Test anxiety can have undesirable consequences for learning and academic achievement. This phenomenon is characterized as a form of performance anxiety that affects a student's ability to prepare for and take an examination (Poorman et al., 2011). Test anxiety impacts an estimated 25 percent to 40 percent of undergraduate students in various areas of study, and prelicensure baccalaureate nursing (BSN) students exhibit higher levels in comparison to students from other disciplines (Gerwing et al., 2015; Wedgeworth, 2016). Heightened levels of test anxiety correlate with lower test scores and grade point averages (Duty et al., 2016; Khalaila, 2015), often preventing nursing students from reaching their full academic potential. This negative impact may impede learning outcomes and program completion rates, ultimately threatening the production of qualified BSN-prepared nurses.
Historically, test anxiety is defined as an unpleasant, emotional state experienced in formal testing situations (Dusek, 1980). As a multidimensional state involving physical, psychological, behavioral, and cognitive components, test anxiety is commonly characterized by worry and emotionality. Worry results from negative thoughts and fear of failure that may be aroused several days before a test and linger beyond test completion. Emotionality is a transient event that accompanies worrisome thoughts and causes physiological manifestations (Spielberger et al., 1983). It diminishes the test-anxious student's ability to concentrate and synthesize information and thereby increases the likelihood of weak test performance.
Test anxiety stems from both state and trait anxiety (Spielberger, 1980). State anxiety is a temporary state that results in worrisome thoughts and apprehension immediately before and during a test. As state anxiety increases, physiological symptoms associated with activation of the autonomic nervous system may occur, including elevated heart rate, restlessness, nausea, and sweating. Often, these symptoms can be debilitating and easily distract the student. Test-anxious students feel threatened when preparing for and taking a test and are thus more likely to display higher levels of state anxiety (Spielberger, 1980). Trait anxiety is an enduring personality trait, represented by a repetitive, long-term tendency to exhibit state anxiety in testing situations with consistent anticipation of negative outcomes. Consequently, the ramifications of test anxiety may extend beyond an immediate testing situation and negatively influence future evaluative experiences.
Tests are an integral part of nursing education and widely utilized to assess the achievement of learning outcomes. Typically, a variety of evaluative strategies are used to assess performance, yet test grades often weigh heavily in the determination of final course grades (Oermann et al., 2009). The testing process in prelicensure BSN programs is unique in comparison to that of other disciplines. Many programs incorporate high grading standards, stringent benchmarks, and strict progression policies. These requirements are necessary to facilitate program completion, prepare students for the National Council Licensure Examination (NCLEX-RN(R)), and ensure readiness for clinical practice. BSN students are expected to utilize high-level thinking to analyze and apply learned concepts in didactic and clinical learning environments. Accordingly, objective tests are designed to assess the student's ability to make sound clinical judgments that ultimately impact patient care (Poorman et al., 2011). Many students are challenged by these demanding expectations and fear failure. Consequently, prelicensure BSN students identify testing as a top-ranked academic stressor (Brodersen, 2017), placing them at higher risk of developing test anxiety.
Test-anxious students often feel unprepared, lack confidence, and manifest self-doubt. Higher levels of test anxiety are linked with impaired problem-solving ability, lower test scores, declining grade point averages, and increased attrition (Duty et al., 2016; Miller & Sawatzky, 2017). Consequently, this phenomenon may threaten program completion rates, successful transition to professional practice, and the production of qualified nurses in the workplace. Given the negative impact of test anxiety, this study examined the impact of a therapy dog intervention on self-reported and physiological measures of test anxiety among prelicensure BSN students.
BACKGROUND
The detrimental effects of collegiate test anxiety create a growing concern for educators that has generated research to investigate the effectiveness of anxiety-reducing interventions before and during a test. Implementation of relaxation techniques before taking a test has successfully decreased self-reported test anxiety in nonnursing undergraduate students (Clinton & Meester, 2019) and prelicensure BSN students (Manansingh et al., 2019; Ward, 2019; Zargarzadeh & Shirazi, 2014). Prato and Yucha (2013) also found that physiological symptoms associated with BSN test anxiety, including heart rate and respiratory rate, decreased with the use of pretest relaxation techniques. Listening to classical music while taking a test decreased self-reported test anxiety in nonnursing undergraduate students (Goldenberg et al., 2013) and BSN students (Lai et al., 2008). Aromatherapy with essential oils has been explored in BSN students (Johnson, 2014; Kavurmaci et al., 2015), leading to decreased test anxiety during an exam. Although findings of these studies are significant, self-report measures serve as the primary basis of evidence. Although nursing education research has superficially probed the topic of test anxiety, the lack of substantial evidence is noteworthy.
More recently, animal-assisted intervention (AAI) with therapy dogs has emerged in the literature as a cutting-edge strategy to moderate the effects of student anxiety. AAI is a goal-oriented intervention that incorporates the use of animals for therapeutic gains in humans. Evidence stems from knowledge that interaction with dogs enhances physical, cognitive, behavioral, and emotional functioning of humans (International Association of Human-Animal Interaction Organizations, 2018), and research has recognized this positive impact as it applies to various aspects of human well-being. Described as a therapeutic healing environment, therapy dog interactions have positively impacted clinical outcomes in adults and children. In adults, interaction with dogs has effectively reduced pain (Harper et al., 2015) and decreased blood pressure and heart rate (Krause-Parello & Kolassa, 2016). Therapy dogs have also calmed patients with Alzheimer's disease (Bono et al., 2015) and reduced the incidence of depressive symptoms in older adults (Olsen et al., 2016). In children with attention-deficit/hyperactivity disorder, dog interaction has enhanced outcomes of cognitive-behavioral interventions (Schuck et al., 2015). Dogs have also reduced anxiety in the pediatric dental setting (Vincent et al., 2020), reduced distress in children undergoing venipuncture (Vagnoli et al., 2015), and decreased pain in pediatric oncology inpatients (Chubak et al., 2017). Interestingly, therapy dog visitation reduced hospital staff nurse burnout while increasing compassion and job satisfaction (Ginex et al., 2018).
Therapy dog interventions are gaining momentum on college campuses worldwide with recent attention in the literature. Undergraduate nonnursing students have benefitted from structured interaction with dogs, as noted by a decrease in self-reported anxiety (Barker et al., 2016; Muckle & Lasikewicz, 2017) and blood pressure readings prior to taking a test (Jarolmen & Patel, 2018; Muckle & Lasikiewicz, 2017). Moreover, therapy dogs have shown promise exclusively with prelicensure BSN students. A multidisciplinary study involving both nonnursing and BSN students (48 percent) resulted in decreased levels of test anxiety across multiple measures including self-report, blood pressure, heart rate, and salivary cortisol levels (Delgado et al., 2018). Anderson (2018) also noted a significant decrease in test anxiety, as shown by a decrease in State-Trait Anxiety Inventory (STAI) for adults scores prior to a dosage calculation examination. To lend further support, self-reported test anxiety (Hall & Duke, 2020) and salivary cortisol levels (Gebhart et al., 2020) decreased following interaction with therapy dogs; however, these studies were inclusive of associate degree nursing students only.
Therapy dogs have an innate ability to provide comfort and joy in many different contexts, including academia. Multiple disciplines have examined therapeutic benefits of the human-dog bond, yet this topic is poorly developed in the context of nursing students. It is important to gain a better understanding of this intervention and ability to elicit positive change in prelicensure BSN students. Ideally, further investigation of a therapy dog intervention may decrease test anxiety, improve learning outcomes, and promote student success.
METHOD
The purpose of this study was to examine the impact of a therapy dog intervention on psychological and physiological measures of test anxiety in prelicensure BSN students. This intervention involved structured interaction with certified therapy dogs prior to the first test of the semester.
Sample and Setting
Eligible participants included prelicensure BSN students enrolled in a junior-level adult health nursing theory course during the fall 2019 semester. Following institutional review board approval, a convenience sample was recruited from three programs in the northeastern region of Pennsylvania. These BSN programs included two public universities and one private university. Exclusion criteria included a known or suspected allergic response to dogs, personal fear of dogs, or history of adrenal disorder. The sample was composed of 91 predominantly female (91.2 percent), white students (91.2 percent), with a mean age of 20.33 years. Most students reported taking the course for the first time (94.5 percent).
Design and Instruments
A one-group, pretest/posttest quasi-experimental design was used to examine the impact of a therapy dog intervention on repeated measures of test anxiety. In the absence of randomization, this design allows for comparison of scores before and after the intervention (Polit & Beck, 2012).
Multiple instruments were employed to measure psychological and physiological indicators of anxiety. The STAI for adults was used to measure self-reported anxiety. The STAI, originally developed by Spielberger et al. (1983), is a well-established tool utilized extensively in multidisciplinary research worldwide. It has two subscales designed to measure both state and trait anxiety; this study used only the state anxiety subscale. The 20-item subscale measures subjective feelings of apprehension, tension, nervousness, and worry. Brief statements, such as "I feel calm" and "I feel upset," evaluate how participants feel in the immediate moment. Total scores range from 20 to 80, with higher scores indicating greater anxiety. Internal consistency alpha coefficients for the combined subscales range from .86 to .95 (Julian, 2011; Spielberger et al., 1983); a meta-analysis of 45 studies confirmed an internal consistency coefficient of 0.91 for the state anxiety subscale alone (Barnes et al., 2002). In the current study, Cronbach's alphas for the pretest and posttest STAI were .920 and .885, respectively.
Anxiety has the potential to induce physiological manifestations, including elevation in heart rate and cortisol levels. A noninvasive transcutaneous pulse oximeter was used to measure heart rate. This device uses a light-emitting diode to measure heart rate through the variation of capillary blood volume in subcutaneous tissue of the fingertip. Research shows that pulse oximeter technology serves as a valid and accurate instrument to assess heart rate in healthy adults (Losa-Inglesias et al., 2016). When comparing radial pulse palpation and finger pulse oximetry, high reliability and consistency between systems were reported with intraclass correlation coefficients greater than .972. Furthermore, a paired t-test indicated no significant differences between these heart rate measurement systems (p = .592).
Salivary cortisol levels also served as a physiological measure of anxiety. When an individual perceives a threat, such as a test, the hypothalamic-pituitary adrenal axis is stimulated to produce cortisol, a glucocorticoid hormone. Upon stimulation of the hypothalamic-pituitary adrenal axis, the adrenal cortex releases cortisol in response to stress and anxiety. Technological advances have enabled reliable salivary cortisol testing and have revolutionized this form of measurement. Salivary cortisol closely parallels serum cortisol concentrations; it is known to be reliable and consistent (r = .91) within and between various reputable laboratories (Calvi et al., 2017).
Data Collection
To maintain confidentiality, each participant received a six-digit identification code that was used consistently for all elements of data collection and analysis. Data were collected from each participant 60 to 90 minutes before the scheduled nursing test; collection took place immediately before and after the dog intervention, permitting an analysis of change. Pretest and posttest measurements were acquired in an identical manner. Participants completed a paper version of the STAI state anxiety subscale, followed by collection of a salivary cortisol specimen vial oral self-swab. Consumption of food and fluids within 60 minutes of sampling, especially those that consist of high carbohydrate content, may directly affect salivary cortisol levels (Strahler et al., 2017). Therefore, participants were asked to refrain from eating or drinking within 60 minutes of collection. Next, heart rate was measured with a calibrated pulse oximeter placed on the participant's fingertip for at least 30 seconds. To minimize the likelihood of an elevated heart rate secondary to recent activity, each participant was granted a period of seated rest for a minimum of five minutes prior to this measurement.
Immediately following completion of pretest measurements, participants proceeded to a designated area to interact with multiple therapy dogs. The length of interaction was 20 to 30 minutes, and participants were able to pet the dogs, engage in playful activity, and display affection toward the dogs according to personal comfort level. The number of therapy dogs was dependent on the number of participants at each site; a ratio of five or fewer students per dog was enforced.
Certified therapy dogs were used for the intervention and accompanied by a handler, a person required to perform this service on a voluntary basis. The dogs obtained certification from reputable organizations, including Therapy Dogs International and Alliance of Therapy Dogs. Certification requires dogs to be properly licensed and trained with a minimum of basic obedience; in addition, the dog and its handler must successfully complete at least 50 individual therapy visits prior to certification. The dog must possess a friendly disposition and calm behavior, and handlers must provide veterinary documentation of dog health status and immunizations; all dogs are expected to be well-groomed prior to each human interaction.
Following the dog interaction, participants returned to the designated data collection area where posttest measurements were obtained in the same manner as before with at least five minutes of seated rest prior to measuring heart rate. To ensure consistency and reliability throughout data collection, one trained research assistant conducted heart rate measurements and was responsible for supervising and packaging all salivary cortisol specimens per standardized protocol. Each specimen was packaged in a cryo-storage container and frozen to -4 degrees Fahrenheit within four hours of collection. Once all data were collected, the specimens were shipped to Salimetrics(R) according to laboratory and shipping guidelines.
Quantitative data were analyzed using the Statistical Package for Social Sciences(R) Version 26. Descriptive data and paired samples t-tests comprised the basis of this analysis.
RESULTS
Following the structured therapy dog intervention, a significant decrease in all mean scores from pretest to posttest was noted. As summarized in Table 1, the STAI mean pretest score was 48.85 points (SD = 10.40) and the mean posttest score was 32.72 points (SD = 7.49), with a total score difference of -16.48 points (SD = 8.32). The mean pretest heart rate was 82.18 bpm (SD = 14.51) and the mean posttest heart rate was 75.78 bpm (SD = 13.03), resulting in a difference of -6.40 bpm. Mean pretest salivary cortisol levels were 0.670 [mu]g/dl (SD = 0.427) and mean posttest levels were 0.639 [mu]g/dl (SD = 0.405), resulting in a difference of -0.031 [mu]g/dl. Typically, salivary cortisol values in adults range from 0.012 to 3.000 [mu]g/dl. Although this measurement is commonly used to diagnose medical conditions involving adrenal dysfunction, this study analyzed salivary cortisol specifically as an indicator of anxiety. Accordingly, the change in cortisol levels from pretest to posttest was more relevant than assessing the normality of cortisol levels.
Paired-samples t-tests revealed a significant difference between pretest and posttest mean scores for each measurement. Table 2 summarizes these results, indicating mean score differences for the STAI, t(89) = 19.233, p = .00, d = 1.73; heart rate measurements, t(90) = 6.927, p = .00, d = 0.726; and salivary cortisol levels, t(86) = 7.031, p = .00, d = 0.754.
DISCUSSION
This study demonstrated causality between the therapy dog intervention and reduction in situational test anxiety. Paired samples t-tests indicate the intervention invoked a statistically significant reduction in student anxiety across multiple measures prior to a nursing course test. This conclusion is derived from a decrease in STAI scores, heart rate, and salivary cortisol levels immediately following interaction with therapy dogs. Self-report using the STAI subscale served as a measure of psychological test anxiety whereas heart rate and salivary cortisol levels provided measurement of physiological test anxiety. Moreover, a large effect size for each measurement established clinical significance, which lends support to the practical value and meaningful nature of the intervention. Acute test anxiety tends to cause worrisome thoughts and negativity, often accompanied by an elevation in heart rate and cortisol secretion. As a result, the test-anxious student may have difficulty with attention throughout the test and trouble synthesizing information. The results of this study support therapy dog interaction as an effective strategy to reduce anxiety-provoked physiological responses in the immediate moment, ideally allowing the student to cope better with negative and worrisome thoughts prior to and during a test. Ultimately, the calming effect of dogs may help the student to focus better, optimize test performance, and maximize their true potential.
Prior research shows that dog interventions have been successful in reducing test anxiety among undergraduate students in various majors (Barker et al., 2016; Jarolmen & Patel, 2018; Muckle & Lasikewicz, 2017). The current study corroborates this previous research, providing more evidence that dog interaction can reduce test anxiety specifically in nursing students. The reduction in self-reported test anxiety and physiological symptoms is congruent with current literature (Anderson, 2018; Delgado et al., 2018), illustrating effectiveness of the intervention. This study enhances empirical evidence and confirms the suitability of a therapy dog intervention for test-anxious BSN students.
Unfortunately, the use of therapy dogs to reduce test anxiety is a topic that is not well developed in higher education. More specifically, in nursing education, there is a dearth of studies to provide substantial evidence. The current study acknowledges this gap by expanding upon prior knowledge. Unlike previous research, a multisite approach was utilized while simultaneously incorporating three measurements to quantify test anxiety. In addition to raising awareness of therapy dog benefits, this study adds to the inadequacy of scientific evidence and emphasizes the value of this intervention.
Implications for Nursing Education
The dynamic nature of nursing education places unique stressors on BSN students. Rigorous benchmarks, stringent clinical requirements, and strict program progression policies may contribute to an elevated level of test anxiety. An added stressor is the use of high-stakes testing, which culminates in the NCLEX-RN. BSN students identify high-stakes testing as a major trigger of test anxiety (Brodersen, 2017; Keller, 2016), and nursing tests are becoming more stringent to keep pace with the dynamic health care environment. As test performance is directly related to the academic success of BSN students, elevated anxiety may undermine academic achievement.
Test anxiety may not impact all students. However, the negative relationship between test anxiety and academic performance is concerning for nurse educators and administrators. Program completion rates may decline and threaten the production of qualified nurses, something that also creates concern for employers. The cycle of unresolved BSN test anxiety may transcend through professional role transition, manifesting as performance anxiety in clinical practice. The transition to professional practice, coupled with role reality, can create high levels of anxiety for the novice nurse, especially for a test-anxious individual. The impact on well-being may lead to emotional instability, job dissatisfaction, role burnout, and attrition, negatively influencing the recruitment and retention of a stable workforce.
Faculty need to use supportive strategies to interrupt the negative cycle of test anxiety. Although a dog intervention may not fully eliminate the problem or exert a long-term effect, the potential for self-reflection is an added benefit. Exposure to therapy dogs may help students gain awareness of personal anxiety before a test, ideally facilitating self-reflection and introspection. In alignment with recent American Association of Colleges of Nursing (2021) Core Competencies for Professional Nursing Education, the process of self-reflection is encouraged as a means of fostering resilience and well-being. Ultimately, this process will help students self-implement strategies that reduce test anxiety before it becomes problematic and interferes with performance.
Although the majority of test anxiety research relates to undergraduate students, test anxiety is also prevalent in graduate students, especially those enrolled in the health care professions (Macauley et al., 2018). Implementation of test anxiety-reducing strategies, such as therapy dog intervention, may also benefit graduate nursing students. Given the consequences, it is critical that educators identify, examine, and mitigate nursing student test anxiety at an early stage.
In an endeavor to develop sound, evidence-based educational practices, exploration of test anxiety is significant to the future of nursing education. An understanding of effective strategies to prevent or gain control over anxiety is important to student success. Nurse educators are in a unique position to help students overcome test anxiety and maximize their full potential (Poorman et al., 2019). The findings and recommendations from this study may assist educators and program administrators in the development and implementation of therapy dog interventions. This mutually beneficial strategy is not only cost-effective and intriguing but also engaging and simple to implement in a supportive environment. It is well known that dogs provide comfort and reduce personal stress while enhancing physical, emotional, and cognitive functioning of many people, including college students. This intervention holds great promise and can be beneficial for nursing students as they prepare for graduation, licensing exams, graduate education, and the transition to professional practice.
Implications for Future Study
Nurse educators are encouraged to implement and investigate therapy dog interventions; additional research is necessary to validate results of this study and increase awareness. To enhance generalizability of the findings, utilization of randomized control designs and longitudinal approaches is needed. A more rigorous design may provide further evidence to support the use of therapy dogs to alleviate test anxiety.
Future studies should incorporate a diversified population of prelicensure BSN students by recruiting from rural, suburban, and urban institutions. Although gender and racial disparity are recognized as limitations of many nursing education studies, it is important to strive for improvement. Deeper exploration of this intervention may also expand to non-BSN students, nontraditional students, and graduate students.
In addition to the testing process, nursing student anxiety is derived from a variety of academic and clinical influences. Anxiety related to clinical simulation and clinical practice is described in the nursing literature, yet few evidence-based interventions aimed to minimize this type of anxiety have been explored. The classroom environment may also contribute to student anxiety, yet this topic is not well described, especially in nursing education. Likewise, the benefit of therapy dog interventions may extend beyond testing. Ultimately, additional study across disciplines may benefit a diverse collegiate population and provide a gateway for ground-breaking interdisciplinary collaboration.
It is also important to assess the impact on learning outcomes. Although this study has shown that interaction with dogs can reduce test anxiety across various psychological and physiological measures, the impact on test scores and final course grades is unknown. Ideally, a longitudinal study may provide a well-suited research design to address the impact of AAI on grades. Current evidence is limited; these suggestions can be used to pave the way for future innovative research. Ultimately, our goal is to promote student success, produce competent graduate nurses, and create a strong BSN-prepared workforce.
Limitations
Noted strengths of this study include multiple quantitative measurements and the use of three BSN programs; however, limitations warrant discussion. This one-group, quasi-experimental study lacked randomization and a control group for comparison, relying solely on the analysis of change from pretest to posttest measurements. The convenience sample, narrow geographic location, and lack of participant diversity produced relatively homogenous demographic characteristics. Although the overrepresentation of female participants adequately represents the national BSN population, the lack of racial diversity may threaten general application of these findings. Although these limitations are common in nursing education research, future studies should aim to incorporate a diverse sample from a larger geographical area.
Social desirability is recognized as a potential limitation when using self-report measures, such as the STAI. One must consider this effect in the setting of research, implying that participants may respond in a socially acceptable or preferred manner while being studied (McCambridge et al., 2014). Thus, confounding issues such as honest reporting and response bias may limit the true validity of a subjective questionnaire. Fortunately, this study did not rely on self-report as the sole measurement of test anxiety.
Issues related to cortisol measurement may result in limitations. Salivary cortisol testing is a complex measurement that may be impacted by timing of collection and extraneous variables. Samples collected in the early morning often provide the most sensitive cortisol measurement; however, samples in this study were collected at various times of the day in correlation with timing of the nursing test. Although a paired-samples t-test resulted in statistical significance, the skewed distribution of cortisol levels may be attributed to inconsistent measurement times. In addition, consumption of substances such as food, fluids, caffeine, alcohol, nicotine, and medications should be considered as potentially confounding variables when measuring salivary cortisol, despite inconsistencies reported in the literature. Other variables, such as personal history of anxiety or other mental health issues, hours of sleep, and change in eating habits, may exert a negative impact on study outcomes and should be considered when designing future studies.
CONCLUSION
Nursing education places unique stressors on BSN students, often culminating in heightened test anxiety. Reducing this type of anxiety may enhance academic performance, learning outcomes, and program completion rates. Accordingly, an understanding of effective strategies to prevent or gain control over anxiety is imperative. This study contributes to the sparse body of nursing education literature related to the effective use of therapy dog interventions to reduce prelicensure BSN student test anxiety.
Implications and recommendations from this study may be used to assist educators and administrators with the implementation of therapy dog interventions designed to reduce nursing student test anxiety. As a cost-effective, exciting, and easy-to-implement strategy, therapy dog interventions have a great potential to decrease prelicensure BSN test anxiety, widely impact learning outcomes, and enhance program success.
REFERENCES