From novice to professional: comparing professional identity formation in undergraduate medical students: a cross-sectional study

Abstract

Background

Professional identity formation (PIF) is central to medical education, shaping students into competent and ethical practitioners. Despite its recognized importance, data on PIF’s cross-sectional variation across year groups in Pakistani undergraduate medical education remain limited. This study assessed PIF across all five MBBS years and examined associations with academic year group and gender.

Methods

A comparative cross-sectional study was conducted (November 2021–April 2022) at Liaquat College of Medicine and Dentistry, Karachi. All 450 MBBS students (years one to five) were invited via universal sampling. A 27-item adapted version of the Professional Identity Five Factor Scale (PIFFS), covering five dimensions of professional identity, was administered online. Kruskal-Wallis tests with Dunn’s post-hoc comparisons (Benjamini-Hochberg FDR correction) were used for year-group comparisons; Mann-Whitney U tests were applied for gender comparisons. Effect sizes were calculated using eta-squared (η²) and rank-biserial correlation (r). Statistical significance was set at p ≤ .05.

Results

Of 450 students, 361 participated (80% response rate; Cronbach’s α = 0.76). Overall PIF scores ranged from 3.5 to 3.8 across year groups, with lower values in the fourth-year cohort and higher values in the final year; these are between-group differences and do not represent within-student change. Kruskal-Wallis tests revealed significant between-group differences for all five dimensions (p < .05); Dunn’s post-hoc comparisons identified Year 4 as the primary driver of significant pairwise differences in four of the five dimensions. Omnibus effect sizes were small (η² range: 0.021–0.047); pairwise effect sizes were small to medium. Significant gender differences were found for Dimensions C, D, and E, with male students scoring higher on all three (p < .05). Among first-year students who had already identified a preferred specialty, 68% were female (p = .034).

Conclusions

PIF scores varied across undergraduate medical year groups, with consistently lower scores in the fourth-year cohort compared to adjacent year groups across multiple dimensions. These findings suggest that PIF remains moderate overall and that the fourth-year period may benefit from enhanced mentorship and reflective learning support. Longitudinal and multi-centre studies are needed to establish developmental trajectories.

Background

Identity formation

Identity formation is a foundational developmental process that shapes how individuals perceive themselvesand their place in society. Identity reflects a sense of belonging, shaped through continuous self-evaluation and lived experience [1, 2]. Empirical work confirms that identity formation is not merely a theoretical construct but carries measurable consequences for adolescent health outcomes. Adolescents who achieve a consolidated identity demonstrate better psychological well-being, higher subjective quality of life, greater sense of purpose, and lower levels of anxiety, depression, and identity confusion [3, 4]. Conversely, identity diffusion or foreclosure — characterized by low exploration or commitment made in the absence of exploration — is associated with poorer outcomes, including diminished academic performance, less adaptive coping strategies, and increased risk of maladjustment [5, 6].

Identity formation begins early in life and evolves through various phases. Professional identity (PI) is formed when individuals join a community and begin to identify and act as a part of it [2]. Wald [7] describes PI in medicine as a gradual internalization of the values and norms of the profession. It transforms individuals from laypersons to professionals, shaped by psychosocial factors, interpersonal characteristics, and social interactions [8]. Hence, professional identity formation in medicine is not just about acquiring skills and knowledge; it is about becominga doctor in the fullest sense — someone who embodies the values, ethics, and social purpose of medicine [9, 10]. It underpins professionalism, ethical practice, resilience, and compassionate care, serving as the foundation of trust between the medical profession and society [5, 7, 8].

Medical professionalsare expected to uphold the values and ethics of their profession and to demonstrate attributes including clinical competence, ethical integrity, empathy, effective communication, lifelong learning commitment, and professionalism in all interactions. Forming PI shapes the students into ethical, empathetic doctors who embody these attributes. In medical education PI is formed through reflective learning, understanding societal roles, and practical engagement [11, 12]. The Carnegie Foundation [13–16] and researchers like Cruess [5] and Olive [12] endorse deliberate inclusion of PIF in formal curriculum, especially in early education when learners rely on imitation.

Through early exposure to legitimate peripheral participation [17] and gradual progression to full participation through experiential learning, they adopt values such as compassion and integrity [18]. Role modeling helps integrate personal traits with professional norms, leading to successful inclusion in the medical community [9]. Clinical rotations are pivotal in this transition as they provide the practical context for applying these developing professional values.

Yakov et al. identified three levels of identity development [19]: linking experiences, comparing them, and looking into the future. Local [20] and International [21] studies agree that Professional Identity Formation (PIF) increases over time, especially in the clinical years, though students also face challenges due to gaps between idealism and real-world healthcare practices.

Factors affecting PIF

Studies confirm that PI starts forming before medical school, influenced by personal aspirations [22]. Environmental exposures such as media, books, and clinical role models significantly influenced PI development [19]. Non-academic activities like white coat ceremonies [1, 12, 23], teamwork as well as academic experiences [1] such as respecting cadavers, bedside teaching, and Problem-based learning (PBL) groups contribute towards the development of professional values. Clinical skills training boosts early PIF scores by strengthening students’ confidence and competence [20]. Mancini et al. [24] found that strong self-image and perceived competence enhance PI formation.

Mentorship, shadowing and interactions with standardized patients also strengthened PI [25]. During the COVID-19 pandemic, many students felt proud of their profession, as they witnessed healthcare workers’ dedication and resilience, although the intensity and nature of this experience varied across institutions and contexts [26].

PIF and gender difference

Despite the increasing number of female doctors [27–30], many faced barriers in specialty choice and career growth [31–33]. This contributes to burnout [34] and impedes identity formation [31]. Environments with less stress and stronger social support have been associated with enhanced PIF in female students, as evidenced by studies from China [26].

Measuring PIF in UGME

Various tools have been developed to measure PIF, including the Professional Identity Essay (PIE), the Questionnaire on Professional Identity Development (QPID), and narrative reflection analyses. Among these Profession-specific tools like the Professional Identity Five Factor scale (PIFFS) offer a comprehensive way to assess both the development and dimensions of PIF among students [6].

Professionalism was formally introduced into Pakistan’s MBBS curriculum in 2019. However, issues such as a lack of dedicated teaching hours and faculty training in professionalism education are major hurdles [35].

Despite the recognized importance of PIF and its formal inclusion in Pakistan’s MBBS curriculum since 2019, quantitative data on how PIF varies across all five undergraduate years in the Pakistani context remain limited. Local studies are few, and no prior study has used the PIFFS to characterize PIF cross-sectionally across all year groups in a Pakistani medical institution, nor systematically examined gender differences across each PIF dimension using a validated psychometric instrument. This study therefore aimed to: (1) assess PIF across all five years of the MBBS programme using the adapted PIFFS at Liaquat College of Medicine and Dentistry, Karachi; and (2) examine associations between PIF dimensions and both academic year group and gender.

Methodology

A comparative cross-sectional study was conducted from November 2021 to April 2022 at Liaquat College of Medicine and Dentistry, Karachi. A universal sampling approach was employed, and all the 450 MBBS students were included.

Inclusion criteria

• All medical students from the first year to the final year MBBS classes.

Exclusion criteria

• Students who did not give consent.

• Students included in the pilot study.

Data collection & analysis

The Professional Identity Five Factor Scale (PIFFS) [6] was adapted for use in this study with written permission from the original author, Tan Chin Pei, obtained by email in April 2021. The instrument was modified to suit the undergraduate medical education context in Pakistan by replacing generic professional terminology with medical-specific language, including references to the healthcare sector, the MBBS degree, and clinical practice. The adapted version retained the five-factor structure of the original instrument. These modifications were reviewed in a pilot study of 50 MBBS students, and minor refinements were made based on the pilot findings; students included in the pilot study were excluded from the main study. The modified PIFFS questionnaire was then administered online using Google Forms. In the original validation study, the questionnaire’s five dimensions demonstrated reliability ranging from 0.65 to 0.85. The 27-item adapted questionnaire comprised 25 Likert-scaled items and two dichotomous items across five dimensions, namely “knowledge about professional practices”, “experience with the profession”, “having the professional as a role model”, “professional self-efficacy”, and “preference for a particular profession”. All the items were rated on a five-point Likert scale (Strongly Disagree = 1, Disagree = 2, Neutral = 3, Agree = 4, Strongly Agree = 5) except the dichotomous (“Yes”/“No”) items: an item on attendance at internship programmes, reported at the item level using the chi-square test across year groups (Table 7), and the specialty-preference item, analyzed using the chi-square test to assess the association between gender and specialty preference within each year group (Table 8). Dichotomous items were not included in the dimension-level composite scores.

Table 7.

Association of Preference for a Particular Specialty scores with Academic Year among Undergraduate Medical Students (n = 361)

Dimension E	Items	Academic Year					P value
		Median (IQR)
		1 st Year	2nd Year	3rd Year	4th year	Final Year
Preference for a Particular Specialty	My future work is related to the degree program I am enrolled in	5 (4–5)	5 (4–5)	5 (4–5)	4 (4–5)	4 (4–5)	0.020*
	My medical education is relevant and useful in preparing me for my future career	4 (4–5)	5 (4–5)	4 (4–5)	4 (4–5)	4 (4–5)	0.255
	I have attended the internship programs†	10.4%	10.1%	10.1%	25.3%	54.2%	< 0.001*
	I have interacted with professionals in the industry outside of my college or through events organized in my college	3 (2–4)	3 (2–4)	4 (2–4)	4 (2–4)	4 (3–4)	0.380

Values are Median (IQR) unless otherwise indicated. Likert items scaled 1–5 (Strongly Disagree to Strongly Agree). P-values for Likert items from Kruskal-Wallis H tests; binary item from Pearson chi-square. * p ≤.05

† Binary item (Yes/No). Values shown are percentage answering “Yes”. Chi-square: χ² = 51.41, df = 4. This item was not included in the Dimension E composite score

Table 8.

Dimension E – Preference for a Particular Specialty

Academic Year	Do you already know what kind of work or specialty you prefer?	Gender		Total	P -Value
		Female	Male
1 st Year	Total (Yes)	57 (40, 70.2%)	39 (19, 48.7%)	96 (59, 61.5%)	0.034
2nd Year	Total (Yes)	44 (33, 75.0%)	25 (19, 76.0%)	69 (52, 75.4%)	0.926
3rd Year	Total (Yes)	43 (29, 67.4%)	26 (21, 80.8%)	69 (50, 72.5%)	0.230
4th year	Total (Yes)	49 (32, 65.3%)	30 (22, 73.3%)	79 (54, 68.4%)	0.457
Final Year	Total (Yes)	18 (14, 77.8%)	30 (28, 93.3%)	48 (42, 87.5%)	0.115
Total	Total (Yes)	211 (148, 70.1%)	150 (109, 72.7%)	361 (257, 71.2%)	0.602

Values are shown as total n (n who answered “Yes”, % of that group answering “Yes”); for example, “57 (40, 70.2%)” denotes 57 female students, of whom 40 (70.2%) answered “Yes”. P-values are from chi-square tests of the association between gender and the specialty-preference item within each year group. * p ≤.05

An electronic consent was circulated, emphasizing voluntary participation. Multiple reminders boosted response rates. The Google Form was configured to require a response for every item before submission; accordingly, there were no missing data, and 100% of dimension-level observations were complete. No imputation or listwise deletion was necessary. The data were analyzed using SPSS Version 23.0.

Normality testing using Kolmogorov-Smirnov and Shapiro-Wilk tests confirmed that the data were not normally distributed; quantitative variables were therefore summarised using medians and interquartile ranges (IQR). Frequencies and percentages were reported for each year group and by gender. Dimension-level scores were calculated as the median of each respondent’s mean score across the items within that dimension; the overall PIF score represents the mean of these five dimension-level medians. Between-group differences across academic year groups were assessed using the Kruskal-Wallis H test. Where Kruskal-Wallis tests were significant, pairwise post-hoc comparisons were conducted using Dunn’s test with Benjamini-Hochberg false discovery rate (FDR) correction for multiple comparisons. Effect sizes for Kruskal-Wallis tests were calculated using eta-squared (η²; small ≥ 0.01, medium ≥ 0.06, large ≥ 0.14). Effect sizes for pairwise comparisons were calculated using rank-biserial correlation (r; small ≥ 0.1, medium ≥ 0.3, large ≥ 0.5). For gender comparisons (two independent groups), the Mann-Whitney U test was applied with rank-biserial r as the effect size measure. The binary specialty-preference item (Dimension E) was analysed using the Pearson chi-square test to assess associations between gender and specialty preference within each year group. Five Kruskal-Wallis tests were conducted simultaneously; although Benjamini-Hochberg correction was applied to post-hoc pairwise comparisons, a residual risk of Type I error across the omnibus tests should be noted. The significance level was set at p ≤.05 for all tests.

Ethical considerations

This study was conducted in accordance with the Declaration of Helsinki. Ethical approval was obtained from the Institutional Review Board (IRB) of Jinnah Sindh Medical University, Karachi, Pakistan (Reference No: JSMU/IRB/2021/−478), where the principal investigator (B.S.) was enrolled as a postgraduate student at the time of the study. Institutional ethical approval was additionally obtained from Liaquat College of Medicine and Dentistry (Reference No: IRB/M-000017/21), where the study was conducted, prior to data collection. Participation in the study was entirely voluntary. Informed consent was obtained electronically from all participants; the consent form was embedded at the beginning of the online questionnaire and clearly described the purpose of the study, the voluntary nature of participation, and the right to withdraw at any time without consequence. Students who did not provide consent were excluded from the study.

Clinical trial number

Clinical Trial Number: Not Applicable.

Results

Of the target Population of 450 students in the MBBS, 361 students consented to participate in the study. The overall response rate was 80%. Of the 361 participants, 211 (58.4%) were female and 150 (41.6%) were male; the gender distribution within each year group is presented in Table 8.

The Cronbach’s alpha of the overall scale was 0.76, with per-dimension values: Dimension A α = 0.68, Dimension B α = 0.72, Dimension C α = 0.75, Dimension D α = 0.80, Dimension E α = 0.70.

As shown in Table 1, the overall Professional Identity Score was higher from the first year to the third year of MBBS, lower in the fourth year, and higher again in the final year. Kruskal-Wallis tests revealed statistically significant between-group differences for all five dimensions: Dimension A (H = 20.86, p <.001, η² = 0.047), Dimension B (H = 12.24, p =.016, η² = 0.023), Dimension C (H = 11.33, p =.023, η² = 0.021), Dimension D (H = 11.35, p =.023, η² = 0.021), and Dimension E (H = 12.27, p =.016, η² = 0.023). All omnibus effect sizes (η²) were small, although several pairwise comparisons reached medium magnitude (Table 2). Post-hoc Dunn’s tests with Benjamini-Hochberg correction identified Year 4 as the cohort driving most significant pairwise differences across Dimensions A, B, C, and D. Full pairwise comparisons and rank-biserial effect sizes (r) are presented in Table 2.

Table 1.

Total Score on PIF

Dimension – Median (IQR)	1 st Year (n = 96)	2nd Year (n = 69)	3rd Year (n = 69)	4th Year (n = 79)	Final Year (n = 48)
Knowledge About Professional Practices	4.25 (3.50–4.33)	4.50 (3.50–4.60)	4.25 (3.33–4.30)	4.00 (3.16–4.00)	4.00 (3.16–4.00)
Experience with the Profession	4.00 (2.50–4.20)	4.00 (2.16–4.00)	3.80 (2.33–3.83)	3.80 (2.00–3.80)	3.90 (2.66–4.16)
Having the Professional as a Role Model	3.50 (3.00–4.80)	3.75 (3.50–4.80)	3.75 (3.43–4.80)	3.25 (3.00–4.00)	3.88 (3.20–4.00)
Professional Self-Efficacy	3.20 (3.16–4.50)	3.40 (3.33–4.50)	3.40 (3.33–4.33)	3.00 (2.83–4.00)	3.70 (3.00–4.16)
Preference for a Particular Specialty	3.40 (2.66–4.00)	3.40 (2.83–4.00)	3.60 (3.00–4.16)	3.40 (2.66–4.00)	3.50 (3.00–4.16)
Overall Score	3.6	3.7	3.7	3.5	3.8

Values are dimension-level medians with interquartile ranges (IQR; 25th–75th percentiles), computed from each respondent’s mean score across the items of that dimension. The Overall Score is a single derived value per year group (the mean of the five dimension medians) and therefore has no associated IQR. Dimension E (Preference for a Particular Specialty) comprises three Likert-scaled items; its two dichotomous items (internship attendance and specialty preference) are reported at item level in Tables 7 and 8

Table 2.

Kruskal-Wallis and Post-Hoc Significant Pairwise Comparisons (Dunn’s Test, BH-Corrected)

Dimension	H (df = 4)	p-value	η²	Significant Pairs (BH-corrected adj. p, r)
A – Knowledge About Professional Practices	20.86	< 0.001	0.047	1 st vs. 4th Year** (p=.006, r = −.29); 1 st vs. Final Year* (p=.018, r = −.27); 2nd vs. 4th Year** (p=.003, r = −.35); 2nd vs. Final Year** (p=.007, r = −.32); 3rd vs. 4th Year* (p=.050, r = −.20)
B – Experience with the Profession	12.24	0.016	0.023	2nd vs. 4th Year* (p=.011, r = −.31)
C – Having the Professional as a Role Model	11.33	0.023	0.021	2nd vs. 4th Year* (p=.035, r = −.27); 3rd vs. 4th Year* (p=.035, r = −.28)
D – Professional Self-Efficacy	11.35	0.023	0.021	4th vs. Final Year* (p=.019, r = + 0.29)
E – Preference for a Particular Specialty	12.27	0.016	0.023	1 st vs. 3rd Year* (p=.029, r = + 0.28); 2nd vs. 3rd Year* (p=.048, r = + 0.26)

BH Benjamini-Hochberg FDR correction. η²: small ≥ 0.01, medium ≥ 0.06, large ≥ 0.14. r = rank-biserial correlation (small ≥ 0.1, medium ≥ 0.3, large ≥ 0.5). Negative r: first group scored higher than the second; positive r: first group scored lower. All non-listed pairs: ns. * p ≤.05; ** p <.01; *** p <.001

Dimension A – Knowledge about professional practices

As shown in Table 3, Dimension A showed the strongest between-group difference of all five dimensions (H = 20.86, p <.001, η² = 0.047). While item-level medians were broadly consistent across years, a notable decline was observed in the fourth-year cohort, consistent with the post-hoc results in Table 2 identifying Year 4 as the primary driver of significant pairwise differences in this dimension.

Table 3.

Association of Knowledge About Professional Practices Scores with Academic Year among Undergraduate Medical Students (n = 361)

Dimension A	Items	Academic Year					P value
		Median (IQR)
		1 st Year	2nd Year	3rd Year	4th year	Final Year
Knowledge About Professional Practices	I know the nature of the work I will do in my future	4 (4–5)	4 (4–5)	4 (4–4)	4(4–4)	4(4–4)	0.011*
	In most work environments, professionals with different backgrounds work together. I know of the different types of professionals I will be collaborating with	4 (3–4)	4 (3–4)	4(3–4)	4 (3–4)	4 (3–4)	0.680
	I have a good idea of the roles and responsibilities of my future job	4 (4–5)	4 (4–5)	4 (4–4)	4 (3–4)	4 (3–4)	0.003*
	I know what kind of, tools and equipment I will handle in my future Occupation	4 (3–4)	4 (3–4)	4 (3–4)	4 (3–4)	4 (3–4)	0.087
	I am aware of the impact of the decisions I make as a professional in the healthcare industry	4 (4–4)	4 (4–4)	4 (3–4)	4 (3–4)	4 (3–4)	0.242
	I have a good idea about the rules and regulations in the industry	4 (3–4)	4 (3–4)	4 (3–4)	4 (3–4)	4 (3–4)	0.618

Values are Median (IQR). Likert items were scored on a 1–5 scale (Strongly Disagree = 1 to Strongly Agree = 5). P-values are from Kruskal-Wallis H tests comparing academic year groups. Number of students per cohort: 1 st Year, n = 96; 2nd Year, n = 69; 3rd Year, n = 69; 4th Year, n = 79; Final Year, n = 48 (total N = 361). * p ≤.05

Dimension B – Experience with the profession

In Experience with the Profession (Dimension B), dimension-level scores were high and broadly stable across all year groups (medians 3.80–4.00; Table 1). The Kruskal-Wallis test indicated a small but statistically significant difference across academic years (H = 12.24, p =.016, η² = 0.02). The only significant pairwise contrast was between second-year students, who recorded the highest scores, and fourth-year students, whose scores were lower (Dunn’s BH-adjusted p =.011; Table 2). At the item level (Table 4), responses were heterogeneous: students across all years strongly endorsed personally knowing someone in their future profession, whereas experiential items — such as part-time work related to their studies, and prior work experience before MBBS — received consistently lower endorsement.

Table 4.

Association of Experience with the Profession with Academic Year among Undergraduate Medical Students (n = 361)

Dimension B	Items	Academic Year					P value
		Median (IQR)
		1 st Year	2nd Year	3rd Year	4th Year	Final Year
Experience with the Profession	I am working part-time or am running a business related to what I am studying	2 (2–3)	2 (2–3)	2 (2–3)	2 (2–3)	3 (2–4)	0.019*
	I am part of an interest group (inside or outside of my institution related to my profession	2 (2–4)	3 (2–4)	3 (2–4)	3 (2–4)	4 (3–4)	0.102
	I know personally some people who work in my future profession	4(4–5)	4(3–5)	4(3–5)	3(2–4)	4(3–5)	< 0.001*
	I follow developments in my future industry in newspapers and on television.	3(3–4)	3(2–4)	4(3–4)	3(2–4)	4(3–4)	0.046*
	Before I entered MBBS, I already had some prior work experience related to the profession of my choice	2(2–3)	3(2–4)	2(2–3)	2(2–3)	3(2–4)	0.471

Values are Median (IQR). Likert items were scored on a 1–5 scale (Strongly Disagree = 1 to Strongly Agree = 5). P-values are from Kruskal-Wallis H tests comparing academic year groups. Number of students per cohort: 1 st Year, n = 96; 2nd Year, n = 69; 3rd Year, n = 69; 4th Year, n = 79; Final Year, n = 48 (total N = 361). * p ≤.05

Dimension C – Having the professional as a role model

In Dimension C (Table 5), it was found that the overall score for all the years is the same, i.e., there is no increase or decrease in the overall score of PIF in students of the First year to Final year MBBS for Dimension C, except in 4th year MBBS, where a slight drop in scores was observed.

Table 5.

Association of Having the Professional as a Role Model with Academic Year among Undergraduate Medical Students (n = 361)

Dimension C	Items	Academic Year					P Value
		Median (IQR)
		1 st Year	2nd Year	3rd Year	4th year	Final Year
Having the Professional as a Role Model	When working on problems/cases in class, I imagine myself to be in the shoes of a professional in my future Work environment.	4 (4–5)	4(4–5)	4(4–5)	4(3–4)	4(3–4)	< 0.001*
	I concentrate on what I believe I would need to know and be able to do when I enter in practical field	4 (4–5)	4 (4–5)	4 (4–5)	4 (4–4)	4 (4–4)	0.001*
	I believe I can already think and reason like a professional in a healthcare organization	4(3–4)	4(3–4)	4(3–4)	3(3–4)	4(3–4)	0.271
	I admire most, those teachers who are professionals in the area that I would like to enter.	4(3–5)	4(4–5)	4(4–5)	4(4–4)	4(3–4)	0.164
	I admire professionals who are already working in my future work environment.	4(4–5)	4(4–5)	4(4–5)	4(4–4)	4(3–4)	0.010*

Values are Median (IQR). Likert items were scored on a 1–5 scale (Strongly Disagree = 1 to Strongly Agree = 5). P-values are from Kruskal-Wallis H tests comparing academic year groups. Number of students per cohort: 1 st Year, n = 96; 2nd Year, n = 69; 3rd Year, n = 69; 4th Year, n = 79; Final Year, n = 48 (total N = 361). * p ≤.05

Dimension D – Professional self efficacy

In Professional Self Efficacy (Dimension D) Table 6, it was found that the overall score for all the years increased slightly, with only slight decrease in the scores of Fourth Year MBBS.

Table 6.

Association of Professional Self-Efficacy with Academic Year among Undergraduate Medical Students (n = 361)

Dimension D	Items	Academic Year					P Value
		Median (IQR)
		1 st Year	2nd Year	3rd Year	4th Year	Final Year
Professional Self-Efficacy	I am sure I will have no problems dressing and behaving professionally in my industry	4 (4–5)	4 (4–5)	4 (4–5)	4 (3–4)	4 (3–5)	0.001*
	I feel poorly prepared for a real job.	3 (2–4)	3 (2–4)	3 (2–3)	3 (2–4)	3 (2–3)	0.041*
	I believe that I will easily get along with my colleagues, get their cooperation, and have informal conversations with them	4 (3–4)	4 (4–4)	4 (3–4)	3 (3–4)	4 (3–4)	0.001*
	I am confident that I can do an excellent job in the future.	4 (4–5)	5 (4–5)	4 (4–5)	4 (3–4)	4 (4–5)	< 0.001*
	I have no doubt that I will master all the skills necessary to succeed in my future work	4 (4–5)	4 (4–5)	4 (4–5)	4 (3–4)	4 (3–4)	0.020*
	I am not sure about the kind of challenges faced by the professional in the industry I will work in	3 (2–4)	3 (2–4)	4 (3–4)	3 (3–4)	4 (3–4)	0.002*

Values are Median (IQR). Likert items were scored on a 1–5 scale (Strongly Disagree = 1 to Strongly Agree = 5). P-values are from Kruskal-Wallis H tests comparing academic year groups. Number of students per cohort: 1 st Year, n = 96; 2nd Year, n = 69; 3rd Year, n = 69; 4th Year, n = 79; Final Year, n = 48 (total N = 361). * p ≤.05

Dimension E – Preference for a particular Specialty

In Preference for a Particular Specialty Table 7 shows that Perceptions of degree alignment significantly declined from a median score of 5 (IQR 4–5) in early years to 4 (IQR 4–5) in the final two years (p =.020), while curriculum relevance remained statistically stable throughout (p =.255). Internship participation expanded significantly across year groups, rising from 10.4% in the first year to 54.2% by the final year (p <.001). However, student interactions with external industry professionals did not differ significantly across academic year groups (p =.380).

It was observed that for the 1 st year MBBS χ² = 4.501, p =.034 (Table 8), which infers that there is a statistically significant association between gender and the question – Do you already know what kind of work or specialty you prefer? Of the 59 students who said they already know what kind of work or specialty they prefer, 68% were female students.

The results of the students of second year MBBS (χ² = 0.009, p =.926), 3rd year MBBS (χ² = 1.442, p =.230), 4th year (χ² = 0.554, p =.457) and Final Year MBBS (χ² = 2.489, p =.115) (Table 8) indicates that there is no statistically significant association between gender and the question – Do you already know what kind of work or specialty you prefer? This means that both males and females knew equally well what kind of work or specialty they would prefer. Table 9 presents Mann-Whitney U test results comparing male and female students across the five PIFFS dimensions. Significant gender differences were found for Dimension C – Having the Professional as a Role Model (U = 17975.5, p =.021, r = −.14), Dimension D – Professional Self-Efficacy (U = 19475.5, p <.001, r = −.24), and Dimension E – Preference for a Particular Specialty (U = 18214.5, p =.011, r = −.16), with male students scoring significantly higher on all three dimensions. No significant gender differences were found for Dimension A (p =.132) or Dimension B (p =.248). Effect sizes were negligible to small (|r| range: 0.07–0.24).

Table 9.

Mann-Whitney U Test Results for Gender Comparisons Across PIFFS Dimensions (n = 361)

Dimension	U	p-value	r	Sig.
A – Knowledge About Professional Practices	14302.5	0.132	+ 0.09	ns
B – Experience with the Profession	16,867	0.248	−0.07	ns
C – Having the Professional as a Role Model	17975.5	0.021*	−0.14	Male > Female
D – Professional Self-Efficacy	19475.5	< 0.001***	−0.24	Male > Female
E – Preference for a Particular Specialty	18214.5	0.011*	−0.16	Male > Female

r = rank-biserial correlation (small ≥ 0.1, medium ≥ 0.3, large ≥ 0.5). Negative r = males scored higher than females. ns = not significant. * p ≤.05; ** p <.01; *** p <.001

Discussion

The higher proportion of female students in all the co-educational medical colleges of Pakistan [36] is similar to the global trend [37]. The final-year cohort contributed the smallest number of participants (n = 48), possibly reflecting the demanding schedules of clinical rotations.

The Professional Identity Formation (PIF) scores demonstrated a pattern of higher values in years one through three, lower values in the fourth-year cohort, and higher values again in the final year. As this is a cross-sectional study, these are between-group differences and may reflect cohort effects rather than within-student developmental change. Notably, data collection (November 2021–April 2022) coincided with the tail end of Pakistan’s COVID-19 fifth wave, and Year 4 students were among those most affected by clinical placement disruptions during this period; this contextual factor should be considered when interpreting the lower scores observed in that cohort. These patterns are broadly consistent with findings reported in the international literature [25, 38, 39].

Knowledge about professional practices (Dimension A)

Dimension A investigates students’ awareness of professional norms, responsibilities, and jargon. A decline in the 4th year may be due to students’ realization, upon clinical exposure, that theoretical knowledge alone is insufficient for real-world practice [21, 40–42].

Self-perceived understanding of medical roles differed significantly across year groups, notably for the items “I know the nature of the work I will do in my future” (p =.011) and “I have a good idea of the roles and responsibilities of my future job” (p =.003), with junior cohorts scoring higher than senior cohorts, particularly Year 4 (Table 2). By contrast, the wider literature reports that medical students’ awareness deepens over time, progressing from naive assumptions to practical understanding of their future roles [22, 43, 44].

It has been reported that knowledge often originates from media portrayals rather than actual clinical experiences, especially in early years [19, 45, 46]. Nevertheless, some institutions, integrate early ethics courses, which may contribute to an improved conceptualization of professional conduct [47].

A near-significant gender association (p =.054) was observed for the item “I know what kind of tools and equipment I will handle in my future occupation”. Thismay suggest female students are more conscientious in preparing for hands-on roles, particularly in male-dominated fields such as surgery [48].

Experience with the profession (Dimension B)

This dimension explores students’ real-life experiences and their developing self-image as professionals. Scores were high across all year groups and varied only modestly, broadly stable through the junior years with a slight decline in the senior cohorts. The significant difference between second- and fourth-year students may reflect the reorientation and initial stress that accompany the increased clinical responsibilities of the fourth year.

Statistically significant associations were found in three items, notably “I am working part-time or am running a business relating to what I’m studying” (p =.019). Final year students were more likely to be engaged in part-time work related to their field. Studies from Iran [49] and New Zealand [50] support the prevalence of part-time work among senior medical students for financial and skill-building purposes [51, 52].

Knowing someone in the same profession was also significantly associated with academic year group (p<.001), reinforcing the idea that mentorship and familial ties influence professional identity [53]. Research shows that such early influences, especially from family members, significantly shape career aspirations [54–56].

At the dimension level, no significant overall gender difference was observed for Dimension B (Mann-Whitney U: p =.248; Table 9). However, at the item level, statistically significant associations between gender and individual items were observed for most items (p <.05), with the exception of part-time work (p=.095). These item-level findings should be interpreted in the context of the non-significant dimension-level result. Male students, due to greater social freedom in contexts like Pakistan, may have more opportunities to engage in professional environments aiding PIF development. Conversely, both genders benefit from involvement in interest groups, which provide networking and career support [57].

Having a professional as a role model (Dimension C)

This dimension measures how students perceive learning from professionals in their field. Overall, scores remained consistent throughout the years, with a slight dip in the fourth year. Learning from role models is an active process involving critical observation and internalization of professional behaviours [6].

Significant relationships with academic progression were noted in items like “When working on problems/cases in class, I imagine myself in the shoes of a professional” (p <.001) and “I concentrate on what I believe I would need to know in the practical field” (p=.001). These findings are supported by studies indicating that final year students demonstrate greater awareness of critical professional skills [58–61]. The literature confirms that clinical mentors who reference evidence-based practices also serve as powerful role models, helping students better understand medical reasoning and application [62].

Admiration for teachers as role models did not differ significantly across year groups (p =.164), whereas admiration for professionals already working in students’ future work environment declined modestly in the senior cohorts (p=.010). Studies confirm that students often look up to faculty for their teaching abilities and professionalism with accessibility and commitment to teaching being particularly valued traits among role models [61, 63–65]. In contrast to the within-dimension consistency over time, a significant gender difference was observed for Dimension C (Mann-Whitney U = 17975.5, p =.021, r = −.14; Table 9), with male students scoring higher than female students on role model identification. Although the effect size was small, this finding suggests that role model identification may differ by gender, possibly reflecting differences in access to or perception of professional role models within the hierarchical structure of medical education in Pakistan. This warrants further qualitative investigation.

Professional self-efficacy (Dimension D)

Professional self-efficacy, or the belief in one’s ability to perform job responsibilities, progressed from the 1 st to 3rd year, dropped in the fourth year, and increased again in the final year. This trend aligns with the literature, which reports that despite increasing awareness, students still often feel unprepared for the real-world responsibilities [43].

Statistically significant associations were observed across all items (e.g., p =.001, p =.041), indicating a strong relationship between academic progression and self-efficacy.

At the dimension level, Dimension D showed the strongest gender difference of all five dimensions (Mann-Whitney U = 19475.5, p <.001, r = −.24; Table 9), with male students scoring significantly higher than female students on Professional Self-Efficacy overall. At the item level, however, one notable exception was observed: female students scored higher on the item “I believe that I will easily get along with my colleagues” (p=.003). This item reflects students’ anticipatory orientation toward peer relationships; the specific interpersonal competencies underlying this expectation cannot be inferred from this single item alone [66–69]. The contrasting direction of this item relative to the overall dimension-level finding warrants cautious interpretation.

Preference for a particular specialty (Dimension E)

The specialty preferences evolve as students progress through their education [42, 70–73]. In the first year, a significant gender difference was found; among the 59 first-year students who reported having already identified a preferred specialty, 68% were female (p =.034). In later years, both genders demonstrated comparable decisiveness.

Literature indicates that career decisions are influenced by mentorship [59], understanding of work environments [59], cultural factors [59, 71], and intrinsic motivations [74, 75]. In Pakistan, cultural norms [76] and parental advice [77] and convenient working hours [78] significantly shape female students’ choices, especially when considering work-life balance and social expectations [77].

These findings suggest that academic exposure and awareness of medical specialties are associated with greater career decisiveness. Notably, the overall PIFFS scores across all years remained moderate (3.5–3.8 out of 5). The consistently lower scores in the fourth-year cohort across multiple dimensions — a between-group pattern that may partly reflect COVID-19 disruptions during data collection — likely mirror the reality shock of early clinical exposure, where students confront the gap between theoretical ideals and clinical realities, facing increased responsibilities, and navigating the complexities of the healthcare system for the first time. This pattern warrants further investigation and suggests that the fourth-year period may benefit from enhanced mentorship and reflective learning support; longitudinal studies are needed to determine whether this represents a developmental effect or a cohort-specific phenomenon.

Conclusion

PIF among medical students is a multifactorial phenomenon. This cross-sectional study found significant between-group differences in PIF scores across MBBS year groups, with consistently lower scores in the fourth-year cohort across multiple dimensions. These differences may reflect developmental variation, cohort-specific factors, or contextual influences such as pandemic-related disruptions, and should not be interpreted as evidence of within-student change [79]. These findings allow us to focus on strategies like role models, mentors, and curricular and co-curricular activities to promote PIF among medical students and integrate some constructive strategies into the curriculum. Findyartini et al. reported higher PIF scores among senior students [25].

The moderate overall PIFFS scores (3.5–3.8 out of 5) suggest that PIF is still developing across the undergraduate years. These findings highlight the value of structured strategies — including role modeling, mentorship, and curricular and co-curricular activities — to support PIF development. Final-year students recorded the highest overall PIF scores, although the pattern across senior years was not uniform: the fourth-year cohort recorded the lowest. Longitudinal, multi-centre, and qualitative studies across Pakistani medical institutions are needed to establish whether the observed between-group patterns reflect developmental change, cohort effects, or contextual factors.

Limitations

This study has several limitations that should be considered when interpreting the findings. First, the cross-sectional design compares independent cohorts and cannot establish within-student developmental change; between-group differences may reflect cohort effects rather than individual progression. Second, the study was conducted at a single institution in Karachi, limiting generalisability to other medical colleges in Pakistan or other settings. Third, all data are based on student self-report, which is subject to social desirability bias and may not reflect actual professional behaviours or identity. Fourth, data were collected between November 2021 and April 2022, a period overlapping with COVID-19 disruptions in Pakistan, which may have differentially affected year groups — particularly Year 4 students whose clinical placements were most disrupted — and limits interpretation of findings in non-pandemic settings. Fifth, the year groups were unequal in size, which may have affected statistical power for smaller cohorts; in addition, the final-year cohort was the only majority-male group (30 male, 18 female students), and as male students scored higher on Dimensions C, D, and E, the comparatively higher final-year scores may partly reflect gender composition rather than year-group differences. Sixth, the PIFFS was modified for this study without formal re-validation in the Pakistani context, and per-dimension reliability estimates should be interpreted accordingly. Seventh, although Benjamini-Hochberg correction was applied to post-hoc comparisons, the conduct of five simultaneous Kruskal-Wallis tests carries a residual risk of Type I error. Eighth, the online administration of the Google Form did not include a formal mechanism to prevent duplicate submissions beyond a single-response-per-account setting. Ninth, although a universal sampling approach was employed, the final sample is self-selected (participation was voluntary), and non-responders (20%) may differ systematically from respondents. Finally, the study relies entirely on quantitative self-report data and lacks qualitative data to explain the fourth-year pattern; the “reality shock” hypothesis is plausible but was not directly measured and should be treated as speculative pending qualitative or longitudinal investigation.

Abbreviations

PIF

Professional Identity Formation

PI

Professional Identity

PBL

Problem Based Learning

PIE

Professional Identity Essay

PIFFS

Professional Identity Five Factor Scale

QPID

Questionnaire on Professional Identity Development

Authors’ contributions

BS conceived and designed the study, collected the data, conducted the statistical analysis, and drafted the manuscript. SKA revised the manuscript critically. SS participated in the discussion. SKA and SS contributed to the study design and provided supervisory guidance. All authors read and approved the final manuscript.

Funding

No funding was received for conduct of this research.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The ethical approval of the proposed study was sought from the Institutional Review Board (IRB) of Jinnah Sindh Medical University (Ref No: JSMU/IRB/2021/-478). Institutional ethical approval was additionally obtained from Liaquat College of Medicine and Dentistry (Reference No: IRB/M-000017/21), where the study was conducted. Informed consent from study participants was taken electronically. The consent form was attached at the beginning of the questionnaire, mentioning all the information about the study and options of voluntary participation.

Competing interests

The authors declare no competing interests.