mounted activities and
and graduate-level counseling students. Program                                  horse-human interactions, including observation
staff received extensive training through printed                                of equine behavior, engagement in equine
lessons, a handbook, hands-on program training                                   management       (e.g.,    grooming),      in-hand
with horses and child actors, as well as training                                horsemanship activities, some riding, and
in theoretical rationale of the program, principles                              personal and group reflection activities. A
of cognitive behavioral counseling, and child                                    description of weekly program objectives and a
development through video and slide                                              selection of sample activities are described in
presentations. Weekly activities - based on                                      Table 1.

Table 1
Outline of Lesson Objectives by Week

Week                     Lesson Objective                                                              Sample Activity
1                        Basic safety: Meet horses and staff                                           Observing horse behavior and herd
2                        Respect: Self, others and horses                                              Moving horses using 4 phases of
                                                                                                       direct or indirect pressure3
3                        Communication: Verbal and non-verbal                                          Leading horses, interpreting horse
                                                                                                       body language
4                        Leadership: Assertive and aggressive cues                                     Driving4 activity using body
                                                                                                       language and phases
5                        Trust: Coping with perceptions of stress                                      Riding and leading

6                        Boundaries                                                                    Driving activity using indirect
7                        Overcoming challenges and building confidence                                 Desensitizing5 horses

8                        Enhancing self-regulation and relaxation                                      Horse massage6, riding

9                        Prepare for parents/visitors day                                              Incorporating horsemanship skills
                                                                                                       for team challenge
10                       Parents/visitors day                                                          Participants ‘teach’ parents
                                                                                                       horsemanship skills
11                       Program wrap up                                                               Obstacle course, riding and

  Pressure refers to a directly physical or implied signal to the horse with fingers/hand, or leg until the desired behavior or the equine is fulfilled
(e.g., moving away from the pressure).
  Driving refers to signaling the horse to move forward by pointing toward the desired space, and applying pressure behind the equine’s shoulder
to ‘drive’ them forward of the equestrian.
  Desensitizing refers to exercises to condition the equine to ignore a stimulus or object, thus avoiding ‘spooking.’
  Horse Massage refers to an exercise taught to participants in which they stroke the horse with slow, rhythmic strokes down the equines’ body to
‘relax’ the equine before riding.


    Children in the experimental group were             counselors7, distribution of flyers and
assigned to the same facilitation team for the          advertisements, and information meetings in
entire 11-week session that included one equine,        schools and the community. Criteria for program
two child participants, an equine specialist, and       participation were that 1) parents and children
facilitator. Each day, a total of eight child           had to be able to communicate effectively in
participants - divided across four equine teams         English, 2) the child did not have a serious
headed by four experienced equines –                    physical or mental disability and 3) the child had
participated. The program was conducted four            to attend the 5th through – 8th grade. Recruitment
times a week, serving 32 children each week per         for the study occurred at the informational
11-week session. The weekly lessons were                meetings, during which parents and children
conducted in the afternoons on weekdays and             received information from the research team
included transportation of participants from            about the study, its screening, assessment, and
school to the program site and back immediately         assignment procedures as well as information
following their regular school day. Children who        about the process of consent and assent8.
were waitlisted participated after children in the      Interested parents were asked to complete a
experimental group completed the program.               standardized measure of child social competence
                                                        for screening and sample selection purposes, for
Procedures                                              which they were paid five dollars (see measures
    The study was approved by the University            section for description of social competence
Committee on Research Involving Human                   measure). Children in each grade were
Subjects at Washington State University. Equine         subsequently rank-ordered by their social
welfare of horses in the program was carefully          competence score with the aim of recruiting
monitored under the university’s Institutional          approximately equal numbers of boys and girls
Animal Care and Use Committee (IACUC)                   per grade, as well as selecting children with
regulations and campus veterinarian. In addition,       relatively lower social competence scores.
equines were evaluated daily by a PATH, Intl.           Participation in the study was voluntary and
certified instructor for physical, mental, and          children could attend the program without
emotional well-being before, during and after           enrolling in the study.
each session. Although equine participation in               Selected study participants were randomly
the program never elicited such behavior,               assigned to a treatment group or wait-listed
program policy required that equines would be           control group. Children assigned to the
immediately       released    from      program         experimental        group      started    program
participation if they were to display stress            participation 2 weeks later, whereas children
behavior to ensure their wellbeing. In addition,        assigned to the waitlisted control group were
any inappropriate or potentially harmful
behavior of humans interacting with equines             7
                                                          Children referred by school counselors were children who were
                                                        receiving school counseling services for academic and/or
required immediate intervention to prevent              behavioral adjustment issues, or those whose parents had consulted
undue distress by the equine.                           with school counseling staff about the presence of stress in the
    Program participants were recruited by              8
                                                          Using a slide presentation, parents were given information about
program staff through referral by school                the consent and assent process, along with hard copies of both
                                                        forms, and a typed script of the assent procedures, which described
                                                        study procedures and children’s rights using developmentally
                                                        appropriate language (i.e., children 11 and younger, children 12
                                                        and older). The Principal Investigator (PI) assented all child
                                                        participants in person.


offered program participation in the following           schools in a rural university town in the Pacific
session – 16 weeks later. Parents of both groups         Northwest area of the United States. Sixty-four
provided ratings of child social competence              children were selected (N boys = 37; N girls = 27;
again at the posttest (i.e., at the end of the 11-       M age = 10.93 yrs) for study participation and
week program). Parents of waitlisted children            random assignment to an experimental group (N
were asked to report on their child’s social             = 33) or waitlisted control condition (N = 31).
competence a third time, after their child               There were no significant differences between
completed their 11-week session.                         participants’ age (p = .988), gender (p = .143),
                                                         referral status (p = .300), grade level (p = .874),
Sample Recruitment, Screening and Assignment             or pretest social competence (p =.599) by group
    In the first year of this two-year study,            status (e.g., between experimental or waitlisted).
program staff recruited a total of 78 interested         Sample characteristics are reported in Table 2.
children and their families from 10 different

Table 2
Sample Descriptive and T-Tests Comparing Group Differences
                         Assigned (N = 64)       Treated (N = 54)
                         M (SD)                  M (SD)
Social Competence
Age                      10.95 (1.06)            10.93 (1.08)
Male                     26                      25
Female                   38                      29
Referred                 10                      8
         5th             30                      26
         6th             21                      18
         7th             5                       3
         8               8                       7

    Sample enrollment, selection, randomization,
follow up and analyses counts are described in a         failure to complete the posttest assessment. Of
flow diagram (Figure 1). A total of 54                   all children originally assigned to the waitlist
participants completed data collection at pretest        control condition, a total of 20 eventually
and posttest. Attrition in the experimental group        participated in the active program, which was
was due to two active withdrawals immediately            conducted 4 weeks after children in the
following assignment, one participant moving             experimental group completed the program.
out of the area after 6 weeks of program                 There were no significant group differences
participation, and one active withdrawal in week         (comparing waitlisted or experimental group) by
four due to allergy issues. Attrition in the control     age (p = .637), referral status (p = .581), grade
group occurred due to active withdrawal by               level (p = .639), social competence (p = .548), or
three participants following random assignment,          gender (p = .089) between participants who
and three passive withdrawals as evidenced by            completed the study and those who withdrew.


Figure 1

Flow Diagram describing sample enrollment, selection, randomization, follow up and analyses

                                            Assessed for eligibility (N = 78)

                                                                           Excluded (N = 14)
                                                                            Not meeting selection criteria
                                                                             (N= 14)

                                                  Randomized (N = 64)

 Allocated to Experimental Condition (N =                              Allocated to Control Condition (N =
 33)                                                                   31)
  Received allocated intervention (N = 31)                             Received allocated intervention

 Lost to follow-up                                                     Lost to follow-up
  Moved away from area (N = 1)                                         Failure to complete post-test
  Allergy issues (N = 1)                                                assignment (N=3)

 Analyzed                                                              Analyzed
  Included based on original data (N = 29)                             Included based on original data (N =
  Included based on pooled data in intent-to-                           25)
   treat analyses (N = 33)                                              Included based on pooled data in
                                                                         intent-to-treat analyses (N = 31)


Measures                                                     T-scores were used as dependent variables
    As part of the screening and pre-test               to analyze group differences and estimate
measurement process, parents were asked to              regression models. T-scores were also used to
complete the parent report form of the Devereux         categorize children into 3 groups ranging from
Student Strength Assessment (DESSA), a                  low to high (0-40 = needs instruction, 41-59 =
measure of child social competence (LeBuffe,            typical, and 60-72 = strength) to enable
Shapiro & Naglieri, 2009). The DESSA is a               interpretation of causal changes for intervention
behavior rating scale that assesses the social-         and prevention purposes. The DESSA has
emotional competencies that serve as protective         excellent internal reliability and shows
factors for children in kindergarten through            significant, moderate-to-high correlations with
eighth grade. In keeping with the prevention            widely used measures with good psychometric
perspective of the program designers, an                properties (LeBuffe, Shapiro & Naglieri, 2009),
assessment tool was chosen that was strength-           including the BERS–2 (Epstein, 2004) and the
based and did not assess risk factors or                Behavior Assessment System for Children–2
maladaptive behaviors. The measure contains             (Reynolds & Kamphaus, 2004). Completion of
72-items asking respondents to indicate how             the DESSA by parents occurred at the beginning
often various child behaviors occurred based on         (α1 = .98) and end (α2 = .98), of the first 11-
a 5 point Likert scale ranging from 0 (never), 1        week program, and parents of waitlisted children
(rarely), 2 (occasionally), 3 (frequently), to 4        reported once more (posttest2) more at the end
(very frequently) over the past 4 weeks. The            of the second 11-week program (α3 = .98).
DESSA features a social competence composite
score derived from 8 subscales. Examples of             Results
items and subscales include Optimistic Thinking
(α = .87; e.g., How often did the child say good        Descriptive analyses
things about the future?), Self-Management (α =               Descriptive analyses on observed data
.86; e.g., How often did the child focus on a task      revealed that overall pretest social competence
despite a problem or distraction?), Goal-               levels of children in the study group were in the
Directed Behavior (α = .89; e.g., How often did         normal range (M = 41.41; SD = 10.16). The
the child keep trying when unsuccessful?), Self-        distribution in the sample suggests that the goal
Awareness (α = .82; e.g., How often did the             of recruiting children with relatively lower
child show an awareness of his/her personal             levels of social competence was met, as
strengths?), Social-Awareness (α = .81; e.g.,           approximately half of the study sample was
How often did the child respect another person’s        classified as needing instruction in social
opinion?), Personal Responsibility (α = .87; e.g.,      competence (N needs instruction = 30, N typical = 31, N
How often did the child prepare for school,             strengths = 3). A series of one-way ANOVAs
activities, or upcoming events?), Decision              revealed no statistically significant differences in
Making (α = .91; e.g., How often did the child          pretest levels of child social competence
accept responsibility for what she/he did?), and        between children in the control and experimental
Relationship Skills (α = .93; e.g., How often did       group composites scores (M control = 40.71 (9.55),
the child greet a person in polite way?).               Min = 28, Max = 69; M experimental = 42.06
DESSA (LeBuffe, Shapiro & Naglieri, 2009)               (10.81), Min = 28, Max = 72) or any subscale
raw scores were converted to T-scores using the         scores (Table 3). Also, there was no evidence to
DESSA coding procedure.                                 suggest that the distribution of social
                                                        competence classification groups (i.e., needing


instruction, typical or strength) was different (p          subjects variable is the posttest, and the
= .649) between the control (N needs instruction = 13,      between-subjects variable constitutes the
N typical = 7, N strengths = 1) and experimental            condition (control versus experimental group).
group control (N needs instruction = 17, N typical = 14,    Results support the first hypothesis (Table 3) by
N strengths = 2). There were significant differences        demonstrating a significant treatment effect of
in pretest child social competence by gender                program participation as illustrated by
showing higher social competence scores for                 significant   group       differences    between
girls (p = .010; M boys = 37.63 (9.03); M girls =           experimental and control groups’ social
44.16 (10.16)) and referral status (p = .009; M             competence composite scores (p = .020). Based
referred = 34.45 (6.85); M unreferred = 43.08 (10.14)),     on calculations using pooled variances, this
although boys and girls were equally likely to be           represents an effect size of d = .61, which is a
referred (p = .851).                                        moderate effect (Cohen, 1988). Researchers
       To determine whether participation in the            explored which aspects of social competence
program significantly affected child social                 appeared most significantly improved by
competence we conducted a series of intent-to-              conducting a series of ANOVAs comparing T-
treat analyses. This is a strategy for the analysis         scores on all subscales by treatment group.
of randomized controlled trials that prevents               Using Tukey’s method to correct for multiple
overestimation of treatment effects by                      comparisons, results revealed that self-
comparing participants in the groups to which               awareness (p = .000) and self-management (p =
they were originally randomly assigned,                     .001) had most significantly - and positively -
regardless of whether they subsequently                     responded to the program participation (Table
withdrew or deviated from the protocol (Fisher              3).
et al., 1990). To accommodate this approach,
missing data on posttest assessments on social              Hypothesis 2
competence (i.e., subscale and composite T-                      Next, we tested the second hypothesis which
scores of child social competence measures at               stated that program participation by waitlisted
posttest) for 10 participants were thus imputed             control group children will increase their levels
using five imputations according to multiple                of social competence. Using t-tests, we
imputation procedures (Rubin, 1987; Rubin,                  compared social competence scores of waitlisted
1996). Doing so retains the full sample for                 children obtained after they completed the 11-
calculating intention-to-treat estimates and,               week session to their own pretest scores
under appropriate assumptions, should yield                 obtained during program screening. We found
unbiased estimates of the intention to treat. The           that waitlisted children in the control group who
posttest results described in the next section thus         completed the program at a later date
reflect pooled estimates across five imputed                demonstrated significantly higher posttest levels
datasets that were generated.                               of social competence (M = 43.71; SD = 11.07)
                                                            after program completion compared to their own
Hypothesis 1                                                pretest scores (p = .000).
   To test the first hypothesis - participation in
the program will improve child social                       Hypothesis 3
competence - researchers conducted a one-way                    Testing the third hypothesis- positive effects
ANOVA comparing child T-scores on social                    of program participation on posttest levels of
competence measured at posttest between                     child social competence are independent of
experimental and control group. The within-

10 | H A I B

Table 3
Experimental Effects of Equine Facilitated Learning Program on Child Social Competence

Pretest                      Control              Experimental          ANOVA
                             M T-scores (SD)      M T-scores (SD)       p-value

Social Competence COMP             40.71 (9.55)   42.06 (10.81)         .599
Personal Responsibility            42.31 (8.90)   48.52 (9.12)          .441
Optimistic Thinking                43.54 (11.07)  47.72 (11.23)         .957
Goal Directed Behavior             37.85 (8.77)   43.64 (8.60)          .572
Social Awareness                   43.88 (11.40)  48.00 (10.13)         .873
Decision Making                    42.85 (10.99)  50.00 (10.60)         .864
Relationship Skills                43.92 (11.00)  49.25 (10.27)         .830
Self Awareness                     41.92 (10.40)  50.65 (9.07)          .823
Self Management                    41.65 (11.75)  51.07 (9.10)          .168
Posttest                           Control        Experimental          ANOVA
                                   T-Scores       T-Scores              p-values
Social Competence COMP             41.45 (10.71)  47.85 (10.41)         .020*
Personal Responsibility            42.72 (8.60)   48.34 (8.87)          .009**
Optimistic Thinking                43.59 (10.50)  47.14 (11.30)         .317
Goal-Directed Behavior             37.89 (8.38)   43.39 (8.15)          .005**
Social Awareness                   44.28 (11.70)  48.30 (10.31)         .227
Decision Making                    43.93 (10.70)  49.38 (10.01)         .039*
Relationship Skills                43.84 (10.39)  48.99 (9.90)          .041*
Self Awareness                     42.13 (11.04)  50.98 (8.87)          .001**
Self Management                    42.33 (11.56)  51.35 (8.84)          .000***
Note.*p< .05. **p

Table 4
Lagged Dependent Variable Model Predicting Posttest Social Competence
Variable                 Unstandardized B                  SE         Beta     p-value
Model 1 (Original Data)
Constant                          25.39                    12.85               .054
Whether Experimental              4.84                     .220       2.015    .021*
Social Competence Time 1          .686                     .108       .666     .000***
Child Gender                      -.111                    2.15       -.041    .673
Child Age                         -1.00                    .959       -1.00    .301
Whether Referred                  -2.22                    3.00       -.073    .463
Model 2 (Imputed Data/Pooled Estimates)
Constant                          27.594                   12.55               .030
Whether Experimental              4.97                     2.15       2.31     .026*
Social Competence Time 1          .662                     .103       6.43     .000***
Child Gender                      -1.38                    2.07       -.666    .507
Child Age                         -1.08                    .933       -1.16    .249
Whether Referred                  -2.36                    2.63       -.898    .370
    Note.*p< .05. **p

Discussion                                              equine facilitated learning may thus have
                                                        produced multiple benefits that pertain to
     The current study employed a randomized            youths’ personal, social, and academic lives.
controlled experimental trial to examine the            Second, these results represent a novel
causal effects of participation in an 11-week           contribution to the literature on equine assisted
equine facilitated learning program on child            intervention, as there is no prior work that has
social competence of 5th-8th grade children. This       employed an experimental design to examine
study responds to a call in the literature of           causal effects of equine facilitated learning
human animal interaction to move the field of           programs on child social competence. Third,
animal assisted intervention out of its ‘fringe’        results echo findings from prior correlational,
status and establishing credibility by means of         anecdotal, and case study evidence, which
carefully controlled trials and valid efficacy          suggest significant positive associations between
studies (Kruger & Serpell, 2010). Results               participation in equine facilitated programs and
indicated support for our first hypothesis              various aspects of adjustment and wellbeing
suggesting a moderate, positive effect (d = .61; p      (Bowers & MacDonald, 2001; Brown &
= .020) on the social competence of 5th-8th grade       Alexander, 1991; Chandler, 2005; Gatty, 2001;
children in response to program participation.          Kaiser et al., 2004; Katcher & Wilkins, 1994;
Positive effects were observed on various               Klontz, Bivens & Klontz, 2007; Macauley &
aspects of social competence, including                 Gutierrez, 2004: MacDonald, 2004; MacDonald
improvements in children’s self-awareness, self-        & Cappo, 2003; Mann & Williams, 2002;
management, personal responsibility, decision           Schultz, Remick-Barlow, & Robbins, 2007;
making, goal directed behavior, and relationship        Trotter, 2005; Trotter et al., 2008; Vidrine,
skills. In addition, children who were originally       Owen-Smith, & Faulkner, 2002). Fourth, the
in the waitlisted control group showed similar          results give greater credence to the claim of
gains in social competence after they completed         therapeutic      horsemanship       professionals,
the program, suggesting support for the second          participants, and parents who have reported and
hypothesis, which predicted significant within-         experienced significant positive effects of equine
subject improvements for waitlisted children.           facilitated activities firsthand. Faced with
Support for the third hypothesis demonstrated           skepticism about the efficacy of equine
that treatment effects were independent of              facilitated programs by potential funders and
children’s pretest social competence levels,            third party payers, therapeutic professionals and
referral status, gender, and age.                       clients can now point to causal evidence. This
     The results are important for the following        may not only increase the public’s confidence in
reasons. First, the fact that participation in an       equine programs’ ability to positively affect
equine facilitated program conducted as an after-       child development, but also translate into
school program effectively enhanced children’s          increased structural support to increase
social competence is exciting, since research           accessibility to such programs. Equine
indicates that promotion of personal and social         facilitated learning programs may serve as
skills in after-school settings are also known to       viable alternatives to after-school programs
improve youths’ feelings of self-confidence and         focusing on athletic or academic achievement, to
self-esteem, school bonding (positive feelings          provide after - school opportunities to children
and attitudes toward school), positive social           with different interests and needs. Last, the
behaviors, school grades, and achievement test          results directly inform research questions for
scores (Durlak & Weissberg, 2007). As such,             future research, particularly research focused on

13 | H A I B

understanding of the mechanisms that may                internal distress. In addition to affecting peer
underlie these positive effects.                        relations, these tendencies lead children to be
                                                        more susceptible to developing some form of
Interpretation of Results                               psychopathology. Thus, social cognitions,
     Although this study did not examine the            especially regarding self-efficacy, personal
pathways referred to in the introduction, we            agency, self-control, and cognitive distortion, as
discuss several processes we think are important        well as tools to help children regulate their
to consider in obtaining positive program               physiological arousal and emotions, may have
effects. First, while we believe that equines           enhanced social competence through stress
possess key characteristics (e.g., prey animal,         reduction. Examples of program activities that
herd animal) that influence the behavior of             targeted stress reduction included silent
humans interacting with them, we do not believe         grooming, rhythmic stroking of the horse,
that equines’ innate characteristics alone              encouraging slow breathing, muscle relaxation
fostered gains in social competence. Program            and rhythmicity during mounted activity, as well
facilitators helped children recognize analogies        as encouraging optimistic thinking, reflecting on
between horse - human interactions to those             emotions experienced during the lesson, and
between humans in various social contexts (e.g.,        utilizing cognitive, behavioral, and physiological
school, family). For example, during one of the         coping strategies towards adaptive self-
horse activities, children were taught to use           management.
gradually increasing amount of finger pressure               Third, apart from equine facilitation,
on the horse - hair, skin, muscle, bone - as a way      children may have benefitted from experiencing
of ‘asking’ the horse to move its front or hind         an optimal social niche that enabled them to
quarters. When the child started the activity           improve on social skills important to success in
using ‘muscle’ instead of ‘hair’, facilitators          other, more complex social settings. For
asked the child to reflect on whether he or she         example, large, unstructured peer group settings
would use the analogous behavior - shove or             (e.g., recess) can be particularly difficult
push - to ‘ask’ another person to move out of           situations for children with poorer social
the way, or if he or she may have used a                competencies, who may have difficulty initiating
different communication tool to express the             and maintaining normative peer interactions.
request. By engaging children in practice and           The smaller, more structured peer interaction
reflection within the context of a horsemanship         setting may have provided an appropriate and
activity, children may have been more open to           positive learning environment as equine team
reflect on their behavior during interactions with      members (horse specialists and facilitators)
non-equine partners.                                    provided support to guide positive peer
     Second, embedded in lessons targeting              interactions, friendship development, and
children’s goal-directed behavior and social            modeling of social competence. In addition,
skills were several activities designed to reduce       given that the sample was not exclusively
‘stress.’ The rationale for including such              targeting referred children, program participants
activities was that children who experience             experienced normative peers as pro-social
stress - through exposure to stressful life events      models. Participation in the program may have
or by perceiving common events as stressful -           provided less competent children with
tend to have distorted perceptions of the degree        opportunities to interact with more competent
of threat present in social situations. They also       social partners, leading to positive social
lack effective coping skills to manage their            interactions with peers. These serve as models

14 | H A I B

for subsequent interactions outside the context          been less willing to interact with horses, or
of the intervention. This small, structured format       whose parents were less supportive of their
may have also counteracted the alienation                child’s program participation. While we
experienced by children who are receiving                informally observed individual differences in the
special education services or targeted                   extent to which children - and parents -
interventions that may make them vulnerable to           expressed enthusiasm about the efficacy of
labeling and the influence of less competent             equine assisted activities, we did not measure or
peers. In sum, we suggest that engaging in               model these differences. Hence, this limitation
equine facilitated activities enhance social             needs to be considered when determining
competence when conducted in an appropriate              whether equine facilitated growth and learning
learning environment that provides optimal               programs can be used as universal prevention
opportunities for promoting self-efficacy,               programs. Similarly, because of the relatively
performance accomplishment, and personal                 short-term nature of the intervention and lack of
agency (Kruger & Serpell, 2010).                         longitudinal data at this time, researchers could
                                                         not ascertain whether the program was of
Strengths and Limitations                                sufficient duration or intensity to potentially
    There are several strengths and limitations to       alter developmental pathways of children at
consider when interpreting results of this study.        significant risk for developing mental health
The main limitation of the study is that the             problems. Last, equine-facilitated activities
results are based on parents’ perception of child        involve a great deal of human-human
social competence. While parents tend to be              interaction, which the participants in the control
appropriate reporters of children’s social               group did not receive. The design of the study
competencies as they intimately know the child,          would have benefitted from an additional control
they are, by definition, not blind to their child’s      group to compare effects on social competence
treatment status. As such, it is possible that the       from the equine facilitated activities to the
results reflect parents’ positive expectations           effects on social competence from interventions
about the efficacy of the program. This                  that do not involve interactions with equines.
limitation would have been of less concern if                 One of the main strengths of this study and
teacher reports on child social competence had           its findings is that the results are based on an
been included, assuming that researchers would           experimental design, which allows researchers
have been more successful in keeping the child’s         to make causal inferences. This constitutes a
treatment status blind to teachers. Teacher              novel contribution to research on human-animal
reports were not collected as several teachers           interaction and child development generally, and
expressed concern about the fact that measures           equine facilitated interventions specifically.
were collected at the beginning of the school            Second, results yielded a moderate effect size
year, when teachers had not yet interacted much          that was statistically significant despite the small
with the children.                                       sample size and being based on a calculation
    Second, although participation in the                method using pooled variances. In addition,
intervention was offered to all 5th-8th grade            results were robust when children’s pretest
students in the geographical area, students and          levels of social competence were controlled,
parents had a choice to enroll their child in the        along with other potential confounding
program. As such, the study cannot address               variables.    Last, although the sample was
whether this equine facilitated intervention             relatively small, participants were drawn from
would be effective for children who may have             ten public and private schools across two

15 | H A I B

different towns, and featured referred and non-         References
referred children of various ages, somewhat
alleviating concerns about the generalizability of      Bandura, A. (1986). Social foundations of
the sample.                                                 thought and action: A social cognitive
                                                            theory. Englewood Cliffs, N.J.: Prentice
Conclusion                                                  Hall.
    Given the causal design of this study, this         Bowers, M. J. & MacDonald, P. M. (2001). The
study provides the broader field of therapeutic             effectiveness of equine-facilitate
horsemanship professionals with scientific                  psychotherapy with at-risk adolescents: A
evidence that can be used to generate support for           pilot study. Journal of Psychology and
expanding         innovative,     evidence-based            Behavioral Sciences, 15, 62–76.
approaches through equine assisted activities.          Brickel, C. M. (1982). Pet-facilitated
That said, observed program effects are specific            psychotherapy: a theoretical explanation via
to the program under study, which is an                     attention shifts. Psychological reports, 50,
established, equine facilitated learning program            71-74. doi:10.2466/pr0.1982.50.1.71
with a structured curriculum and a team of              Brown, L. & Alexander, J. (1991). Self Esteem
experienced, PATH, Intl. certified instructors,             Index examiner’s manual. Austin, TX: Pro-
horse specialists, therapists, and facilitators.            ed.
Until more is known about the mechanisms                Budiansky, S. (1997). The nature of horses:
underlying its observed effects, prevention                 Exploring equine evolution, intelligence,
scientists should pair enthusiasm with caution              and behavior. New York: The Free Press.
before encouraging scaling up equine facilitated        Cain, G. (1975). Regression and Selection
programs in school and community settings to                Models to Improve Nonexperimental
serve as targeted prevention programs. In fact,             Comparisons, in C. A. Bennett and A. A.
future research should attempt to elucidate                 Lumsdaine (Eds.), Evaluation and
which components of equine facilitated learning             Experiment: Some Critical Issues in
are central to program success. For example, is it          Assessing Social Programs. New York:
grooming horses, riding horses, or being in the             Academic Press.
presence of horses or other children that drove         Chandler, C. (2005). Animal assisted therapy in
program effects? Or, was it the cognitive-                  counseling. New York, NY: Routledge.
behavior learning processes practiced during            Cohen, J. (1988). Statistical power analysis for
natural horsemanship activities that uniquely               the behavioral sciences (2nd edition).
enhanced children’s social competencies?                    Hillsdale, NJ: Erlbaum.
Similarly, to what extent does the equine               Dodge, K. A., & Somberg, D. R. (1987). Hostile
component of this type of program infer benefits            attributional biases among aggressive boys
above and beyond evidence-based prevention                  are exacerbated under conditions of threats
focusing on human to human interaction? Last,               to the self. Child Development, 58, 213–224.
future research should examine effects of equine            doi:10.2307/1130303
assisted interventions on other developmental           Durlak, J. A., & Weissberg, R. P. (2007). Meta-
outcomes.                                                   analysis of youth intervention programs.
                                                            Collaborative for Academic, Social and
                                                            Emotional Learning. Retrieved April 24,
                                                            2012, from http://casel.org/publications/the-

16 | H A I B

    impact-of-after-school-programs-that-              Katcher, A. H., & Wilkins, G. G. (1994). The
    promote-personal-and-social-skills.                   use of animal assisted therapy and education
Fisher, L.D., Dixon, D.O., Herson, J.,                    with attention-deficit hyperactive and
    Frankowski, R.K., Hearon, M.S., Pearce,               conduct disorders. Interactions, 12(4), 1–5.
    K.E. (1990). Intention to treat in clinical           doi: 10.1145/1070960.1070962
    trials. In: Pearce KE, editor. Statistical         Klontz, B. T., Bivens, A., Leinart, D., & Klontz,
    issues in drug research and development.              T. (2007). The effectiveness of equine-
    New York: Marcel Dekker; pp. 331–350.                 assisted experiential therapy: Results of an
Gatty, C. M. (2001). Psychosocial impact of               open clinical trial. Society & Animals, 15,
    therapeutic riding: A pilot study. Equine-            257-267. doi: 10.1163/156853007X217195
    Facilitated Mental Health Association, 5(2),       Kruger, K.A., & Serpell, J. A. (2010). Animal-
    8.                                                    assisted interventions in mental health:
Greenberg, M. T., Domitrovic, C. & Bumbarger,             definitions and theoretical foundations. In A.
    B. (2000). Preventing mental disorders in             Fine (Ed.) Handbook on Animal-Assisted
    school-age children: A review of the                  Therapy. New York: Academic Press.
    effectiveness of prevention programs.              LeBuffe, P.A., Shapiro, V.B., & Naglieri, J.A.
    Retrieved from Pennsylvania State                     (2009). The Devereux Student Strengths
    University, College of Health and Human               Assessment (DESSA). Lewisville, NC:
    Development website:                                  Kaplan Press.
    http://prevention.psu.edu/pubs/documents/M         Lemon, B.W., Bengston V.L., & Peterson, J.A.
    entalDisordersfullreport.pdf                          (1972). An Exploration of the Activity
Greenwald, A. (2001). The effect of a                     Theory of Aging: Activity Types and Life
    therapeutic horsemanship program on                   Satisfaction Among In-movers to a
    emotionally disturbed boys. Dissertation              Retirement Community, Journal of
    Abstracts International: Section B: The               Gerontology, 27, 511-523. doi:
    Sciences and Engineering, 62(2-B), 1078.              10.1093/geronj/27.4.511
Griffin, K.W., Epstein, J.A., Botvin, G.J., &          Macauley, B. & Guierrez, K. (2004). The
    Spoth, R.L. (2001). Social competence and             effectiveness of hippotherapy for children
    substance use among rural youth: mediating            with language-learning disabilities,
    role of social benefit expectancies of use.           Communication Disorders Quarterly, 24,
    Journal of Youth and Adolescence, 30, 485-            205-217. doi:
    498. doi:0047-2891/01/0800-0485                       10.1177/15257401040250040501
Huang, L., Stroul, B., Friedman, R., Mrazek, P.,       MacDonald, P. M. (2004). The effects of equine-
    Friesen, B., Pires, S. & Mayberg, S. (2005).          facilitated therapy with at-risk adolescents:
    Transforming mental health care for                   A summary of empirical research across
    children and their families. American                 multiple centers and programs. The Center
    Psychologist, 60, 615-627.                            for the Interaction of Animals and Society
    doi:10.1037/0003-066X.60.6.615                        (CIAS). Philadelphia: University of
Kaiser, L., Spence, L. J., Lavergne, A. G., &             Pennsylvania School of Veterinary
    Vanden Bosch, K. L. (2004). Can a week of             Medicine. Available at:
    therapeutic riding make a difference?—A               www2.vet.upenn.edu/research/centers/cias/
    pilot study. Anthrozoos, 17, 63–72. doi:               pdf/CIAS_AAI_white_paper.pdf
    10.2752/089279304786991918                         MacDonald, P. M., & Cappo, J. (2003). Equine-
                                                          facilitated therapy with ‘‘at-risk’’ youth:

17 | H A I B

   Does it work? Strides, 9, 30–31. doi:                  psychotherapy: A mental health
   10.1007/s10566-006-9031-x                              promotion/intervention modality for
Mann, D. & Williams, D. (2002). Equine                    children who have experienced intra-family
   assisted family therapy for high-risk youth:           violence. Health & Social Care in the
   Defining a model of treatment and                      Community, 15, 265-271. doi:
   measuring effectiveness, Unpublished                   10.1111/j.1365-2524.2006.00684.x
   manuscript.                                        Serpell, J. A. (2000). Creatures of the
McGreevy, P. D. (2007). The advent of                     unconscious: companion animals as
   equitation science. The Veterinary Journal,            mediators. In A. L. Podberscek, E. S. Paul &
   174, 492-500. doi:                                     J. A. Serpell (Eds.), Companion Animals
   10.1016/j.tvjl.2006.09.008,                            and Us (pp. 108-121). Cambridge:
McGreevy, P. D. & McLean, A. N. (2007).                   Cambridge University Press.
   Roles of learning theory and ethology in           Shonkoff, J., & Phillips, D. (2000). From
   equitation, Journal of Veterinary Behavior:            neurons to neighborhoods: The science of
   Clinical Applications and Research, 2, 108-            early childhood development. Washington,
   118, doi: 10.1016/j.jveb.2007.05.003.                  DC: National Academy Press.
McGreevy, P. D. & McLean, A. N. (2010).               Sørlie, M., Hagen, K.A., & Ogden, T., (2008).
   Equestrian Science. United Kingdom:                    Social competence and antisocial behavior:
   Blackwell Publishing.                                  Continuity and distinctiveness across early
O’Connell, M.E., Boat, T., & Warner, K.E.                 adolescence. Journal of Research on
   (2009). Preventing Mental, Emotional, and              Adolescence, 18, 121-144.
   Behavioral Disorders Among Young People:               doi:10.1111/j.1532-7795.2008.00553.x
   Progress and Possibilities. Washington,            Sørlie, M. A. & Nordahl, T. (1998).
   D.C.: The National Academic Press.                     Problematferd i skolen: hovedfunn,
Reynolds, C.R., & Kamphaus, R.W. (2004).                  forklaringer og pedagogiske implikasjoner:
   Behavior assessment system for children                hovedrapport fra forskningsprosjektet
   (2nd ed.). Circle Pines, MN: American                  “Skole og samspillsvansker” [Problem
   Guidance Service.                                      behavior in school: Main findings,
Rockhill, C.M., Vander Stoep, A., McCauley,               explanations and educational implications:
   E., & Katon, W.J. (2009). Social                       Main report from the research project
   competence and social support as mediators             “School and difficult social interactions”].
   between comorbid depressive and conduct                Oslo: NOVA.
   problems and functional outcomes in middle         Trotter, K.S. (2005). The efficacy of equine
   school children. Journal of Adolescence,               assisted group counseling with at-risk
   32, 535-553. doi:                                      children and adolescents. Dissertation.
   10.1016/j.adolescence.2008.06.011                  Trotter, K.S., Chandler, C.K., Goodwin-Bond,
Rubin, D.B. (1987). Multiple Imputation for               D. & Casey, J. (2008). A comparative study
   Nonresponse in Surveys. J. Wiley & Sons,               of the efficacy of group equine assisted
   New York.                                              counseling with at-risk children and
Rubin, D.B. (1996). Multiple imputation after             adolescents. Journal of Creativity in Mental
   18+ year. Journal of the American                      Health, 3, 254- 284. doi:
   Statistical Association, 91, 473-489.                  10.1080/15401380802356880
Schultz, P. N., Remick-Barlow, G. A., &               Vidrine, M., Owen-Smith, P., & Faulkner, P.
   Robbins, L. (2007). Equine-assisted                    (2002). Equine-facilitated group

18 | H A I B
You can also read
NEXT SLIDES ... Cancel