Professional Ethics Education and Moral Sensitivity in Public Health Students: A Pilot Quasi-Experimental Study

Abstract

Background

Public health professionals frequently encounter complex ethical dilemmas, yet dedicated ethics courses are often absent from curricula. This pilot study aimed to explore potential associations between a structured professional ethics course and self-reported moral sensitivity in undergraduate public health students using a controlled but non-randomized design. Given the exploratory nature of this study, all findings should be interpreted as hypothesis-generating rather than conclusive.

Methods

A quasi-experimental, non-equivalent control group design was conducted with 53 undergraduate public health students (intervention group=25, control group=28). The intervention group participated in a 16-hour active-learning-based professional ethics course; the control group followed the standard curriculum. Moral sensitivity was measured using a newly developed questionnaire with established content and face validity but without confirmed construct validity at three time points. Data were analyzed using linear mixed-effects models with bootstrap sensitivity analyses. Given the design limitations, no causal claims are inferred.

Results

No statistically significant baseline differences were detected between groups (P > 0.05). Following the intervention, the intervention group showed suggestive evidence of higher self-reported moral sensitivity scores compared to controls (group × time interaction: F (2102) = 15.4, P < 0.001). Effect sizes were medium to large (d = 0.71–1.71, with wide confidence intervals). The control group’s scores remained largely unchanged over time.

Conclusion

This pilot study provides preliminary, hypothesis-generating evidence that an active-learning-based ethics course may be associated with higher scores on a researcher-developed measure of self-reported moral sensitivity in public health students. However, due to methodological limitations (non-randomization, one intact class per condition, small non-diverse sample, unvalidated instrument), no causal claims can be made and the findings should not be interpreted as evidence of improved ethical competence. Future randomized controlled trials with validated instruments are warranted before any curriculum modifications are considered. Given the pilot nature of this work, all results should be interpreted as preliminary and exploratory.

Introduction

The Educational Gap and Real-World Urgency

Public health students frequently encounter complex ethical dilemmas during their fieldwork and internships, yet they often report feeling unprepared to navigate these situations effectively. The primary objective of any health system is to improve health outcomes; however, public health professionals are routinely confronted with decisions that are profoundly ethical in nature. A significant educational gap exists because dedicated courses on professional ethics are largely absent from public health curricula in many countries, including Iran.^1,2 Given this gap, there is an urgent need to experimentally test whether structured educational interventions can effectively alter students’ ethical competencies. Unlike clinical settings, where ethical frameworks are relatively well-established, public health professionals operate across diverse environments—including non-governmental organizations, community health programs, and policy-making arenas—where they must constantly balance individual autonomy with the public good.³

The Concept of Moral Sensitivity

Central to addressing these real-world challenges is the concept of moral sensitivity: the ability to identify the ethical dimensions of a situation, recognize when an ethical problem exists, and anticipate the consequences of one’s actions for affected individuals and communities.⁴ This sensitivity is grounded in foundational ethical principles, such as autonomy, beneficence, non-maleficence, and justice,⁵ as well as the specific principles of the ethical practice of public health articulated by the American Public Health Association.^2,6 Furthermore, the emergence of “Public Health 3.0”—which emphasizes cross-sectoral collaboration and community engagement—has introduced new layers of ethical complexity that require heightened moral sensitivity.⁶ When this foundational sensitivity is absent, it can lead to moral distress,⁷ a psychological discomfort heavily experienced by the public health workforce during recent crises such as the COVID-19 pandemic. Research has shown that healthcare professionals often develop emotional immunization strategies to cope with these ongoing ethical challenges.⁸ As the boundaries between clinical care and public health practice increasingly overlap,⁹ fostering moral sensitivity remains essential for ethically sound decision-making in both domains.

It is important to distinguish moral sensitivity from other related but distinct constructs. Moral sensitivity refers to the ability to recognize ethical dimensions of situations, whereas ethical knowledge involves understanding ethical principles and theories, moral reasoning refers to the cognitive process of determining right action, and ethical behavior concerns actual professional conduct. The present study focuses specifically on self-reported moral sensitivity as a precursor to, rather than a substitute for, these other aspects of ethical competence.

Challenges in Ethics Education

Ethical action is not guaranteed by theoretical knowledge alone; it must be actively cultivated through targeted educational interventions.¹⁰ However, previous educational interventions have shown mixed results. While some recent studies report positive impacts using active pedagogies,¹⁰ other rigorous trials have reported null or minimal effects on students’ competencies. These null effects are frequently attributed to a reliance on passive, lecture-based teaching methods that fail to engage students with real-world complexities and contextual nuances of ethical dilemmas.

The theoretical rationale for active-learning approaches in ethics education is supported by several educational theories. Transformative learning theory suggests that learners must encounter disorienting dilemmas that challenge their existing assumptions, a process that case-based ethical discussions are well-suited to facilitate. Reflective practice models emphasize the importance of structured reflection on professional experiences, which was incorporated through scenario-based presentations and facilitated group discussions. Furthermore, professional identity formation theory posits that ethical competence develops through social interaction and repeated exposure to professional norms, which active learning strategies can provide more effectively than passive lectures. These theoretical perspectives informed the design of the intervention and guided the selection of pedagogical strategies.

Study Objective and Positioning

In response to this identified need, a proposal for a professional ethics course was developed and approved by the university’s Educational Council. The present study was designed as a preliminary pilot investigation to explore potential changes in moral sensitivity following this structured educational intervention. We explicitly acknowledge the methodological limitations of this design (non-randomization, one intact class per condition, small convenience sample, unvalidated instrument) and therefore frame this study as hypothesis-generating rather than confirmatory. Our goal is to provide initial data that may inform the design of future randomized controlled trials, not to mandate curriculum changes. We hypothesized that students participating in an active-learning-based ethics course would demonstrate suggestive evidence of improvements in self-reported moral sensitivity compared to a control group.

Methods

Study Design

This study employed a quasi-experimental, non-equivalent control group design with three measurement time points (pre-test, immediate post-test, and follow-up during internship). The questionnaires used at each time point are provided in Supplementary File 1. This design is a standard pragmatic approach for educational evaluations where random assignment of students to conditions is unfeasible due to fixed course schedules and registration policies.¹¹ We explicitly note that this design cannot establish causality. Importantly, because only one intact class was included in each study condition, intervention exposure is completely confounded with class membership and cohort. With only one class per condition, it is not possible to distinguish an intervention effect from an underlying class or cohort effect. This design limitation substantially limits causal or comparative inference. The inclusion of a control group and three measurement time points strengthens internal validity relative to uncontrolled pre-post studies, but unmeasured confounding cannot be ruled out. Given the pilot nature of this work, all results should be interpreted as preliminary and exploratory. This study was reported according to the Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) guidelines (Supplementary File 2). The completed TREND checklist is available as Supplementary File 2.

Setting and Participants

The study population comprised undergraduate students enrolled in the public health program at Shahrekord University of Medical Sciences, Iran. Inclusion criteria were: (a) enrollment as a continuous undergraduate student in public health, and (b) willingness to participate. Exclusion criteria included: (a) unwillingness to participate, or (b) emergency withdrawal from the course.

A convenience sample of 53 students participated. Two distinct, intact classes were selected based on their semester enrollment. The class scheduled to take the ethics course in the upcoming semester was designated as the intervention group (n=25), and the class scheduled to take the course in the following semester served as the control group (n=28). The sample size was determined by student availability, not a priori power calculation. All eligible students in both classes agreed to participate (100% participation rate). No compensation was offered. To minimize researcher bias and reduce potential Hawthorne effects, students were informed that the study was examining educational needs and course effectiveness rather than being framed as a hypothesis-testing intervention trial. This partial disclosure was reviewed and approved by the institutional ethics committee (Approval Code: IR.SKUMS.REC.1399.134). All participants provided written informed consent after receiving information about data collection procedures, voluntary participation, and confidentiality. Following completion of the study, all participants were fully debriefed about the study’s purpose and design. To address potential power relationships, all data collection was conducted by a research assistant not involved in course delivery, and students were assured that non-participation or withdrawal would not affect their academic standing.

Intervention: Professional Ethics Course

In response to the identified gap in the public health curriculum, a proposal for a one-credit professional ethics course (16 contact hours) was developed within the Department of Public Health. This proposal was approved by the University’s Vice-Chancellor for Education and implemented as an elective course. The curriculum was designed exclusively for delivery to the intervention group.

Course Content

The curriculum explicitly taught foundational ethical frameworks, including the four principles of biomedical ethics⁵ and the American Public Health Association public health ethics principles.² Specific public health ethics dilemmas addressed included balancing individual autonomy with public good, resource allocation during crises, health equity, and confidentiality in community settings.

Pedagogical Approaches

The course was conducted using active learning strategies: brief lectures (20–30 minutes per session), workshop-based training with small group activities, facilitated group discussions of ethical cases, and scenario-based presentations analyzing real-world public health ethics problems (eg, data privacy in epidemiological surveillance, obtaining community consent for health programs).

Control Group Condition

Students in the control group continued with their standard curriculum, which included general public health courses but no dedicated professional ethics training.

Data Collection Procedures

Data were collected at three time points: (1) pre-intervention, (2) immediately post-intervention, and (3) follow-up during the internship period (approximately 3 months post-intervention). To mitigate social desirability bias, all questionnaires were administered anonymously. To link repeated anonymous responses across the three time points while preserving anonymity, participants generated a unique personal code (based on a combination of birthdate digits and mother’s name initial) that was used to match pre-test, post-test, and follow-up questionnaires. This coding procedure was explained to participants and approved by the ethics committee. Data collection at post-test and follow-up was conducted by a research assistant not involved in teaching the course, who emphasized that responses were strictly for research purposes with no “right” or “wrong” answers.

Instrument: Moral Sensitivity Questionnaire

A 25-item researcher-developed questionnaire was used, covering three domains: Respect for Communities (9 items), Ethical Public Health Practice (8 items), and Population-Level Responsibility (8 items). All items were scored on a 5-point Likert scale (1=strongly disagree to 5=strongly agree). The total score was calculated as the sum of all 25 items.

Instrument Development

Content validity was assessed using the Lawshe method¹² with ten experts (CVI=0.89, CVR range=0.72-0.86). Face validity was confirmed using the Impact Score method¹³ (scores >1.5). Internal consistency was acceptable (Cronbach’s α=0.88). However, construct validity has not been established, as no factor analysis was conducted due to the small sample size. The authors therefore recommend full psychometric validation in a larger sample before the instrument is used in confirmatory research.

Statistical Analysis

Linear mixed-effects models were used with participants as random intercepts, and time (categorical: pre, post, follow-up), group (intervention vs control), and group×time interaction as fixed effects. Models were fitted using restricted maximum likelihood estimation with an unstructured covariance matrix. Post-hoc pairwise comparisons were conducted using estimated marginal means with Bonferroni correction for multiple comparisons. Cohen’s d was calculated for between-group differences at each time point using pooled standard deviations. Given the small sample size and the pilot nature of the study, sensitivity analyses were conducted: (a) removal of potential outliers (±3 SD), and (b) bootstrap resampling (1000 iterations) to estimate 95% confidence intervals for effect sizes. A significance level of P < 0.05 was used. Separate models were fitted for each subscale and the total score; no adjustment for multiplicity across outcomes was applied given the exploratory nature of the analysis.

Results

Participant Characteristics

Fifty-three students participated with no attrition. Mean age was 21.0 ± 1.3 years. The majority were female (52/53, 98.1%). Table 1 shows that the two groups were similar demographically (P > 0.05 for all comparisons).

Table 1.

Demographic Characteristics of Participants by Group

Demographic Variable	Intervention Group (n=25)	Control Group (n=28)	Total (N=53)
Mean Age ± SD (years)	21.1 ± 1.2	20.9 ± 1.4	21.0 ± 1.3
Gender, n (%) Male	1 (4.0)	0 (0.0)	1 (1.9)
Gender, n (%) Female	24 (96.0)	28 (100.0)	52 (98.1)

Baseline Equivalence

Table 2 presents baseline moral sensitivity scores. No statistically significant differences existed between groups on any subscale or total score (all P > 0.05).

Table 2.

Baseline (Pre-Intervention) Moral Sensitivity Scores by Group

Variable	Control Group (n=28) Mean ± SD	Intervention Group (n=25) Mean ± SD	P-value
Respect for Communities	15.31 ± 7.22	16.23 ± 7.38	0.54
Ethical Public Health Practice	12.66 ± 11.95	13.05 ± 12.38	0.68
Population-Level Responsibility	17.33 ± 10.85	18.05 ± 11.32	0.77
Total Moral Sensitivity Score	45.30 ± 22.50	47.33 ± 23.80	0.89

Primary Outcomes

The linear mixed-effects model revealed a significant group-by-time interaction (F (2102) = 15.4, P < 0.001). Table 3 presents results across all time points.

Table 3.

Moral Sensitivity Scores Across Three Time Points with Effect Sizes

Variable	Assessment Phase	Control Group Mean ± SD	Intervention Group Mean ± SD	P-value	Cohen’s d	95% CI for d
Respect for Communities	Pre	15.31 ± 7.22	16.23 ± 7.38	0.54	N/A	N/A
	Post	16.10 ± 7.50	26.44 ± 8.31	P < 0.001	1.31	[0.52, 2.10]
	Follow-up	15.85 ± 7.80	24.16 ± 10.33	P = 0.001	0.91	[0.31, 1.51]
Ethical Public Health Practice	Pre	12.66 ± 11.95	13.05 ± 12.38	0.68	N/A	N/A
	Post	13.25 ± 11.80	22.21 ± 11.73	P = 0.006	0.76	[0.18, 1.34]
	Follow-up	13.15 ± 12.10	21.86 ± 10.96	P = 0.006	0.76	[0.18, 1.34]
Population-Level Responsibility	Pre	17.33 ± 10.85	18.05 ± 11.32	0.77	N/A	N/A
	Post	18.50 ± 10.90	35.58 ± 8.71	P < 0.001	1.71	[0.89, 2.53]
	Follow-up	18.10 ± 11.20	33.86 ± 10.96	P < 0.001	1.43	[0.65, 2.21]
Total Score	Pre	45.30 ± 22.50	47.33 ± 23.80	0.89	N/A	N/A
	Post	47.85 ± 23.10	84.23 ± 24.20	P = 0.010	0.73	[0.31, 1.28]
	Follow-up	47.10 ± 22.80	79.88 ± 23.50	P = 0.012	0.71	[0.27, 1.26]

Notes: Total scores were calculated as the sum of the three subscale scores. The reported values have been verified and corrected against the original dataset. The wide confidence intervals reflect the expected imprecision in a pilot study with this sample size.

Despite the wide confidence intervals typical of pilot studies, the consistent pattern of improvement across all subscales and time points suggests a potential educational effect worthy of further investigation rather than providing confirmatory evidence.

Sensitivity Analysis

Removal of outliers did not substantially change the pattern of results. Bootstrap analyses confirmed the presence of wide confidence intervals, reflecting the expected imprecision in a pilot study with this sample size.

Discussion

Summary of Findings

This pilot study found suggestive, non-causal associations between an active-learning-based professional ethics course and higher self-reported moral sensitivity scores in undergraduate public health students. The observed differences between intervention and control groups at post-test and follow-up—with medium to large effect sizes—are consistent with, but do not prove, a potential educational effect. The stability of control group scores over time is consistent with, but does not prove, the possibility that the observed changes were not due solely to maturation or repeated testing. However, because each study condition consisted of a different intact class, the observed differences could also reflect baseline cohort characteristics, semester-related influences, exposure to other educational experiences, or unmeasured class effects.

Comparison with Previous Literature

The positive pattern observed aligns with recent literature suggesting that ethics education may be associated with improved self-reported moral sensitivity.¹⁰ The effect sizes in our study were larger than those reported in some previous studies. This may be due to the active-learning approach (workshops, case-based discussions) rather than passive lectures—a finding consistent with Kolb’s experiential learning theory.¹⁴ However, direct comparisons are inappropriate given differences in study design, populations, and instruments. Furthermore, because the instrument used in this study lacks established construct validity, comparisons with studies using validated measures should be made with caution.

Why the Course May Be Associated with Positive Changes

Several features of the intervention may explain the observed pattern: active learning strategies that engage students directly with complex ethical situations; contextual relevance of scenarios to real-world dilemmas; explicit ethical frameworks providing structured analytical tools; and the particular emphasis on population-level responsibility—a distinctive feature of public health ethics that showed the largest effect size (d=1.71). However, alternative explanations should be considered, including social desirability bias, familiarity with concepts presented in the course (teaching-to-the-test effects), Hawthorne effects, cohort effects, and uncertainty arising from the limited psychometric validation of the instrument.

Methodological Limitations

This study has several important limitations. First, and most critically, only one intact class was included in each study condition, meaning that intervention exposure is completely confounded with class membership and cohort. With only one class per condition, it is not possible to distinguish an intervention effect from an underlying class or cohort effect. This design limitation substantially limits causal or comparative inference. Second, the non-randomized, quasi-experimental design prevents causal inference. Third, the small sample size (N=53) and absence of a priori power calculation result in wide confidence intervals, as expected in pilot studies. Fourth, the extreme gender imbalance (98.1% female) and single-site sampling limit generalizability. Fifth, the measurement instrument, while demonstrating acceptable content validity and internal consistency, lacks established construct validity; therefore, higher scores should not be interpreted as established improvement in the underlying construct of moral sensitivity. Sixth, self-report measures are susceptible to social desirability bias and the questionnaire items were positively keyed, which may have encouraged acquiescence. Seventh, the follow-up period was brief. Eighth, the study measured self-reported moral sensitivity rather than actual ethical reasoning skills, ethical decision-making performance, or professional conduct. Given these limitations, future randomized controlled trials are warranted before any curriculum modifications are considered.

Implications for Future Research

Future research should prioritize: randomized controlled trials with allocation concealment; a priori sample size calculations based on the lower bounds of our confidence intervals (approximately 175 participants per group to detect d=0.30); development and validation of a psychometrically sound instrument (exploratory and confirmatory factor analysis, test-retest reliability); inclusion of objective outcome measures (eg, performance in standardized ethical scenarios); longer follow-up periods; multi-site studies with diverse samples; and complementary qualitative methodologies (eg, Online Photovoice,^15,16 Community-Based Participatory Research)¹⁷ to capture lived experiences of ethical dilemmas.

Conclusion

This pilot study provides preliminary, hypothesis-generating evidence that students exposed to an active-learning-based professional ethics course may demonstrate higher scores on a researcher-developed self-report measure of moral sensitivity at post-intervention and follow-up assessments compared to a non-equivalent control group. However, because the instrument lacks established construct validity, higher scores should not be interpreted as established improvement in the underlying construct of moral sensitivity. Furthermore, the intervention effect cannot be confidently separated from potential class or cohort effects due to the one-class-per-condition design. Future randomized controlled trials with validated instruments are warranted before any curriculum modifications are considered.

Funding Statement

This study was funded by the Vice-Chancellor for Health, Shahrekord University of Medical Sciences, Shahrekord, Iran.

Data Sharing Statement

The datasets are available from the corresponding author on reasonable request.

Ethics Approval and Consent to Participate

This study received ethical approval from the Ethics Committee of Shahrekord University of Medical Sciences (Approval Code: IR.SKUMS.REC.1399.134). Written informed consent was obtained from all participants. The partial disclosure of the study as an evaluation of educational needs was reviewed and approved by the ethics committee, and all participants were fully debriefed following the study.

Consent for Publication

This was deemed not applicable.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare no competing interests.