Changes in medical students’ perceptions of ethical climate during their medical training: a follow-up study of a student cohort Authors: Dora Vukorepa, Mariano Kaliterna, Ružica Bočina, Marin Viđak, Ana Marušić, Marija Franka Žuljević Published: 2026 ABSTRACT BackgroundEthical climate is an important component of professional development in medical education, yet little is known about how medical students’ perceptions of ethical climate evolve during training or whether these perceptions are associated with career intentions. This study aimed to examine longitudinal changes in perceived and desired ethical climate among medical students and to explore potential associations with intended residency choice.MethodsWe conducted a follow-up study of a student cohort among medical students enrolled in the Croatian-language medical program at the University of Split School of Medicine. Ethical climate was assessed using the Ethical Climate Questionnaire (ECQ), which evaluates nine ethical climate domains. Data were collected during the third and the sixth year of study. Students completed both perceived and desired versions of the ECQ. Paired analyses were performed for students who participated at both time points, while independent-sample analyses compared third and sixth-year cohorts. Ethical climate domains were also examined in relation to students’ desired residency choice.ResultsA total of 70 third year and 77 sixth-year students participated, with 57 students included in paired analyses. The two dominant climates were Company profit and Laws and professional codes. Perceived ethical climate scores decreased significantly from the third to the sixth year in the Company rules and procedures (p < 0.001), while other domains remained stable. No statistically significant differences were observed in desired ethical climate domains between the third and sixth study years after Bonferroni correction. No associations were found between ethical climate domains and intended residency choice.ConclusionsMedical students’ perceptions of institutional ethical climate change during undergraduate training, particularly regarding organizational structure, while ethical ideals remain largely stable. Ethical climate perceptions do not appear associated with residency choice, suggesting that ethical development and career decision-making may follow distinct trajectories. These findings highlight the importance of addressing the hidden curriculum and supporting structured reflection on ethical tensions encountered during clinical training.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12909-026-09494-z. Background Ethical climate is viewed as a type of organizational work climate reflecting organizational practices, procedures, and policies with moral consequences [18], and it serves as a descriptive map of ethical decision-making and actions within an organization, based on philosophical and moral theories [15]. Ethical climate has been investigated in different settings, mostly in business [6], but also in healthcare [1, 24] and higher education institutions [17, 23, 26]. Because ethical climate is a complex and subtle phenomenon, the question remains how changes in ethical climate at a workplace can be assessed and measured [11]. Ethical climate is usually measured using ethical climate questionnaires, among which the Ethical Climate Questionnaire (ECQ), developed by Victor and Cullen and later refined to its current 36-item version, is one of the most widely used instruments [1, 6]. Within medical education, ethical climate is particularly relevant, as students undergo professional socialization in complex clinical environments where formal ethical standards interact with organizational constraints, hierarchical structures, and everyday practice [13, 26]. In this study, ethical climate refers to organizational ethical climate, that is, shared perceptions of institutional practices, procedures, and policies with moral consequences, rather than medical ethics in the narrower sense of ethical issues arising within the physician–patient relationship. Previous research in organizational settings showed that perceptions of ethical climate influence both decision-making and subsequent behaviour in response to ethical dilemmas [18]. In business settings, a statistically significant negative relationship between organizational ethical climate and the seriousness of ethical violations has been reported, as well as a positive relationship between ethical climate and success in responding to ethical issues [3]. In healthcare contexts, ethical climate has additionally been associated with job satisfaction and professional well-being [1, 16, 25]. Although ethical climate has been measured in a variety of professional settings, relatively few studies have examined ethical climate in institutions of higher education [17]. Cross-sectional studies conducted at the University of Split demonstrated that students from different faculties and disciplines, although not formal employees, perceive organizational ethical climate similarly to staff, indicating that institutional affiliation provides comparable insight into organizational functioning [26]. Furthermore, faculties from different disciplines exhibited similar dominant ethical climates, largely based on rules and laws, suggesting organizational similarity within a publicly funded university [17, 26]. Recent hospital-based research in the same regional context also identified rule- and law-oriented climates as dominant, underscoring the institutional nature of ethical environments [14]. There are not many studies testing interventions for changing organizational climate [26]. An important aspect of organizational change is harmonization between the perceived and desired ethical climate [4]. However, most ECQ studies focus exclusively on perceived ethical climate, with limited attention to desired ethical climate or to longitudinal change [8]. This study builds on the premise that students perceive organizational ethical climate in a manner comparable to employees [26]. However, it remains unclear whether these perceptions change during medical training as students’ progress from preclinical to clinical education. Moreover, longitudinal evidence on ethical climate development in student populations is scarce, despite previous findings of changes in moral reasoning during medical education [13]. Therefore, this follow-up study of a student cohort aimed to examine perceived and desired ethical climate among medical students during their third and sixth study years. In addition, we explored whether changes in ethical climate dimensions were associated with students’ intended residency choice. Methods Study design and setting We conducted a follow-up study of a student cohort at the University of Split School of Medicine, Croatia. The study was conducted exclusively among Croatian medical students at the University of Split School of Medicine, Croatia, which is a publicly funded institution with an integrated undergraduate and graduate six-year medical program. Data collection The study followed a single cohort of medical students enrolled in the academic year 2019/2020, with data collected at three time points: during the first, third, and sixth year of study. Data collection began in the academic year 2019/2020, when participants were first-year medical students in the medical program at the University of Split School of Medicine. The second wave of data collection was conducted in the academic year 2021/2022, when the same group of participants had reached the third year of study. The final data collection was completed in the academic year 2024/2025, when the participants were in their sixth year of study. All surveys were administered during scheduled teaching sessions within the courses held by the Department of Medical Humanities. Data collection was carried out exclusively online using the SurveyMonkey platform, which ensured standardized administration and facilitated access for all students (Survey-Monkey Inc., San Mateo, CA, USA). The present analysis focused on the third- and sixth-year measurements because these time points represent two key stages of the medical curriculum: the end of the predominantly preclinical phase and the final year of clinical training. This comparison allowed us to examine changes in students’ ethical climate perceptions across an important period of professional socialization during undergraduate medical education. Participants The study population consisted of all medical students who enrolled in the Croatian-language medical program at the University of Split School of Medicine in the academic year 2019/2020. To be eligible for inclusion, students had to provide informed consent and participate in the online survey at third and/or sixth year of the study. Participants were excluded if more than two questionnaire items were missing from their survey (within any ECQ sub-scale). In cases where only a single response was missing within the Ethical Climate Questionnaire (ECQ), the missing value was imputed using the mean of the corresponding item column across the cohort. This approach ensured that minimal missingness did not reduce the overall sample size or bias the results. Consequently, the longitudinal analyses were conducted on data from the third and sixth study years. Initially, all third-year medical students enrolled in the Croatian-language program at the University of Split School of Medicine, who began their studies in the academic year 2019/2020, were invited to participate in the study. Survey The questionnaire used in this study consisted of three sections. The first section collected demographic data, including participants’ age, gender, and year of study. In the final (sixth) year of study, the questionnaire included an additional set of questions in which students were asked to report their preferred medical residency, whether their preference had changed since the beginning of their studies (and, if so, to specify their initial choice), as well as their current grade point average. Perceptions of the organizational ethical climate were assessed using the Ethical Climate Questionnaire (ECQ) developed by Victor and Cullen [6]. The instrument, previously translated into Croatian, back-translated, and validated in higher education settings, consists of 36 Likert-type items (0–5 scale). The ECQ evaluates nine theoretical climate types: 1) Self-interest, 2) Company profit, 3) Efficiency, 4) Friendship, 5) Team interest, 6) Social responsibility, 7) Personal morality, 8) Rules and procedures, and 9) Laws and professional codes. Each climate type is represented by four items (score range 0–20), with the highest-scoring climate type considered dominant. Higher scores indicate stronger agreement with the values represented by each ethical climate type signifying either their greater perceived presence within the current institutional environment (perceived ECQ) or their greater desirability as features of an ideal workplace climate (desired ECQ). Students completed two versions of the ECQ at each measurement point: one assessing the perceived climate and another assessing the desired climate. Due to the large number of distinct specialties, responses were grouped into five broader categories for analytical purposes: 1) internal medicine–related, 2) surgical, 3) general practice, 4) neuropsychiatric, and 5) diagnostic and other specialties (e.g., radiology, pathology, cytology). Students who reported an interest in non-clinical careers (e.g., public health, epidemiology, medical education, research, or administrative medicine) were classified as a separate category labelled non-clinical/academic. This classification follows prior research grouping non-patient-care fields independently from clinical specialties [5, 10]. The English version of the full survey given to participants is provided in Appendix 1. Ethical aspects To allow pairing of responses while preserving confidentiality, participants generated a unique six-character alphanumeric code based on personal identifiers (the first letter of their mother’s and father’s names, first letter of their own name, the last two digits of their birth year, and first letter of birthplace). This pseudonymization method enabled matching across time points. Once datasets were matched, all codes were removed and the data analysed in anonymized form. The study was approved by the Ethics Committee of the University of Split School of Medicine in accordance with the Declaration of Helsinki (Class No.: 003–08/20–03/0005, Reg. No.: 2181–198-0304–20–0068). At the beginning of the online questionnaire, all participants were presented with an information sheet describing the purpose and procedures of the study, as well as assurances of confidentiality and anonymity. Consent to participate was obtained electronically: participants were required to mark a checkbox confirming that they had read the information and agreed to take part in the study. Before completing the questionnaire, students were informed about the purpose of the study, participation was voluntary, and no additional instruction or educational intervention regarding ethical climate was provided before survey completion, in order to avoid influencing participants’ responses. Statistical analysis All data were analyzed using Jamovi software, version 2.3.28 (The jamovi project, Sydney, Australia). Descriptive statistics were calculated for all variables obtained at the third- and sixth-year measurements, including measures of central tendency and dispersion. For all longitudinal analyses, we matched all questionnaires that had complete responses across both measurement points, thereby constructing a fully paired dataset for within-subject comparisons over time. The data were, therefore, organized in two complementary formats to enable both cross-sectional and longitudinal analyses. The first (“vertical”) dataset included all available responses from the third- and sixth-year cohorts and was used for independent-sample comparisons between study years (Mann–Whitney U test). The second (“horizontal”) dataset contained only paired observations from students who completed the ECQ at both time points, with each row representing a single student’s scores from the third and sixth years. This dataset was used for within-subject analyses of longitudinal changes (Wilcoxon signed-rank test). The normality of data distribution was assessed using the Shapiro–Wilk test, which indicated non-normal distribution for most ECQ domain scores. Therefore, non-parametric tests were used for further analyses. To examine within-subject changes in ethical climate perceptions between the third and sixth study years, the Wilcoxon signed-rank test was applied to the paired dataset (n = 57). Differences between independent groups of third- and sixth-year students were analyzed using the Mann–Whitney U test. Descriptive statistics were expressed as medians and interquartile ranges (IQR), while categorical variables were presented as frequencies and percentages. Statistical significance was set at p < 0.05 for all analyses. Effect sizes were calculated for both paired and independent comparisons using rank-biserial correlation and are reported alongside p-values to provide information on the magnitude of observed differences. Results Participant flow and descriptive statistics Initially, all third-year medical students enrolled in the Croatian-language program at the University of Split School of Medicine, who began their studies in the academic year 2019/2020, were invited to participate in the study. At the third year of study, 72 students completed the questionnaire, of whom two were excluded due to incomplete data, resulting in 70 valid responses. During the second data collection point, conducted in the sixth year of the medical program, a total of 77 students participated. After matching the datasets from both time points, 57 students were identified as paired respondents and included in the longitudinal analysis (Fig. 1). The paired sample was smaller than the total number of valid responses because paired analyses included only students who completed the ECQ at both time points and whose responses could be reliably matched using the study identification code.Fig. 1Flow chart of survey responses Flow chart of survey responses The median age of participants was 21 years in the third year and 24 years in the sixth year, consistent with the longitudinal design. Women comprised 63% of the third-year sample and 73% of the sixth-year sample. The median grade point average (GPA) among sixth-year students was 4.2 (IQR 4.1–4.45) (Table 1).Table 1Descriptive statistics of all included study participantsVariable3rd year (n = 70)6th year (n = 77)Age, median (IQR)21 (21–22)24 (22–25)Female, n (%)44 (62,8)56 (72,7)GPA, median (IQR)-4.2 (4.1–4.45)Completed both surveys, n (%)57 (75.0%) Table 1Descriptive statistics of all included study participantsVariable3rd year (n = 70)6th year (n = 77)Age, median (IQR)21 (21–22)24 (22–25)Female, n (%)44 (62,8)56 (72,7)GPA, median (IQR)-4.2 (4.1–4.45)Completed both surveys, n (%)57 (75.0%) Descriptive statistics of all included study participants A total of 77 sixth-year students indicated their preferred specialty choice. Internal medicine-related fields were the most frequently selected, chosen by 26 students (34%). Surgical specialties were similarly well represented, with 24 students (31%) expressing interest in this category. Neuropsychiatric specialties were selected by 11 students (14%), while diagnostic and other fields, including radiology and pathology, were chosen by 14 students (18%). Only one student reported a preference for general practice, and one student indicated interest in a non-clinical or academic career path (full results listed by each residency are provided in Appendix 2, Table S1). Ethical climate results In the independent sample analysis, Company profit emerged as the highest-scoring perceived ethical climate domain at both time points, while Company profit also represented the dominant domain in the desired ethical climate. In contrast, among students with paired data, Laws and professional codes constituted the dominant perceived ethical climate, whereas Company profit remained the highest-scoring domain in the desired ethical climate. Descriptive data for all nine ECQ domains are presented in Tables 2 and 3.Table 2Changes in perceived and desired ethical climate domains between the third and sixth year of study among paired respondentsVariableYear 3 (n = 57)Year 6 (n = 57)p*Effect size†Md (IQR)rank-biserial rPerceived ethical climate: Self interest13.0 (10.0–14.0)12.0 (11.0–14.0)0.9880.003 Friendship12.0 (9.0–15.0)12.0 (10.0–14.0)0.6560.072 Personal morality11.0 (8.0–12.0)11.0 (9.0–12.0)0.939−0.013 Company rules and procedures12.0 (10.0–14.0)9.0 (7.0–10.0)< 0.0010.785 Team interest13.0 (12.0–14.0)12.0 (10.0–14.0)0.1000.263 Company profit16.0 (13.0–17.0)14.0 (12.0–16.0)0.0050.453 Efficiency15.0 (13.0–16.0)14.0 (12.0–16.0)0.0920.275 Social responsibility15.0 (12.0–16.0)14.0 (11.0–15.0)0.2720.174 Laws and professional codes16.0 (13.0–17.0)15.0 (12.0–16.0)0.2180.209Desired ethical climate: Self interest10.0 (8.0–13.0)11.0 (8.0–13.0)0.134−0.243 Friendship15.0 (13.0–18.0)14.0 (12.0–17.0)0.4970.111 Personal morality11.0 (8.0–13.0)11.0 (8.0–12.0)0.2490.195 Company rules and procedures10.0 (8.0–13.0)10.0 (8.0- 13.0)0.4840.107 Team interest16.0 (13.0–18.0)14.0 (11.0–18.0)0.2510.188 Company profit17.0 (14.0–18.0)16.0 (13.0–17.0)0.0360.339 Efficiency16.0 (14.0–17.0)14.0 (12.0–16.0)0.0780.297 Social responsibility16.0 (14.0–18.0)16.0 (13.0–18.0)0.758−0.052 Laws and professional codes16.0 (14.0–18.0)16.0 (14.0–18.0)0.6700.071We performed a Bonferroni correction due to a large number of comparisons in the analysis, making the target P-value cut off for significance P = 0.003 (0.05/18)*Paired Wilcoxon rank t-test†Effect sizes for Wilcoxon signed-rank tests were reported as rank-biserial correlationsTable 3Comparison of perceived and desired ethical climate domain scores between third- and sixth-year studentsVariableYear 3 (n = 70)Year 6 (n = 77)p*Effect size†Md (IQR)rank-biserial rPerceived ethical climate: Self interest13.0 (10.0–14.0)12.0 (10.0–14.0)0.7330.032 Friendship12.0 (9.0–15.0)12.0 (9.0–15.0)0.9020.011 Personal morality11.0 (8.25–12.0)11.0 (9.0–12.0)0.7890.025 Company rules and procedures12.0 (10.0–14.0)9.0 (7.0–10.0)< 0.0010.564 Team interest13.0 (11.0–14.0)12.0 (10.0–14.0)0.1720.130 Company profit16.0 (13.0–17.0)14.0 (12.0–17.0)0.0890.162 Efficiency14.0 (13.0–16.0)14.0 (11.0–16.0)0.4650.069 Social responsibility14.5 (12.0–16.0)14.0 (11.0–16.0)0.2700.105 Laws and professional codes15.5 (13.0–17.0)15.4 (12.0–17.0)0.3930.081Desired ethical climate: Self interest10.0 (8.0–13.0)11.0 (8.0–13.0)0.2890.101 Friendship15.0 (12.3–18.0)15.0 (12.0–17.0)0.5260.060 Personal morality11.0 (8.0–12.0)11.0 (9.0–13.0)0.7800.026 Company rules and procedures10.0 (8.0–12.8)10.0 (8.0–12.0)0.3450.090 Team interest16.0 (12.3–18.0)14.0 (11.0–18.0)0.3410.090 Company profit17.0 (14.0–18.0)16.0 (13.0–18.0)0.1030.155 Efficiency16.0 (13.0–17.0)14.0 (12.0–16.0)0.0630.177 Social responsibility15.5 (13.0–17.8)16.0 (13.0–18.0)0.7610.029 Laws and professional codes15.5 (13.0–17.8)16.0 (13.0–18.0)0.4390.073We performed a Bonferroni correction due to a large number of comparisons in the analysis, making the target P-value cut off for significance P = 0.003 (0.05/18)*Independent samples Mann–Whitney U test†Effect sizes for Mann–Whitney U tests were reported as rank-biserial correlations Table 2Changes in perceived and desired ethical climate domains between the third and sixth year of study among paired respondentsVariableYear 3 (n = 57)Year 6 (n = 57)p*Effect size†Md (IQR)rank-biserial rPerceived ethical climate: Self interest13.0 (10.0–14.0)12.0 (11.0–14.0)0.9880.003 Friendship12.0 (9.0–15.0)12.0 (10.0–14.0)0.6560.072 Personal morality11.0 (8.0–12.0)11.0 (9.0–12.0)0.939−0.013 Company rules and procedures12.0 (10.0–14.0)9.0 (7.0–10.0)< 0.0010.785 Team interest13.0 (12.0–14.0)12.0 (10.0–14.0)0.1000.263 Company profit16.0 (13.0–17.0)14.0 (12.0–16.0)0.0050.453 Efficiency15.0 (13.0–16.0)14.0 (12.0–16.0)0.0920.275 Social responsibility15.0 (12.0–16.0)14.0 (11.0–15.0)0.2720.174 Laws and professional codes16.0 (13.0–17.0)15.0 (12.0–16.0)0.2180.209Desired ethical climate: Self interest10.0 (8.0–13.0)11.0 (8.0–13.0)0.134−0.243 Friendship15.0 (13.0–18.0)14.0 (12.0–17.0)0.4970.111 Personal morality11.0 (8.0–13.0)11.0 (8.0–12.0)0.2490.195 Company rules and procedures10.0 (8.0–13.0)10.0 (8.0- 13.0)0.4840.107 Team interest16.0 (13.0–18.0)14.0 (11.0–18.0)0.2510.188 Company profit17.0 (14.0–18.0)16.0 (13.0–17.0)0.0360.339 Efficiency16.0 (14.0–17.0)14.0 (12.0–16.0)0.0780.297 Social responsibility16.0 (14.0–18.0)16.0 (13.0–18.0)0.758−0.052 Laws and professional codes16.0 (14.0–18.0)16.0 (14.0–18.0)0.6700.071We performed a Bonferroni correction due to a large number of comparisons in the analysis, making the target P-value cut off for significance P = 0.003 (0.05/18)*Paired Wilcoxon rank t-test†Effect sizes for Wilcoxon signed-rank tests were reported as rank-biserial correlations Changes in perceived and desired ethical climate domains between the third and sixth year of study among paired respondents We performed a Bonferroni correction due to a large number of comparisons in the analysis, making the target P-value cut off for significance P = 0.003 (0.05/18) *Paired Wilcoxon rank t-test †Effect sizes for Wilcoxon signed-rank tests were reported as rank-biserial correlations Table 3Comparison of perceived and desired ethical climate domain scores between third- and sixth-year studentsVariableYear 3 (n = 70)Year 6 (n = 77)p*Effect size†Md (IQR)rank-biserial rPerceived ethical climate: Self interest13.0 (10.0–14.0)12.0 (10.0–14.0)0.7330.032 Friendship12.0 (9.0–15.0)12.0 (9.0–15.0)0.9020.011 Personal morality11.0 (8.25–12.0)11.0 (9.0–12.0)0.7890.025 Company rules and procedures12.0 (10.0–14.0)9.0 (7.0–10.0)< 0.0010.564 Team interest13.0 (11.0–14.0)12.0 (10.0–14.0)0.1720.130 Company profit16.0 (13.0–17.0)14.0 (12.0–17.0)0.0890.162 Efficiency14.0 (13.0–16.0)14.0 (11.0–16.0)0.4650.069 Social responsibility14.5 (12.0–16.0)14.0 (11.0–16.0)0.2700.105 Laws and professional codes15.5 (13.0–17.0)15.4 (12.0–17.0)0.3930.081Desired ethical climate: Self interest10.0 (8.0–13.0)11.0 (8.0–13.0)0.2890.101 Friendship15.0 (12.3–18.0)15.0 (12.0–17.0)0.5260.060 Personal morality11.0 (8.0–12.0)11.0 (9.0–13.0)0.7800.026 Company rules and procedures10.0 (8.0–12.8)10.0 (8.0–12.0)0.3450.090 Team interest16.0 (12.3–18.0)14.0 (11.0–18.0)0.3410.090 Company profit17.0 (14.0–18.0)16.0 (13.0–18.0)0.1030.155 Efficiency16.0 (13.0–17.0)14.0 (12.0–16.0)0.0630.177 Social responsibility15.5 (13.0–17.8)16.0 (13.0–18.0)0.7610.029 Laws and professional codes15.5 (13.0–17.8)16.0 (13.0–18.0)0.4390.073We performed a Bonferroni correction due to a large number of comparisons in the analysis, making the target P-value cut off for significance P = 0.003 (0.05/18)*Independent samples Mann–Whitney U test†Effect sizes for Mann–Whitney U tests were reported as rank-biserial correlations Comparison of perceived and desired ethical climate domain scores between third- and sixth-year students We performed a Bonferroni correction due to a large number of comparisons in the analysis, making the target P-value cut off for significance P = 0.003 (0.05/18) *Independent samples Mann–Whitney U test †Effect sizes for Mann–Whitney U tests were reported as rank-biserial correlations In the paired analysis between third- and sixth-year scores (Table 2), a statistically significant decrease was observed in the perceived Company rules and procedures (median [IQR]: third year 12.0 [10.0–14.0] vs. sixth year 9.0 [7.0–10.0]; p < 0.001) which remained significant after Bonferroni correction. No statistically significant differences were found for the remaining ECQ domains. In the independent sample analysis (Table 3), consistent findings were observed. Sixth-year students reported significantly lower scores in perceived Company rules and procedures (12.0 [10.0–14.0] vs. 9.0 [7.0–10.0]; p < 0.001). No significant differences were found for the remaining ECQ domains. In the paired analysis of desired ethical climate scores, no statistically significant differences were observed between the third and sixth years across any of ECQ domains after Bonferroni correction. Association of ethical climate and desired residency Our initial aim was to perform a multinomial logistic regression analysis to determine whether any of the ethical climate scores could serve as predictors of the desired residency choice in the sixth year. As a preliminary step, we compared each domain of the ethical climate questionnaire (both perceived and desired) across the different residency categories. No statistically significant differences were observed, and therefore there was no rationale to proceed with developing a multinomial regression model (data shown in Appendix 2,Table S2). Discussion Our findings indicate that students’ perceptions of the institutional ethical climate change over time, whereas their desired ethical climate remains largely stable. In contrast, no meaningful association was observed between ethical climate dimensions and students’ intended residency categories. Although Laws and professional codes constituted the dominant perceived ethical climate in the paired sample, the most pronounced longitudinal change was observed in perceived ethical climate domain related to institutional structure, specifically Company rules and procedures. By the sixth year of study, students reported a significantly lower perception of this domain compared with the third year. This decrease reflects a change in students’ perception of the institutional rule-based dimension of ethical climate, rather than a change in medical ethics in the narrower sense or in students’ personal morality. These findings are consistent with previous research conducted among healthcare professionals, suggesting that increased exposure to institutional realities is associated with more critical perceptions of formal rules, procedures, and organizational priorities [21]. Studies in clinical settings have shown that ethical climate perceptions, particularly those related to organizational structures, are sensitive to professional experience and contextual factors rather than static individual characteristics [2, 17]. Similarly, Viđak et al. reported Laws and professional codes as the dominant ethical climate across different faculties, supporting the stability of normative ethical frameworks at the institutional level [26]. As students’ progress through medical training and gain greater insight into everyday institutional functioning, they may become more aware of tensions between formal regulations and real-world clinical practice [16]. The observed decline in perceived Company rules and procedures may reflect early professional exposure to organizational inconsistencies and governance challenges, paralleling findings among Croatian healthcare professionals who report political interference, weak accountability, and ineffective management as key system-level problems [22]. Moreover, qualitative studies show that students often witness unprofessional behaviour and organizational decisions that conflict with formal ethical standards, reinforcing the idea that real world practice may diverge from textbook ethics and driving changes in perceptions of institutional ethical climate over time [7]. In contrast, ethical climate domains reflecting core moral and professional values such as Social responsibility, Personal morality, and Laws and professional codes remained relatively stable across study years. This pattern aligns with evidence from both healthcare and academic settings indicating that fundamental ethical orientations are relatively consistent and less sensitive to change over time [8]. A recent Croatian study of hospital employees (doctors, nurses, and other professionals) used the same ECQ instrument to evaluate ethical climate within a public healthcare system has consistently identified rule-based and law-oriented climates as dominant, suggesting that institutional ethics are often grounded in formalized normative frameworks rather than individual moral variability [14]. Desired ethical climate scores showed fewer and smaller changes compared with perceived ethical climate scores. This distinction suggests that students can differentiate between the ethical characteristics of their current institutional environment and the ethical environment in which they would ideally wish to work. Similar discrepancies between perceived and desired ethical climate have been described in healthcare settings and interpreted as indicators of ethical tension rather than ethical decline [21]. Contrary to our initial expectations, no association was found between ethical climate dimensions and residency choice. This finding contrasts with earlier studies suggesting that ethical attitudes may indirectly influence residency selection, particularly in relation to patient-centeredness or altruistic values [5]. However, our results are consistent with more recent research emphasizing that residency choice is primarily influenced by factors such as lifestyle considerations, workload expectations, perceived prestige, and exposure to role models rather than by institutional ethical climate [12, 26]. The lack of association observed in this study suggests that ethical climate perceptions and career decision-making may represent largely independent dimensions of professional development during undergraduate medical education. The lack of differences in ethical climate across residency categories in our study is conceptually similar to findings reported by Malički et al., who examined ethical climate across different academic disciplines. Although using a different measurement framework, Malički and colleagues assessed organizational ethics through dimensions such as fairness, consistency of rule application, leadership responsibility, and institutional support for ethical conduct. These constructs overlap conceptually with ECQ domains such as Company rules and procedures, Laws and professional codes, and Social responsibility. In both studies, ethical climate perceptions were largely uniform across subgroups within the same institution, suggesting that ethical climate is primarily an institutional characteristic rather than a function of disciplinary or career differentiation [17]. Our study adds to existing literature in several important ways. First, it applies a validated ethical climate instrument longitudinally within a cohort of medical students, an approach rarely used in previous research. Most existing ethical climate research has primarily relied on cross-sectional samples or focused on practicing healthcare professionals, thereby limiting insight into how ethical climate perceptions evolve during medical students’ professional development [14, 17]. Second, by examining both perceived and desired ethical climate, this study distinguishes students’ evaluation of their current institutional environment and their ethical ideals, offering a more detailed understanding of ethical development during medical training. Third, the lack of association between ethical climate and residency choice highlights a potential separation between ethical perception and career planning and suggests that these domains may follow distinct developmental pathways. The longitudinal design of this follow-up study of a student cohort strengthens the interpretation of changes over time by minimizing cohort effects and allowing within-cohort comparisons. However, several limitations should be acknowledged. As a follow-up study conducted at a single institution, causal relationships cannot be inferred, and generalizability may still be limited. Additionally, the relatively small sample size, particularly within certain residency categories, restricted the complexity of multivariable analyses and reduced statistical power. Ethical climate perceptions were assessed using self-report instruments, which may be influenced by response bias and individual interpretation. In addition, findings from a single-institution cohort may not generalize to other contexts, as ethical climate perceptions are shaped by broader institutional and societal environments. In Croatia specifically, population-based evidence indicates mixed levels of trust in the healthcare system and wider variability in institutional trust, which may influence how students interpret organizational priorities and formal rules during training [20]. These findings have important implications for medical education. The observed decline in perceptions of institutional structure and organizational priorities underscores the need for medical schools to critically reflect on how formal rules, procedures, and performance-driven cultures are experienced by students during clinical training [19]. Addressing perceived gaps between institutional values and everyday practice may help support ethical development and professional well-being. Structured opportunities for reflection such as facilitated discussion groups, narrative ethics, or mentored debriefings during clinical training may help students process discrepancies between formal rules and lived practice and transform ethical contradictions into aspects useful for professional growth [9]. Future research should continue to conduct longitudinal studies across multiple institutions to improve generalizability and explore contextual differences in ethical climate. Qualitative studies could provide deeper insight into how students interpret and negotiate ethical tensions encountered during training and in which way the encountered situations may affect their perception of the ethical climate in their surroundings. Additionally, future longitudinal studies might examine how ethical climate perceptions evolve during residency and early professional practice and how they interact with factors such as moral distress, burnout, or professional identity formation. Conclusion This follow-up study of a student cohort shows that medical students’ perceptions of institutional ethical climate change during undergraduate training, particularly in domains related to organizational structure, while their desired ethical climate remains largely stable. No association was observed between ethical climate dimensions and intended residency choice, suggesting that ethical perceptions and career planning may follow largely independent developmental pathways. These findings indicate that increasing clinical exposure is accompanied by greater awareness of organizational realities without a corresponding shift in ethical ideals. The results underscore the importance of addressing ethical climate explicitly within medical curricula, especially during clinical training. Future multi-institutional longitudinal studies incorporating qualitative approaches are needed to further examine ethical climate development during residency and early professional practice. Supplementary Information Supplementary Material 1. Supplementary Material 2. Supplementary Material 1. Supplementary Material 2.