The lack of formal pedagogical education for medical faculty is not a problem unique to the veterinary field (Bing-You et al. 2017; Wilkerson & Irby 1998). Jason and Westberg (1982) found that most medical school faculty had no formal teacher preparation. When seeking to answer how clinical teachers in human medicine acquired knowledge about teaching, Irby (1994) found that they sourced knowledge by observing teachers when they were learners. This observational emulation suggests that role models are a vital key for clinical teachers who have had no formal training in teaching. This can be a slow and arduous process, often involving experimentation and reflection to find what works and what fails (Wilkerson & Irby 1998). Medical teacher education has since grown to include faculty development initiatives and enhanced learning curriculum (Steinert et al. 2016; Strand et al. 2015), yet sourcing pedagogical knowledge for the medical teacher is still a challenge for clinical training environments (Denton et al. 2015). When describing pedagogical concepts that must be mastered by clinical teaching faculty, Irby (2014) listed knowledge of pedagogy and learners, and knowledge integrated in teaching scripts, as foundational. Finally, Chen et al. (2016) concluded that medical faculty developed skills to teach multi-level learners while on the job and that a developmental approach during their appointment accounted for acquiring this pedagogical knowledge. For veterinary medicine, it has been posited that academic clinicians develop a teaching style similar to those they observed while in school (Smith & Lane 2015), but this has not yet been confirmed with empirical evidence. This observational learning could also explain how veterinary educators obtain pedagogical knowledge, but the need to ask these questions remains.

Teachers in other disciplines piece together the knowledge needed to become good teachers in similar ways. Research has shown that nurse educators learn on the job and use mentors in discovering how to effectively teach (Gardner 2014). Ironside, McNelis and Ebright (2014) discovered that nurses in clinical education continue teaching the way that they were taught, which included a focus on memorisation of medical knowledge, time management skills and completing medical records. These disparities highlighted in nursing education are also found in other fields. Interestingly, even teacher educators describe gaps in the development of teaching skills and acquiring pedagogies for educating student teachers. These teacher educators were expected by their professors to learn on site, but they eventually depended on their previous experiences while in school as classroom teachers (Goodwin et al. 2014).

A recurring theme found by researchers is the need for improving teaching practices in higher education through faculty development programmes, but how these programmes should be constructed is a matter of debate (Devlin & Samarawickrema 2010; Glicken 2004; Huston & Weaver 2008; Peluso & Hafler 2011; Sunal et al. 2001). Wilkerson and Irby (1998) suggested that medical school comprehensive faculty development programmes include professional, instructional, leadership and organisational development, and that teaching skills are separate from expertise in the educator’s field. Peer coaching has been advocated as a teacher development strategy, particularly for the older, more experienced faculty member (Huston & Weaver 2008). This approach focuses on collaborative, reflective learning strategies with equally experienced faculty members who meet to discuss and question the problems faced at their levels. It also advocates for the experienced faculty to mentor junior faculty.

Sunal et al. (2001) explored a series of faculty development programmes and studied why faculty members may be inhibited in their effort to become more effective teachers. They suggested that personal efficacy is valued above teacher efficacy. Personal efficacy was defined as ‘a faculty member’s beliefs about his or her ability to facilitate student potential learning’ and teacher efficacy was defined as ‘a faculty member’s belief about their own teaching ability’ (p. 251). Low personal efficacy results in a teacher believing that his or her students will not learn regardless of what the teacher does in that course. Low teacher efficacy results in a teacher believing that because of his or her poor teaching skills, he or she cannot successfully teach the students. Faculty members with higher personal efficacy are more likely to effect change in teaching and instructional strategies. Positive teacher development may not occur unless the educators experience dissatisfaction with their current personal concepts of teaching (Steinert et al. 2016; Strand et al. 2015; Sunal et al. 2001).

Clinical teachers share traits with effective teachers (Smith & Lane 2015). It is beyond the scope of this article to provide a complete review of the literature on what constitutes an effective or successful teacher. However, in reviewing the literature, there are a few salient points worth mentioning. Often, the different roles of a clinical educator cannot be accomplished by one person, and, in fact, clinical education may take a whole team. Clinical teachers have to take their roles beyond that of a role model or mentor, and also have to be the information provider, the case facilitator, the assessor of the learner, the course planner and the resource developer (Harden & Crosby 2000). Furthermore, excellent clinical teaching goes beyond ordinary teaching (Sutkin et al. 2008). Non-cognitive characteristics, such as enthusiasm, relationship skills and the ability to inspire, support, actively involve and communicate with students, dominate the themes of superior clinical teaching. These activities promote an emotional component and a connection with students, which transcends the construction of biomedical knowledge. These traits may be inherent in the excellent teacher, and may not be easily taught in formal teacher development protocols. However, formal educational strategies are still advocated for enhancing the skills of the clinical teacher. Thinking aloud, role modelling, reflection and the guided thinking process, deconstructing concrete experiences and providing the opportunity for active practice and timely feedback have all been identified as necessary for veterinary clinical educators (Smith & Lane 2015). An ethnographic case study of veterinarians in clinical teaching found that the effective educator demonstrated a reflective approach to clinical teaching (Magnier et al. 2014). These teachers sought to provide a challenging yet safe learning environment, while establishing rapport with learners.

At the time of submission of this article, studies on veterinary educators’ formal teacher education prior to their faculty appointment or how veterinary faculty members pieced together the knowledge needed to become teachers were not found. Based on the literature review, it was hypothesised that most veterinary faculty had limited to no formal training in teaching practices prior to their academic appointments. The information obtained will fill the knowledge gap regarding how veterinary faculty members obtain pedagogical knowledge and should lead to suggestions and improvements for current veterinary teacher development strategies.

The educational psychology theory that the project can be rooted in is situated cognition (Brown, Collins & Duguid 1989; Lave & Wenger 1991). This theory, also known as situated learning, could potentially explain how veterinary faculty members have derived a large portion of their education background from their communities and interactions in everyday life, from the collections of people and events they were influenced by while in veterinary school or other social settings. In medical education, those that influence this type of learning could be other students, patients, clinical instructors or environmental settings (Young et al. 2014).

This type of learning can often be described as unintentional. As it is hypothesised that most veterinary faculty will not have formal educational backgrounds, they are likely to use their experiences with previous educators and combine them with current mentoring from senior faculty to be successful. To this point, cognitive apprenticeship as described by Collins et al. (1991) likely will play a role. In this apprenticeship, the novice authentically learns from an expert who passes down traditions and methods. This is a form of sociocultural learning and can be embedded in an academic setting, within the workplace or within medical education (Lyons et al. 2016). In the context of this study, this theory could also help explain how junior faculty could improve their teaching through relationships with expert teachers in their respective fields (Collins et al. 1991; Stalmeijer 2015).

A mixed-methods approach was used to collect and analyse both close- and open-ended data. Qualitative methods that have been successfully employed in the past were used to establish the credibility of this study (Creswell 2013; Shenton 2004). Honesty was highly valued by allowing participants to decline the survey process, encouraging them to be frank and open in their answers, and letting participants withdraw from the research process at any time. Open- and close-ended questions allowed a broader explanation of the research questions. This helped the participants to clarify some of the survey responses and provide richer information. To establish transferability and dependability, transparency was highly valued and the context of the results of the study was made available. This included the number of participants involved, the data collection methods employed and the time of the data collection.

All participants were informed of the scope of the study and individual unsigned consent was obtained. There were no consequences for those who declined to participate. Instructors who were recruited in person received an unsigned consent document for personal records, which included a statement indicating informed consent and a brief description of the study’s purpose and procedures. Instructors who were recruited by email received another version of the consent document which provided the same information as mentioned above.

The Pedagogical Knowledge Acquisition Survey–Faculty (PKAS-F) (Online Appendix 1) was created in the fall of 2016 for the collection of data among veterinary faculty and it arose from multiple discussions and literature reviews during the author’s educational psychology class. It was designed to answer the three research questions in order to fill the gap in the knowledge of veterinary educators. Content was checked by an assistant professor of educational psychology at the institution where the survey was delivered to ensure appropriate organisation and presentation. The content was organised to first capture demographic information, and then subheadings were used to group questions into (1) teacher education training prior to becoming veterinary faculty, (2) teacher education since becoming veterinary faculty, (3) desire to learn, (4) formal teacher education, (5) influences and (6) teaching ability. Close-ended yes/no questions were utilised to capture answers without ambiguity, but were then followed with open-ended questions to develop a richer data set. Where appropriate, a Likert scale was used instead of yes/no responses in order to obtain richer data. The survey was intended to be completed in 30 min or less and was designed to lead veterinary faculty through a self-reflection process. Only the questions that captured information needed to answer the research questions were analysed. This survey was distributed to all consenting participants via an online survey engine. Upon completion, the researchers gathered all data from the survey company. Corresponding items on the PKAS-F were mapped to the research questions (Table 1).

Participants were recruited for the study in the fall of 2016 based on publicly shared online information that indicated the individual’s position as an instructional faculty member at a college of veterinary medicine in the southeastern United States. Email correspondence was used to recruit and secure consent from the participants. The recruitment email served as a disclosure of information about the study to potential participants, and was structured in a way that allowed for unsigned consent.

The sample consisted of veterinary faculty (n = 29) from a southeastern university college of veterinary medicine. A stratified sampling technique was used to capture data from different subgroups of instructional faculty (i.e. clinical instructor [n = 4], assistant professor [n = 5], associate professor [n = 9] and professor [n = 11]), all teaching in veterinary medicine. It was anticipated that sampling from a single university would not be problematic when considering that instructional faculty had arrived at their current positions with highly varied backgrounds and experiences. The land-grant university that the veterinary school is associated with is one of only a few schools to earn both research and community engagement rankings from the Carnegie Foundation for the Advancement of Teaching. The college of veterinary medicine awards Doctor of Veterinary Medicine (DVM) degrees, master’s degrees and Doctor of Philosophy degrees. It has been fully accredited by the American Veterinary Medical Association since 1981. The DVM degree is a 4-year curriculum that is composed of 2 years of classroom and laboratory work and 2 years of clinical education, which is a uniquely innovative approach in veterinary medical education. The clinical rotations vary from 4 to 8 weeks, with an emphasis on skill acquisition and technical competency. Observational, passive and experiential learning all play important roles in the clinical rotations. Clinical faculty along with residents and interns are responsible for delivering the learning experiences in the final 2 years of clinical education.

Codes emerged from inductive and deductive thematic analyses. Memorable quotes were noted and summaries were constructed including the main themes, points and insights expressed by each participant. Next, a coding process provided labels for emerging themes, concepts and examples (Table 2). After comparing and contrasting the various themes and concepts, each research question was answered with descriptive detail (Rubin & Rubin 2012).

Quantitative analysis was conducted using IBM SPSS Statistics version 24. The analysis investigated correlations between the instructors’ current position (question 3) with the instructors’ desire to learn more about teaching and instructional strategies (question 7). The latter question used a Likert scale with six responses from 1 (not at all) to 6 (most definitely). In G*Power Version 3.1.9.2, the statistical test was defined as Correlation: Point Biserial model, and the type of power analysis was defined as post hoc: Compute power – given alpha, sample size and effect size.

This project received Institutional Review Board approval with protocol ID: IRB-17–416 from the Mississippi State University.

Based on the survey results, it was found that most instructors (n = 27; 93%) did not receive formal teacher training, which supported the initial proposition. Only two participants had received prior formal education training: one had a BS in Education and the other had taken two education curriculum graduate-level courses. Much of the faculty who responded to question 10 (n = 20) reported attendance to university-sponsored workshops or seminars (n = 14; 70%) held by the Center for Teaching and Learning once already in their faculty appointments. For instance, Dr Y stated, ‘I have attended seminars at a number of different institutions that I have worked at covering a variety of topics … PowerPoints, organisation, use of clickers, etc.’ (Associate Professor, Southeastern United States). Faculty who attended conferences related to teaching and instruction (n = 3; 15%) described learning about problem-based learning, test writing and clinical instruction. Independent study and self-directed learning was used by a few faculty members (n = 2; 10%), as demonstrated by Dr A’s willingness to personally read books on the subject: ‘In my reading, I noticed several references indicating that enthusiasm is a more powerful explanatory variable than most others in relating to student success’ (Professor, Southeastern United States). One faculty member had taken two graduate courses since entering the faculty appointment and noted,

In an interesting contrast, 21 faculty members reported learning about teaching prior to their faculty appointments by emulating a past teacher they admired as students.

The majority of faculty who responded to question 26 (n = 24) reported obtaining their knowledge from a role model of some sort (n = 21; 88%). Of those responding that role models were a significant source of pedagogical knowledge, four stated that their parents were teachers and served as a significant influence in becoming an educator themselves. When responding to question 27, many participants (n = 20; 69%) indicated that they emulate past or present teachers and most of those (n = 15; 75%) modelled teachers or professors they had when they themselves were students.

Dr C illustrated,

Dr U explained,

Responses to questions 11, 12 and 25 were best summarised and conceptually coded across all participants (n = 29) to provide the most meaningful description of the data. This process was performed similarly to the descriptions in Table 2. The resulting themes were ranked in order according to the frequency of response category and quotes from open-ended questions were used to provide rich descriptions.

Results from a post hoc quantitative analysis of veterinary educators (n = 29) yielded one significant Pearson correlation with a medium effect size, a power of 0.65 and a = 0.05. A significant correlation (r = -0.41; p < 0.05) demonstrated a moderate negative relationship between current job title or position and the level of desire to learn more about effective teaching methods. To elaborate, the relationship indicated decreases in the level of desire to learn more about teaching as job title or position increased among veterinary faculty. In an interesting contrast, most participants who responded to question 32 (n = 27) stated that teaching in veterinary medicine was both an innate and learned ability (n = 24; 89%), indicating that many veterinary faculty understand that it is possible to improve instructional strategies and further develop pedagogical knowledge.

Dr S stated:

Dr A clarified: