The driving forces of research ethics in academia: insights from students based on the theory of planned behavior

Abstract

Background

Research ethics is crucial for protecting participants’ rights in medical studies. Although several studies have explored general awareness and attitudes toward research ethics among medical students, few have systematically examined the behavioral determinants of adherence to Research Ethical Codes (REC) using established theoretical models. This study addresses this gap by identifying the determinants of REC adherence among students at Kermanshah University of Medical Sciences (KUMS), applying the Theory of Planned Behavior (TPB) as a conceptual framework.

Methods

A descriptive-analytical study was conducted in 2024 among KUMS students. A cluster sampling approach was used, with structured questionnaires distributed to gather data on demographics and TPB constructs. Data were analyzed using SPSS, employing Pearson correlation and linear regression analyses.

Results

The study included 271 participants with a mean age of 23.99 years. The adherence level to ethical codes was 69.1%. The adjusted R² value was 0.352, indicating that 35.2% of the variance in adherence to REC behavior was explained by the model. Notably, attitude (B = 0.694, p < 0.001) and intention (B = 0.857, p = 0.002) were significant predictors. The highest adherence was for confidentiality (mean score = 4.04), while the lowest was for obtaining ethical approval before data collection (mean score = 2.68).

Conclusions

The findings indicate a 69.1% adherence to ethical research codes, with attitude and intention identified as key predictors of ethical behavior. These results emphasize the importance of targeted educational strategies to strengthen students’ commitment to ethical research practices.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12910-025-01315-0.

Background

The term research is commonly defined as a systematic and creative process aimed at generating new knowledge or applying existing knowledge in innovative ways to improve understanding and solve problems [1]. Over the past decade, medical research in developing countries has expanded significantly, driven by the urgent need to address local health challenges and reduce global health disparities [2]. Since health researchers involve human participants, fundamental ethical principles for protecting their rights, dignity, and welfare are deemed essential [3]. Some studies have shown that adherence to ethical principles in research in developing countries, including the Middle East, is suboptimal; in many of these studies, there was no plan for obtaining informed consent, or many essential elements of informed consent were omitted from the submitted consent forms. For example, El-Dessouky et al., in a study conducted in two dental schools in Saudi Arabia and Egypt found that less than half of the respondents were familiar with ethical principles in research, and less than one-third were aware of the functions of ethical codes in research [4]. However, despite these findings, there is limited evidence regarding the determinants of ethical research behavior among students in western Iran. This raises the research gap that the present study seeks to address: What are the behavioral factors influencing adherence to Research Ethical Codes (REC) among medical students in Kermanshah University of Medical Sciences (KUMS)?

Globally, biomedical research is governed by a set of standardized editorial and ethical requirements, most notably the updated 2025 recommendations of the International Committee of Medical Journal Editors (ICMJE). These guidelines—formerly known as the Uniform Requirements—continue to shape manuscript preparation and ethical publishing practices across scientific journals, emphasizing transparency, authorship accountability, and protection of research participants [5]. Research ethics charters, composed of formal codes and normative guidelines, define the ethical boundaries of research conduct. These frameworks serve multiple functions: they support students, faculty, and researchers in designing ethically sound proposals; guide ethics committees in evaluating research protocols; and provide a reference for resolving ethical dilemmas throughout the research lifecycle [6].

Foundational declarations such as the Nuremberg Code, the Declaration of Helsinki, and the Belmont Report have played a pivotal role in shaping modern biomedical ethics. Their shared goals include improving the quality and integrity of research involving human subjects, preventing harm, and establishing universal principles for informed consent, risk-benefit assessment, and participant autonomy [7]. In recent years, a growing body of research has sought to examine the level of awareness and understanding of ethical codes among members of academic and scientific institutions. Adjovi (2025) underscores that despite increased global attention to research ethics, there remains a limited grasp of the underlying factors that contribute to unethical behavior in scientific practice. This gap in understanding poses a significant challenge to the development of effective preventive strategies and highlights the need for targeted education, behavioral analysis, and the cultivation of a culture of integrity within research environment [8]. To promote higher education and research integrity, universities must foster a culture of ethical awareness and responsibility [9].

In this regard, the field of research ethics is a complex and ever-evolving topic, and it is essential for researchers to familiarize themselves with general ethical concepts and the information required to ensure approval and publication of research, as well as to avoid retraction of articles [10]. On the other hand, evidence indicates the usefulness of utilizing behavioral analysis models and theories to identify the most significant determinants affecting behavior [11]. One such model is the Theory of Planned Behavior (TPB), which has been widely applied to predict ethical and unethical behaviors in educational and research contexts [12–16]. TPB, proposed by Ajzen, posits that behavioral intention is the most immediate predictor of behavior. This intention is shaped by three key constructs: (a) Attitude toward the behavior – the individual’s positive or negative evaluation of performing the behavior; (b) Subjective norms – perceived social pressure from significant others to perform or not perform the behavior; and (c) Perceived Behavioral Control (PBC) – the perceived ease or difficulty of performing the behavior [17].

In the context of research ethics, TPB provides a valuable framework for understanding how students’ beliefs, social influences, and perceived control affect their intention to adhere to ethical standards. Given the critical importance of ethical conduct in research and the lack of region-specific data in western Iran, the present study aims to identify the behavioral determinants of adherence to REC among students at KUMS, using TPB as the guiding theoretical framework.

Methods

Study overview

This descriptive-analytical study was conducted in 2024 among students at KUMS. A cluster sampling method was employed, with faculties that included MSc or professional doctoral students considered as distinct clusters. Within each cluster, participants were selected using simple random sampling to ensure representativeness. Data collection took place on campus through face-to-face recruitment. Eligible students were invited directly to participate after receiving a full explanation of the study’s objectives and confidentiality protocols. Structured paper-based questionnaires were used to gather data. Participants completed them independently, providing self-reported responses that reflected their views and experiences. Inclusion criteria focused on MSc students and those enrolled in professional doctoral programs in medicine, dentistry, and pharmacy. Students who declined to participate or were absent during data collection were excluded. Additionally, incomplete questionnaires were removed from the final analysis to preserve data integrity. To enhance clarity, a flowchart illustrating the data collection process has been provided in Appendix 1.

Sample size

In estimating the required sample size, it was noted that specific data on adherence to standards REC among KUMS students was not readily available. Consequently, a P value of 0.5 was assumed for the calculations. Based on the university’s official statistics, the total student population was reported to be 5,491. Utilizing a 95% confidence level and a 5% margin of error, the calculated sample size necessary for this study was determined to be 359 participants. Following the data collection phase, incomplete questionnaires were removed from the analysis, resulting in a final sample of 271 completed questionnaires. The overall response rate for this study was 75.4%.

Data collection tools

As previously outlined, the primary data collection instrument employed in this study was a written questionnaire, which participants completed through self-reporting. The questionnaire was structured into two distinct sections. The complete questionnaire form is attached as a supplementary file (Appendix 2).

The first section gathered essential background and demographic information, including variables such as age, gender, faculty of study, educational level, marital status, being employed while studying, familiarity with REC, and prior participation in research ethics workshops. This foundational information is crucial for understanding the context of the participants’ responses.

The second section of the questionnaire was designed to assess constructs derived from the TPB. This included key elements such as attitude, subjective norms, PBC, and intention, alongside the actual behaviors related adherence to standards REC. This section was meticulously crafted based on insights from prior studies [12–16] concerning adherence with ethical principles within academic settings. Furthermore, both the validity and reliability of the questionnaire were rigorously evaluated prior to the commencement of the study. For the constructs within the questionnaire: Attitude was assessed through 6 items (e.g., “Failure to adhere to ethical codes in research leads to the infringement of others’ rights.“). Subjective norms were evaluated with 4 items (e.g., “If I do not fully comply with ethical codes in my research, other students will not endorse it.“). PBC included 2 items (e.g., “I can easily adhere to all ethical codes in my research.“). Intention was measured with 3 items (e.g., “How willing are you to comply with all ethical codes in your research (thesis)?“). Behavior was assessed through 12 items (e.g., “I will not begin data collection before receiving ethical approval.“). A higher score in each construct indicated a more favorable attitude, stronger subjective norms, greater PBC, higher intention, and improved behavior adherence to standards REC. Participants responded to the items using a 5-point Likert scale, ranging from “very low” to “very high.”

Validity and reliability assessment

In this study, the evaluation of face validity, content validity, and internal consistency was paramount. The face validity of the questionnaire was assessed qualitatively through in-depth interviews with 12 experts, leading to adjustments based on their constructive feedback. During these interviews, experts were asked to evaluate specific aspects of each item, including clarity, wording, and cultural appropriateness. Expert feedback was systematically documented through structured note-taking and subsequently subjected to thematic analysis to identify recurring patterns and inform item refinement. Content validity was measured using both qualitative and quantitative approaches. For the qualitative assessment, the same group of 12 experts was consulted to provide their opinions on the difficulty, relevance, and clarity of the items, resulting in further modifications based on their insights. Additionally, to quantitatively assess content validity, a different set of 12 experts was invited to categorize each item as “essential,” “useful but not essential,” or “not essential.” According to Lawshe’s table, the minimum acceptable thresholds for the Content Validity Index (CVI) and Content Validity Ratio (CVR) were established at 0.79 and 0.62, respectively [18, 19]. To ensure greater transparency in the validation process, the CVI and CVR were calculated for each item and subsequently aggregated by construct. Within the “Attitude” domain, CVI values ranged from 0.83 to 1.00, with a mean of 0.93, while CVR values varied between 0.67 and 1.00, yielding a mean of 0.86. In the “Subjective Norms” construct, CVI scores ranged from 0.92 to 1.00 (mean = 0.96), and CVR scores ranged from 0.67 to 1.00 (mean = 0.91). For the “Perceived Behavioral Control” section, CVI values were 0.92 and 1.00 (mean = 0.96), and CVR values were 0.83 and 1.00 (mean = 0.92). In the “Behavioral Intention” domain, CVI scores ranged from 0.92 to 1.00 (mean = 0.97), and CVR scores were uniformly 1.00 across all items. Lastly, in the “Adherence to Standards REC Behaviors” section, CVI values ranged from 0.92 to 1.00 (mean = 0.98), and CVR values ranged from 0.67 to 1.00 (mean = 0.94). These findings indicate that all items met or surpassed the minimum acceptable thresholds, supporting their inclusion in the final version of the instrument. All items were retained in their original form, as the CVR and CVI analyses indicated that each met the established validity thresholds, requiring no revisions or exclusions.

Internal consistency was evaluated using Cronbach’s alpha coefficient for the constructs examined. An alpha value exceeding 0.65 was deemed acceptable [20]. Prior to the main study, a pilot study involving 20 students from KUMS was conducted. This preliminary study aimed to gather feedback on the clarity, comprehensiveness, and completion time of the various tools, as well as to estimate the internal consistency of the instruments. The reliability estimates for the constructs were as follows: attitude (0.83), subjective norms (0.75), PBC (0.82), intention (0.77), and adherence to standards REC behaviors (0.85).

Data analysis

The data collected in this study were analyzed using SPSS Statistics software, version 16, a widely used and reliable tool for quantitative data analysis in social and behavioral sciences. SPSS offers several advantages, including an intuitive interface, robust statistical capabilities, and efficient data management features, making it particularly suitable for descriptive and inferential analysis in academic research. To examine the relationships among the TPB constructs, Pearson correlation tests were utilized. Additionally, linear regression analysis was employed to identify the determinants influencing adherence to REC standards. Both the raw and adjusted model data were reported to enhance transparency and interpretability. Descriptive statistics—including frequency, percentage, mean, and standard deviation—were presented to provide a comprehensive overview of the findings. To quantify behavioral compliance, the mean behavior score was divided by the maximum possible score and multiplied by 100, yielding a percentage value. This method allowed for a standardized interpretation of adherence levels across participants.

Results

The participants had a mean age of 23.99 years (SD = 3.24), with ages ranging from 19 to 36 years. Table 1 presents a demographic overview of the participants, highlighting various characteristics such as sex, marital status, employment status while studying, educational level, field of study, participation in REC workshops, and familiarity with REC. The sample is relatively balanced in terms of sex, with 52% male and 48% female participants. Most respondents are single (89.7%) and not employed while studying (92.6%). In terms of education, a significant majority hold MD degree students (71.2%), while 28.8% have MSc students. The distribution of participants across different schools reveals that the majority are from the Medicine school (39.5%), followed by dentistry (24.0%) and pharmacology (11.1%). Participation in REC workshops is relatively low, with 80.1% not attending. Furthermore, the level of familiarity with REC varies, as 41.0% report very low familiarity, while only 3.0% indicate very high familiarity.

Table 1.

Demographic characteristics of participants

Variables	Frequency	Percent
Sex
Female	130	48.0
Male	141	52.0
Marital Status
Married	28	10.3
Single	243	89.7
Being employed while studying
No	251	92.6
Yes	20	7.4
Level of Education
MSc	78	28.8
MD	193	71.2
School
Medicine	107	39.5
Dentistry	65	24.0
Pharmacology	30	11.1
Nursing and Midwifery	18	6.6
Allied Medical Sciences	17	6.3
Health	25	9.2
Nutrition	9	3.3
History of Participation in REC Workshop*
No	217	80.1
Yes	54	19.9
Level of familiarity with REC standards
Very Low	111	41.0
Low	73	26.9
Medium	62	22.9
High	17	6.3
Very High	8	3.0

^*The REC workshop is a structured educational session designed to enhance students’ understanding of research ethics, including topics such as informed consent, data confidentiality, plagiarism, and responsible conduct of research.

Table 2 provides a comprehensive analysis of the relationships among various psychological constructs, including attitude, subjective norms, PBC, intention, and behavior. The table displays correlation coefficients between these variables, indicating the strength and direction of their relationships. Notably, the strongest correlation is observed between behavior and attitude (r = 0.581), suggesting that a more positive attitude is associated with more favorable behavioral outcomes. Additionally, intention shows substantial correlations with attitude (r = 0.416), subjective norms (r = 0.461), and PBC (r = 0.379), highlighting the importance of these factors in shaping individuals’ intentions. The descriptive statistics for each construct are presented below to illustrate the distribution of participants’ responses. The mean score for attitude was 22.90 with a standard deviation of 5.20, within a possible range of 6 to 30. For subjective norms, the mean was 12.37 (SD = 3.14), with scores ranging from 4 to 20. The construct of PBC had a mean of 6.67 (SD = 1.59), within a range of 2 to 10. The mean score for intention was 10.80 (SD = 2.34), with a possible range of 3 to 15. Finally, adherence to REC behaviors showed a mean of 41.46 (SD = 8.05), with scores ranging from 12 to 60, indicating that students achieved 69.1% of the maximum possible score for their compliance to the REC guidelines.

Table 2.

Summary of correlations and means for Attitude, subjective Norms, PBC, Intention, and adherence to standards REC behaviors

Variables	Attitude	Subjective Norms	PBC Total	Intention	Mean (Standard Deviation)	Score Range
Attitude	1				22.90 (5.20)	6–30
Subjective Norms	0.059	1			12.37 (3.14)	4–20
PBC	0.051	0.501**	1		6.67 (1.59)	2–10
Intention	0.416**	0.461**	0.379**	1	10.80 (2.34)	3–15
Adherence to Standards REC Behaviors	0.581**	0.108	0.138*	0.392**	41.46 (8.05)	12–60

**Correlation is significant at the 0.01 level (2-tailed)

*Correlation is significant at the 0.05 level (2-tailed)

Table 3 presents the results of factors influencing adherence to standards REC behaviors among students. The table includes both crude and adjusted regression coefficients (B) along with their 95% confidence intervals (CI) and significance levels (Sig.). Among the variables analyzed, attitude and intention were found to be significant predictors of compliance behavior. Specifically, the crude model shows a strong positive relationship for attitude with a coefficient of B = 0.878 (95% CI: 0.711, 1.045, Sig. < 0.001), indicating that attitude is positively associated with compliance behavior. Similarly, intention also demonstrated a robust association with compliance behavior, presenting a coefficient of B = 1.428 (95% CI: 0.987, 1.869, Sig. < 0.001) in the crude model. In the adjusted model, intention remained significant with B = 0.857 (95% CI: 0.308, 1.407, Sig. = 0.002), highlighting its critical role in predicting adherence to standards REC behaviors. In the adjusted model, attitude remained significant with B = 0.694 (95% CI: 0.493, 0.894, Sig. < 0.001), highlighting its continued association with adherence to standards REC behaviors.

Table 3.

Determinants of adherence to standards REC behaviors: crude and adjusted models

Variables	Crude		Adjusted
	B (95% CI)	Sig.	B (95% CI)	Sig.
Age	0.191 (−0.145, 0.528)	0.264	-	-
Sex	0.023 (−2.077, 2.122)	0.983	-	-
Marital status	−0.910 (−4.163, 2.344)	0.582	-	-
Being employed while studying	3.741 (−0.349, 7.831)	0.073	−0.502 (−4.688, 3.684)	0.813
Level education	−2.999 (−5.272, −0.726)	0.010	0.272 (−3.843, 4.386)	0.897
School	0.548 (−0.028, 1.123)	0.062	0.611 (−0.387, 1.609)	0.229
History of participation in REC workshop	2.065 (−0.480, 4.610)	0.111	0.929 (−1.820, 3.678)	0.506
Level of familiarity with REC	1.540 (0.577, 2.503)	0.002	0.437 (−0.650, 1.524)	0.429
Attitude	0.878 (0.711, 1.045)	< 0.001	0.694 (0.493, 0.894)	< 0.001
Subjective norms	0.288 (−0.062, 0.638)	0.107	−0.109 (−0.487, 0.269)	0.570
PBC	0.712 (0.035, 1.389)	0.039	−0.103 (−0.797, 0.592)	0.711
Intention	1.428 (0.987, 1.869)	< 0.001	0.857 (0.308, 1.407)	0.002

Other factors such as level of familiarity with REC standards and PBC did not show significant relationships in the adjusted model despite being promising in the crude model. For instance, the crude model indicated that familiarity with REC had a positive impact (B = 1.540, 95% CI: 0.577, 2.503, Sig. = 0.002), but this relationship was not maintained after adjustment (B = 0.429, 95% CI: −0.650, 1.524, Sig. = 0.437). Similarly, PBC was significant in the crude model (B = 0.712, 95% CI: 0.035, 1.389, Sig. = 0.039) but lost significance in the adjusted model (B = −0.103, 95% CI: −0.797, 0.592, Sig. = 0.711). This suggests that while these factors may initially appear relevant, their influence diminishes when accounting for other variables in the model. The model summary indicates a strong relationship between the predictors and the dependent variable, with an R value of 0.617, suggesting a moderate to high correlation. The R Square value of 0.381 indicates that approximately 38.1% of the variance in the dependent variable can be explained by the model, which includes predictors such as job, level of education, school, history of participation in REC workshop, level of familiarity with REC, attitude, PBC, subjective norms, and intention. The Adjusted R Square value of 0.352, which accounts for the number of predictors in the model, suggests that the model remains robust even when considering the complexity added by multiple variables. The standard error of the estimate is 6.55, reflecting the average distance that the observed values fall from the regression line, which provides a measure of the model’s accuracy. Overall, the model demonstrates a significant ability to explain the behavior based on the included predictors.

Figure 1 provides a concise comparison of crude and adjusted regression coefficients for predictors of REC-related behavior. The crude model shows that being employed while studying (B = 3.741), level of education (B = − 2.999), familiarity with REC (B = 1.540), attitude (B = 0.878), PBC (B = 0.712), and intention (B = 1.428) have notable associations with REC-related behavior. After adjustment, only attitude (B = 0.694) and intention (B = 0.857) remain statistically significant, indicating their robust and independent predictive value. Demographic variables such as age, sex, and marital status show minimal influence in both models.

Fig. 1.

Side-by-Side Presentation of Crude and Adjusted Effects for Adherence to Standards REC Behaviors Predictors

The data presented in Table 4 highlights the extent of students’ adherence to standards REC, measured on a scale from 1 to 5. The average scores indicate varying levels of adherence to ethical principles, with the highest compliance observed for the principle of confidentiality and safeguarding participants’ secrets (mean score of 4.04). This suggests that students prioritize the protection of participant information. Conversely, the lowest compliance is seen in the requirement that no data collection should begin before obtaining ethical approval, with a mean score of 2.68. This indicates a significant gap in understanding or valuing the importance of securing ethical approval prior to initiating research activities.

Table 4.

Mean scores and standard deviations of students’ adherence to REC standards

Variable	Mean	Standard Deviation
No data collection should begin before obtaining ethical approval.	2.68	1.22
Informed and voluntary consent must be obtained from study participants, free from coercion, threats, bribery, or seduction.	3.53	1.14
The benefits to society or scientific advancement cannot justify exposing participants to unreasonable harm or limiting their autonomy.	3.16	1.15
Participants should be informed about the research methods, objectives, potential risks, benefits, nature, and duration of the study.	3.49	1.07
Compensation must be provided in case of undue harm during the research.	3.61	1.00
Necessary measures (such as preventive, diagnostic, and therapeutic options) should be provided.	3.64	1.03
The material and moral rights of all parties (participants, researchers, research, and relevant organizations) must be respected in reporting.	3.63	1.10
Participants should be informed that they can withdraw from the study at any time.	3.37	1.04
In clinical trials, participants should be made aware that they may be randomly assigned to one of two groups: test or intervention.	3.31	0.95
Research methods must align with the religious and cultural norms of participants and society.	3.16	1.16
Factors like speed, ease of work, researcher convenience, lower costs, and practicality should never expose participants to additional risks or harms.	3.62	1.09
The principle of confidentiality and the safeguarding of participants’ secrets must be upheld.	4.04	1.05

Discussion

Participants achieved an average score of 69.1% on the questionnaire assessing adherence to ethical codes in research. Consistent with our findings, Abbaszadeh et al. reported a 63.89% adherence level among postgraduate students at Tabriz University, with the highest scores related to respecting participants’ ethical rights [21]. In our study, the lowest adherence was observed in the item concerning data collection prior to ethical approval, while the highest scores were associated with confidentiality and safeguarding participants’ interests. This moderate level of adherence is not unique to our context. For instance, Li et al., [22] conducted a cross-national study—including Western countries—and found considerable variation in ethical compliance among researchers. Reported issues included data manipulation, failure to obtain informed consent, and breaches of confidentiality, suggesting that ethical challenges in research are global and shaped by institutional and cultural factors.

These findings imply that students may prioritize the expediency of completing their research or thesis over securing proper ethical approvals, underscoring the need for further investigation and targeted educational interventions. El-Dessouky et al. found that 44% of researchers in dental schools across two Middle Eastern universities believed that strict adherence to ethical codes delays research progress [4]. In contrast, Mallela et al. reported that only 20% of participants in Indian academic settings shared this view [3]. Taken together, studies from Iran, the Middle East, India, and Western countries reveal that adherence to ethical codes is influenced by a complex interplay of cultural attitudes, institutional support, and perceived procedural barriers.

Overall, the most consistently observed behaviors included maintaining confidentiality and compensating for potential harm, while the least observed was initiating data collection without prior ethical approval. This disparity highlights the urgent need to strengthen ethics education and awareness among students. To address these challenges, academic institutions should implement comprehensive training programs that emphasize the importance of ethical approval prior to data collection. Workshops and seminars can foster a deeper understanding of ethical standards and their implications for research integrity. Moreover, integrating ethics into the curriculum and clarifying the role of ethics committees may help demystify the approval process and reduce resistance. Finally, streamlining the ethical review process could encourage students to seek approval more readily, thereby improving compliance and promoting responsible research practices.

Our findings indicate that intention and attitude are the most significant factors influencing adherence to ethical codes in research among students. However, it is important to clarify that within the TPB, attitude typically influences behavior indirectly through intention. In our study, we examined the direct effect of attitude on behavior in both crude and adjusted regression models. While attitude showed a significant direct association with behavior, this may reflect contextual influences or overlapping constructs in students’ responses. This deviation from the classical TPB structure has been acknowledged, and further research is recommended to explore these dynamics. Numerous studies align with our findings, demonstrating that attitude is a determining factor in students’ behaviors regarding academic issues, such as adhering to ethical principles in education and research [12–15]. For instance, Stone et al. utilized the TPB in their study among students in the United States, revealing that attitude is a crucial predictor of academic misconduct [12]. Similarly, Harding et al. found that attitudes toward cheating serve as a strong predictor of college cheating behavior [13]. Furthermore, the study by Chudzicka-Czupała et al., across seven countries based on TPB (Poland, Ukraine, Romania, Turkey, Switzerland, the United States, and New Zealand), demonstrated that students’ attitudes significantly predict their intention to engage in academic dishonesty, such as cheating [14]. Additional studies from Western countries further support these findings. Leonard et al., found that attitude significantly predicts ethical behavioral intention in academic settings in the United States [23]. Likewise, Stone et al., confirmed that attitude, subjective norms, and PBC are all associated with academic misconduct intentions and behavior [24]. These findings suggest that cultural norms, institutional expectations, and individual autonomy in Western contexts may shape students’ ethical decision-making differently than in collectivist cultures. However, contradictory results have also been reported. For example, Mirzaei-Alavijeh, et al., [16] research among KUMS students indicated that attitude is not a significant predictor of intentions to commit plagiarism.

Our findings indicate that subjective norms and PBC were not significant predictors of adherence to REC standards among students. While PBC was statistically significant in the crude model, it lost its significance in the adjusted model. This observation contrasts with various studies that have reported differing results. Some research emphasizes the positive impact of subjective norms and PBC on students’ academic behaviors [14, 16]. The discrepancies in findings suggest that contextual factors and specific academic environments may influence the role of subjective norms and PBC in predicting ethical behavior. For instance, variations in institutional culture, peer influence, and the level of awareness regarding ethical standards could account for these differences. Furthermore, this discrepancy highlights the complexity of the relationship between attitudes and academic integrity, suggesting that contextual factors and individual differences may play critical roles.

To enhance adherence to ethical codes in research, educational institutions should implement comprehensive training programs that focus on fostering positive attitudes toward academic integrity. These programs should incorporate real-life scenarios and case studies to illustrate the importance of ethical behavior in research and the potential consequences of misconduct. Based on our findings, several real-life scenarios can be integrated into ethics training programs to make the content more tangible and relevant for students. These scenarios are not hypothetical—they reflect actual gaps in ethical adherence observed among students in our study and highlight common challenges in research conduct. For instance, initiating data collection without prior ethical approval emerged as the lowest scoring item, indicating a lack of awareness or undervaluing of institutional review processes. Another scenario involves obtaining consent under pressure or without full disclosure, which compromises the principle of informed and voluntary participation. Misreporting or manipulating data for convenience or academic gain represents a serious breach of research integrity, often driven by performance pressures. Additionally, failing to maintain confidentiality of participant information underscores the need for greater emphasis on data protection and privacy. Including these examples in training programs can stimulate critical discussion, promote ethical reflection, and help students internalize the importance of responsible research practices. Additionally, further research is needed to explore the underlying reasons for contradictory findings in the literature. This could involve qualitative studies that delve into students’ perceptions and experiences regarding ethical conduct. Finally, institutions should establish clear policies that not only outline the consequences of unethical behavior but also promote a culture of integrity. This can be achieved through mentorship programs, peer-led discussions, and the integration of ethics into the curriculum, ensuring that students understand the significance of ethical standards in their academic and professional lives.

Strengths and limitations of the study

This study demonstrates several notable strengths, particularly in its methodological rigor and the application of the TPB to examine adherence to REC among students at KUMS. The use of cluster sampling and structured questionnaires enabled comprehensive data collection on key constructs such as attitude, subjective norms, PBC, and behavioral intention.

To minimize bias, a validated structured questionnaire was employed, random sampling within clusters was conducted, and a multi-step data entry and verification process was implemented. The data were initially entered into SPSS by the second author and subsequently reviewed and refined by the first and corresponding authors. Final statistical analyses were performed by the third author, and all outputs were rechecked by the first and corresponding authors to ensure accuracy and consistency.

Nevertheless, the study has several limitations. Reliance on self-reported data may introduce bias, as participants could potentially overestimate their adherence to ethical standards. Moreover, the cross-sectional design limits the ability to draw causal inferences between TPB constructs and ethical behavior. The lack of significant findings related to subjective norms and perceived behavioral control suggests that contextual factors may influence ethical conduct, highlighting the need for further qualitative research to explore these dimensions.

Additionally, data collection was limited to a single institution (KUMS), which may restrict the generalizability of the findings. Future studies involving multiple universities with diverse cultural and educational backgrounds are recommended to enhance external validity and provide a more comprehensive understanding of ethical adherence in research contexts.

Conclusions

The present study revealed that students demonstrated a moderate level of adherence to ethical research codes (69.1%), with the highest compliance observed in maintaining confidentiality and the lowest in securing ethical approval prior to data collection. These findings underscore critical gaps in ethical awareness and practice, particularly regarding procedural compliance. Importantly, the results align with key constructs of the TPB. While TPB posits that attitude influences behavior primarily through intention, our regression analysis showed that both attitude and intention were significant direct predictors of ethical behavior—even after adjustment for other variables. This suggests that, within this context, attitude may exert a more immediate influence on behavior than traditionally assumed in TPB, possibly due to the internalization of ethical values among participants. Conversely, subjective norms and PBC did not retain significance in the adjusted model, indicating a limited role in shaping actual compliance behavior. The inclusion of both crude and adjusted regression models further clarified the robustness of these relationships. While several variables appeared influential in the crude model (e.g., familiarity with REC, PBC), only attitude and intention remained significant after controlling for confounding factors, reinforcing their independent predictive value. These insights highlight the need for targeted educational interventions that not only inform students about ethical procedures but also cultivate positive attitudes and strengthen intentional commitment to ethical conduct. Institutions should prioritize ethics-focused training and streamline review processes to enhance compliance. Future research may benefit from qualitative approaches to explore the cognitive and contextual factors that mediate the translation of ethical intention into behavior.

Authors’ contributions

F.J. and M.M.A designed the initial proposal and method to accomplish the project. F.J., M.M.A. and H.G. performed specialized analysis of statistical data and also editing the final version. S.V. contributed to the data collection. All authors have read and approved the manuscript.

Funding

This research has been financially supported by the KUMS (research code number: 4020937).

Data availability

“The data are accessible from the corresponding author upon reasonable request.”

Declarations

Ethics approval and consent to participate

This study received ethical approval from the Research Ethics Committee of Kermanshah University of Medical Sciences and was conducted in accordance with national guidelines and standards for medical research in Iran (Approval Code: IR.KUMS.REC.1402.592). All procedures adhered to the ethical principles outlined by the institutional and national research committees, as well as the Declaration of Helsinki (1964) and its subsequent revisions. Before enrolling in the study, participants were provided with detailed information regarding the research objectives, methodology, potential risks and benefits, and their rights as participants. They were clearly informed that participation was entirely voluntary and that they could withdraw from the study at any time without consequence. Each participant received an information sheet and provided written informed consent. To maintain confidentiality, all data were anonymized and securely stored. No personally identifiable information was linked to the responses.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.