Exploring medical educators’ perspectives on teaching effectiveness and student learning | BMC Medical Education

This mixed-methods study was conducted after obtaining ethical approval from the relevant Institutional Review Board. This was a sequential explanatory Quan-Qual study, in which the first phase was a quantitative component, and the second phase comprised a qualitative study. Participant confidentiality in the quantitative study was ensured by anonymously collecting data through coded questionnaires without identifying information. Informed consent was obtained, emphasising that all information would remain confidential and used solely for research purposes. Completed questionnaires were securely stored, and digital data were entered into a password-protected database that was accessible only to the research team. The data analysis focused on aggregated trends rather than individual responses, preventing the identification of any participant. The study’s confidentiality measures were reviewed and approved by the Ethics Committee to ensure compliance with ethical standards. By implementing these steps, participant privacy was rigorously maintained throughout the data collection and analysis, demonstrating a strong commitment to ethical research practices.

The study’s inclusion criteria focused on selecting faculty members with specific qualifications and experience. Faculty with postgraduate qualifications and a minimum of five years of experience teaching theoretical and clinical lessons in the medical or dental field were included. The selection criteria required participants to have published at least five scientific papers in the past five years. This timeframe was chosen to ensure that the educators included in the study were actively engaged in recent scholarly activities and were conversant with current trends and developments in medical and dental education. However, this study had some clear exclusion criteria. Educators not involved in medical education, such as those from engineering, arts, and other non-medical disciplines, were excluded. This was done to ensure the relevance and specificity of this study in the field of medical education. Visiting faculty members were also excluded, as the study aimed to analyse the perspectives of more permanent, established educators in the medical field.

Further exclusion criteria were participants with incomplete data or those who dropped out before completing all phases of the study. Such exclusions were necessary to maintain the integrity and completeness of the research data. Additionally, faculty members who failed to attend the semi-structured interviews, which were a crucial part of the study, and who did not provide a valid reason for their absence were excluded. This criterion was set to ensure that the study results were based on comprehensive and reliable input from fully engaged participants.

Quantitative study methods

Questionnaires were provided to all the postgraduate faculty available at the Dow University of Health Sciences and six other medical and dental schools in Karachi, Pakistan, with an explanation of the purpose of the study. Participants were given two weeks to complete and submit the forms, ensuring adequate time to provide thoughtful responses while maintaining the study’s schedule. The questionnaires were collected manually and represented using a positivist paradigm approach. Reminders were given for unfilled forms and were subsequently approached to be collected. Convenience sampling was used in this study. Participants were informed that they would be contacted for face-to-face in-depth interviews.

Quantitative instrument

To objectively assess teaching approaches, this study used the Approaches to Teaching Inventory (ATI), a validated instrument widely used in educational research [30, 31]. The ATI is designed to measure the extent to which educators adopt teacher-centred or student-centred strategies, aligning well with the study’s objectives of exploring instructional methods (Appendix C). The choice of the ATI over other available tools is based on its strong psychometric properties, widespread acceptance in the academic community, and its specific focus on higher education teaching practices [32, 33]. Additionally, the ATI allows for comparability with other international studies, facilitating a broader understanding of teaching approaches across different contexts [32, 34, 35].

Prosser and Trigwell (1999) classified teaching approaches into two main types: the Information Transmission/Teacher-Focused (ITTF) approach and the Conceptual Change/Student-Focused (CCSF) approach [30, 31, 34]. Within these two overarching categories, they identified five qualitatively different approaches to teaching.

1. 1.

   Approach A: Teacher-focused strategy with the intention of transmitting information to students.

2. 2.

   Approach B: Teacher-focused strategy aimed at helping students acquire the concepts of the discipline.

3. 3.

   Approach C: The teacher–student interaction strategy with the intention that students acquire the concepts of the discipline.

4. 4.

   Approach D: Student-focused strategy aimed at developing students’ conceptions.

5. 5.

   Approach E: Student-focused strategy with the intention of bringing about conceptual changes in students.

The ATI comprises 22 items divided into two main scales: the ITTF and CCSF [20]. The CCSF scale describes an approach intended to change students’ ways of thinking through a teaching strategy that focuses on students. The ITTF scale describes an approach intended to transmit information by using a teacher-focused strategy. Each item is rated on a 5-point Likert scale, where 1 represents “only rarely” and 5 represents “almost always.” This detailed scoring system enables a comprehensive assessment of educators’ teaching orientation and provides valuable quantitative data on their instructional preferences. A total of 130 questionnaires were distributed and 93 questionnaires were returned, with a response rate of 71%.

Pretest

Before launching the main study, a pilot inventory test was conducted with 40 participants, who were not enrolled in the primary investigation. The pre-test assessed comprehension and determined the average time taken to complete the questionnaire. An expert health professional assessed the content validity of the questionnaire, ensuring that the content was appropriate and relevant for the intended audience. To evaluate the reliability of the questionnaire used in this study, we conducted an internal consistency analysis using Cronbach’s alpha. The questionnaire comprised items categorised into two subscales: Teacher-focused and Student-focused.

The Teacher-focused subscale, after the removal of questions 9 and 11, demonstrated a mean score of 3.97 with a standard deviation (SD) of 0.44. The internal consistency for this subscale was determined to be acceptable, with a Cronbach’s alpha value of 0.70. For the student-focused subscale, questions 8 and 15 were excluded to enhance the reliability. The mean score for this subscale was 4.44, with a standard deviation of 0.41. The internal consistency of the student-focused subscale was strong, as indicated by the Cronbach’s alpha value of 0.80.

The overall questionnaire, encompassing all items, yielded a mean score of 4.21 with an S.D. of 0.70. The complete questionnaire exhibited high reliability with a Cronbach’s alpha value of 0.80. These results indicate that the questionnaire, both in its entirety and within its subscales, possesses good internal consistency, making it a reliable tool for assessing the constructs of interest in this study.

Statistical analysis

For the quantitative data analysis in the study, SPSS version 24 served as the primary software tool. Differences between variables were evaluated through cross-tabulation, accompanied by the chi-square test. The central tendency of the scores was determined using the mean values. To ascertain the significance of the relationships between variables, p-values were used.

Qualitative study methods

Guided by the constructivist framework, this study used semi-structured interviews to investigate medical faculty members’ perceptions of their teaching and professional development. Faculty members with both high and low ATI inventory scores were selected for interviews. Participants were equally divided between medical and dental educators, with half from medical colleges and half from dental colleges. Additionally, there was an equal mix of faculty from the government and private institutions, totalling 25 experienced educators who were purposively selected for in-depth qualitative interviews after providing informed consent.

The selection criteria included factors such as availability, previous connection to the researcher, seniority, and their roles within their institutions. This study acknowledges that bias is an inherent aspect of qualitative research, primarily when prior relationships exist between the researcher and participants. Recognising that complete objectivity is unattainable, several measures were implemented to mitigate potential biases and enhance the study’s validity and credibility. The researcher engaged in reflexivity by continually reflecting on personal biases and and how these might influence data collection and interpretation, aiming to minimise subjective influence. Reflexivity was further maintained through the use of a reflective journal, allowing the researcher to document and critically examine personal thoughts and potential biases throughout the research process.

To ensure the accuracy and credibility of the data, after transcribing the interviews, the transcripts were shared with the participants for their review. This process, known as member checking, allowed participants to confirm the accuracy of their statements and add additional thoughts, ensuring that their perspectives were authentically represented. Direct quotes from participants were included in the analysis to accurately convey their viewpoints, thus reducing the influence of the researcher’s interpretations. The study deliberately selected a diverse group of participants, equally divided between medical and dental educators, from both government and private institutions. This diversity minimised the impact of any single relationship on the overall findings. The participants were assured of complete confidentiality and anonymity, encouraging honest and open communication. The interviews were conducted in a neutral, private setting to reduce any potential discomfort or power dynamics stemming from prior relationships. A standardised semi-structured interview guide was used to maintain consistency across all interviews, minimising the likelihood of the researcher’s prior knowledge influencing participants’ responses. Additionally, the co-authors independently reviewed the data and analyses to identify and correct any potential biases in interpretation. By transparently acknowledging the inherent biases and implementing these measures, this study aimed to ensure that the findings accurately reflected educators’ perspectives and experiences while maintaining ethical research practices. The interviews were audio-recorded for research purposes and were accessible only to the researcher and study supervisor.

The interviews began with general questions about the teaching experience and gradually narrowed down in focus, with minimal interruptions to guide the conversation towards the study’s objectives. The researcher refrained from injecting personal views into the discussion and used neutral probes to facilitate the flow of the interview.

The semi-structured interviews comprised eight main questions and probing items for further clarification (Appendix A). The duration of the interviews ranged from 40 to 60 min. All interviews were conducted in English and the participants’ preferred language. In a study that included qualitative data, such as this one, the study’s primary outcomes were the identification and analysis of recurring themes from semi-structured interviews. This could offer insights into everyday challenges and best practices for teaching effectiveness and student learning. Interviews were transcribed verbatim. All audio recordings were converted manually into text. This transcription process was followed by a qualitative content analysis [36]. Qualitative content analysis is instrumental in understanding phenomena and social realities by using written data. A significant advantage of this method is its capability to analyse large datasets without the risk of unwanted interaction effects between the participant and researcher [37].

The analysis involved identifying 94 different codes across the scripts, which facilitated the development of conceptual ideas beyond mere descriptions, as shown in Appendix C. These codes were subsequently categorised into themes. The data were treated as a unified entity and analysed following McCracken’s (1988) five-step method [38]. In the first step, the transcripts were read thoroughly with notations made in the margins to highlight the initial observations. The second step involved developing these observations into preliminary descriptive and interpretive categories, informed by evidence from the transcripts, relevant literature, and guiding theoretical frameworks. The third stage examined these preliminary codes more closely to identify connections and to develop pattern codes. In the fourth stage, basic themes were determined by analysing clusters of comments from respondents and memos made by the researchers. The fifth and final stages entailed examining these themes across all interviews and identifying the predominant themes that emerged from the data. These predominant themes served as answers to the research questions and formed the foundation for the data write-up. Additionally, field notes were consistently taken during the interviews to capture the respondents’ expressions and impressions. This practice was instrumental in understanding the unspoken elements and provided valuable context during the data analysis phase.

Sample size estimation

Convenience sampling was performed in this study. For the quantitative part, which used the Approach to Teaching Inventory (ATI), power analysis helped identify the required sample size to detect an expected effect with a given level of confidence and power. To detect a medium effect size (Cohen’s d = 0.5) with 80% power and a significance level (alpha) of 0.05 was used. Under these assumptions, a two-tailed t-test required a sample size of approximately 128 participants to detect a statistically significant effect [39].

In many real-world research settings, particularly in specialised fields like medical education, there may be constraints on the number of available participants. Given resource limitations, a sample size of 128 participants may represent a sizable proportion of the available population and may be the maximum feasible size. A sample size of 128 is suitable for various statistical tests, such as t-tests, ANOVA, and multiple regression, provided that the number of predictors is not too large. It can also allow for stratified analyses by subgroups if needed. While convenience sampling has drawbacks, such as potential bias and limited generalizability, it is sometimes the most practical method [40].

Qualitative studies often use the concept of “saturation” to determine the sample size [41]. This is typically when no new information or themes are observed in the data. The qualitative sample is supposed to be a subset of the quantitative sample for more in-depth exploration. However, only a quarter of the sample size was used in the qualitative study, with 25 purposively selected participants.