Work-integrated learning in non-professional degrees: Evidence from twelve cohorts of social sciences internships in Hong Kong

Kolb’s experiential learning theory and Holmes’ employability model

Two theoretical frameworks guide this study’s analytical approach. The first is Kolb’s (1984) experiential learning theory, which conceptualises learning as an iterative four-stage cycle: concrete experience, reflective observation, abstract conceptualization, and active experimentation. Applied to internships, this framework implies that workplace placement alone is insufficient for meaningful learning. Students must also have structured opportunities to reflect on their experience, integrate it with prior conceptual knowledge, and experiment with new behaviours— a cycle that depends on both the quality of the placement and the institutional supports surrounding it. Internships that interrupt this cycle, for example through repetitive or irrelevant tasks, inadequate supervision, or limited reflective guidance, may generate engagement short of genuine learning. We use Kolb’s framework to interpret how different institutional arrangements enabled or disrupted the learning cycle across our cohorts. Internship experience, as the core praxis of this framework, enables students to connect academic knowledge with application: the iterative process of “learning, doing, and reflecting” supports personal discovery and growth (Lim and Mustafa, 2013; Simons et al., 2012).

The second framework is Holmes’ (2013) graduate employability model, which reorients attention from “skills possession,” the assumption that a list of competencies can be acquired and demonstrated, toward the more dynamic processes of identity construction and the negotiation of graduate capital in workplace settings. For Holmes, employability is not a fixed attribute but an ongoing achievement, shaped by how graduates position themselves in relation to professional communities and labour market expectations. This is particularly relevant for social sciences graduates, for whom the pathway from degree to career is rarely linear and the relevance of academic content to specific occupational roles is seldom obvious. We draw on Holmes throughout to interpret how internships function as sites where students begin to construct employability identities, navigate workplace cultures, and negotiate the value of their academic formation, including in cases where that negotiation is difficult or ambivalent.

Together, these frameworks position the analysis toward process rather than outcome, attending equally to cases where learning is disrupted, supervision weak, or identity formation frustrated. In what follows, Kolb’s four-stage cycle guides the account of how internship conditions enabled or disrupted learning across four developmental phases, while Holmes’ identity model considers how students came to position themselves as future graduates over time.

Internships in HE: From professional to non-professional degrees

Internships have become an established component of university-industry engagement in professional degree programmes worldwide, where placement is typically tied to occupational accreditation, providing students with a shared professional curriculum and clear benchmarks for competence (Chang and Chu, 2009; Rawlinson and Dewhurst, 2013; Shin et al., 2013; Smith and Smith, 2010). Beyond individual development, internships in professional fields serve institutional functions by strengthening university-industry partnerships and contributing to the “entrepreneurial university” and “triple helix” dynamics prominent in HE strategy (Etzkowitz and Zhou, 2017; Ferreira and Steenkamp, 2015; Garomssa, 2024). For employers, hosting interns provides an efficient pipeline for evaluating prospective hires (Gauthier, 2024; Smith and Green, 2021).

The extension of internship models to non-professional degrees, including social sciences, is a more recent and less theorised development. In the absence of an occupational accreditation framework, the relationship between degree content and internship tasks is less pre-defined, and the burden of making that clear connection falls more heavily on institutional design and individual student reflection (Grantz and Thomas, 1996). Employers hosting social sciences students typically value dispositions and generic competencies (e.g., adaptability, communication, critical thinking, teamwork) rather than discipline-specific knowledge, creating a structural challenge for theory-practice integration. This challenge has become more salient as demand for “soft skills” grows alongside technical literacy in an era of AI-driven automation (Castrillon, 2022; Dondi et al., 2021). The expanding connections between HEIs and industries in non-professional fields have accordingly been welcomed by university management and funding bodies as a mechanism for demonstrating graduate outcomes, as what Dunbar-Morris and Lowe (2023) describe as the “value for money” logic now applied to degree programmes. Faculty members have also found that such partnerships generate opportunities for knowledge exchange and network building beyond academia (Marques et al., 2024; O’Dwyer et al., 2023).

Hong Kong’s performance-driven HE context

Hong Kong’s HEI system is structured around eight publicly funded universities overseen by the University Grants Committee (UGC), which distributes funding and sets performance expectations. Since the 2010s, demographic decline and the expansion of degree places have intensified competition among institutions, placing graduate employability outcomes at the centre of institutional strategy and funding negotiations (Burns, 2020; Lo and Li, 2023). University management has accordingly sought to enhance programme competitiveness through curriculum reform, industry outreach, and the embedding of work-integrated learning modules.

Employability has consequently become a salient concern across many degree programmes in Hong Kong, with learning and assessment increasingly framed in terms of work-related outcomes. It is important, however, to distinguish two senses of “vocational” that are frequently conflated in both policy discourse and academic writings. The first refers to technical and vocational education and training (VET) as a distinct educational sector or tracking system. Research from China and comparable contexts demonstrates that VET participation can carry social stigma and is commonly associated with lower socioeconomic status and academic prestige (Wang, 2022; Wang and Guo, 2019). Students in such systems often perceive academic degree pathways as preferable, a preference shaped not simply by what they learn but by how their credentials are socially valued. This distinction matters when interpreting claims about “vocationalism” in HE.

The second, and analytically distinct, sense is what we term vocationalization within HE— a policy and curricular orientation in which universities, including those offering academic and generalist programmes, increasingly organise learning around demonstrated graduate outcomes, “career readiness,” and labour-market relevance through employability skills, work-integrated learning, and outcomes-oriented assessment. This vocationalization does not redirect students into VET tracks. Rather, it reframes the purposes and performance metrics of academic study within the university sector itself (Smith and Smith, 2010). Critically, this pressure can intensify precisely because students and families continue to value academic degrees: universities face growing expectations from students, employers, and policymakers to demonstrate how non-professional programmes promote employability, even when degrees carry no occupational accreditation. It is in this second sense— vocationalization as a policy logic applied to academic programmes — that characterises the Hong Kong HE landscape, making credit-bearing internships a strategic institutional response rather than merely a curricular enhancement.

Social sciences internships in Hong Kong’s universities

Building on the model established in professional programmes, social sciences curriculum planners in Hong Kong’s HEIs began introducing internship modules in the 2010s to foster career readiness and develop the transferable generic skills demanded across a range of industry contexts. Most social sciences programmes’ prospectuses stated goals such as addressing “societal needs” through training oriented toward career readiness and employability. These internships vary in structure: some are credit-bearing, others voluntary; some offer remuneration to students, others do not; some are undertaken during term time, others in the summer. Despite this diversity, most schemes make explicit reference to internationally recognised competency frameworks, such as the U.S. National Association of Colleges and Employers (NACE), and emphasise generic attributes including communication, teamwork, problem-solving, adaptability, and time management (Scholz et al., 2004).

The typical operational model involves students selecting from a list of vacancies offered by partner organisations, completing a supervised placement of at least 150–200 hours, and submitting reflective assignments (e.g., learning logs, internship reports, or presentations) assessed by both an academic supervisor and an industry mentor. Faculty members serving as internship coordinators manage the placement cycle, broker and sustain university-industry relationships, run preparatory workshops on résumé writing and workplace etiquette, and conduct mid-placement visits. This tripartite structure, in which students, faculty, and host organisations each carry defined responsibilities, is the institutional form that our case study analyses across years of implementation, and it is the configuration through which Kolb’s experiential learning cycle and Holmes’ identity-formation processes are either supported or disrupted.

Research design and positionality

This study adopts a reflective-practitioner approach combined with grounded theoretical coding techniques to analyse longitudinal qualitative and structured data collected across 12 years of internship programme administration. The research team comprises five faculty members who have served in supervisory, coordinative, and programme-leadership roles since the scheme's inception in the mid-2010s, with one author serving as programme leader for five years during a period of substantial transformation. This dual positioning, as both administrators and researchers, affords unique insider access to programme development, the logistical challenges of tripartite coordination, and the evolving mindsets of successive student cohorts. It also introduces the risk of confirmatory bias and promotional framing. To mitigate these risks, we anonymised all student and partner data, and incorporated negative and critical cases as analytically constitutive rather than incidental.

Data sources

The dataset begins in 2012, when the first cohort of students admitted to the programme (from the 2010 intake) undertook their credit-bearing summer internship; no prior records exist because the programme had not yet produced students at the internship stage. Across the 12 cohorts (2012–2023), approximately 670 students completed the internship, generating 554 student evaluation survey responses (response rate approximately 80%), over 150 partner organisation evaluation forms from 20 to 38 participating organisations annually, and an archive of programme committee minutes and internship operation documents spanning the full period. Table 1 summarises these data sources, and analytical uses. Throughout the findings, sources are attributed using the shorthand labels in Table 1, followed by the relevant year or year range (e.g., Student Comment, 2017; Partner Survey, 2019). Quotations presented in the findings were selected to represent the range of experiential patterns identified through coding, including positive, mixed, and critical cases, and are attributed by year and source type. Student and partner names, and organisational identifiers have been removed or replaced with generic descriptors.

OPEN IN VIEWER

Table 1.

Overview of data sources and analytical scope (2012–2023).

Data source	Period	Nature of data	Analytical use
Student surveys	2012–2023	Students’ evaluation on learning outcomes, supervision, internship arrangements etc.	Identification of recurrent themes in skill development, identity formation, supervision quality, and perceived fairness
Student comments	2012–2023	Students’ qualitative comments on experiences, skills acquired, challenges, and career clarity	Open and focused coding of emergent experiential patterns and critical cases
Partner surveys	2012–2023	Internship partners’ ratings on interns’ performance, and programme design etc.	Cross-actor comparison of expectations, performance, and institutional fit
Partner comments	2012–2023	Internship partners’ qualitative comments on supervision of interns, interns’ skill readiness and gaps etc.	Identification of convergences and tensions between student and partners’ perspectives
Programme committee minutes and internship review	2012–2023	Periodic internship evaluations, and discussion of follow-up actions, design changes by the programme management team	Contextualisation of institutional responses and learning

Analytical approach

Data were analysed using an inductive coding approach informed by grounded theory principles (Charmaz, 2006; Corbin and Strauss, 2015). Initial open coding was conducted across all data sources to identify recurring experiences, concerns, and learning outcomes, including both positive and critical remarks. Through focused coding, these initial codes were iteratively grouped into higher-order analytical categories: expectation alignment, skill development, theory-practice integration, supervision and assessment, and institutional mediation. In selective coding, relationships among these categories were examined to identify the study’s core analytical claim that internship effectiveness in non-professional degree programmes depends not on placement alone but on active institutional mediation across the tripartite relationship among students, faculty, and host organisations. The five categories and their integration into a broader account, emerged inductively from the data rather than being imposed a priori, though theories of experiential learning (Kolb, 1984) and graduate employability (Holmes, 2013) served as sensitising concepts that informed the interpretive lens.

To ensure rigour we stepped back regularly to treat provisional categories as contingent, revising them whenever incoming data challenged prior groupings. Constant comparison was applied systematically across the three stakeholder groups (students, internship partners, and faculty) so that convergences and tensions between actor perspectives surfaced analytically rather than being resolved in favour of a single voice. A team member (the third author) also cross-checked and audited the coding process to confirm that categories were grounded in the raw data. Quantitative indicators from the Likert-scale survey items, including satisfaction percentages, skill self-assessments, and career clarity scores, are used selectively throughout the findings to contextualise qualitative patterns at key cohort junctures, while the qualitative corpus carries the main analytical weight.

While the five core thematic categories remained analytically stable across cohorts, their relative salience varied substantially over time, reflecting changes in institutional arrangements, external conditions, and students’ positioning within the internship process. This longitudinal variation in salience is itself an important finding, underscoring the distinctive nature of internships in non-professional social sciences programmes, where learning outcomes are not anchored to occupational certification but are continuously negotiated through institutional mediation, placement design, and broader socio-economic conditions.

Formative phase (2012–2015): Expectation alignment and early tensions

The formative years of internship were characterised not by failure of student effort, but by institutional underdevelopment. The programme was able to mobilise placements and generate meaningful learning for some students, but expectation alignment, theory-practice linkage, and assessment comparability remained significantly uneven. These features are consistent with what Billett (2011) identifies as the “relational contingency” of workplace learning, or the degree to which learning outcomes depend on the alignment of institutional guidance, supervisory practice, and student readiness.

Partner satisfaction data from the formative years, although collected with a less standardised instrument than later years, already demonstrates a consistent baseline. All partners were either “very satisfied” (16.7%) or “satisfied” (83.3%) with student intern performance in the programme’s first partner survey record. They viewed students as “very helpful,” working “efficiently and independently,” and noted that they “demonstrated high calibre in learning new knowledge.” Most responding organizations confirmed willingness to offer placements the following years, providing an early indicator of partnership sustainability (Table 2; Partner Surveys and Comments, 2012 and 2013).

OPEN IN VIEWER

Table 2.

Partner satisfaction and student learning indicators (2012–2023).

Year	Partner surveyed (N)	Partners’ overall satisfaction a	Internship well-organised b	Knowledge & skills helpful for internship b	Phase	Core self-reported skills c
2012	20	100% (Very Sat. 16.7% + Sat. 83.3%)	44.2%	53.5 | 48.8%	Formative	Task management; basic teamwork; report writing; communication
2013	20	100% (Very Sat. 16.7% + Sat. 83.3%)	68.9%	65.6% | 57.4%
2014	24	100% (Very Sat. 25% + Sat. 75%)	80%	62% | 66%
2015	20	100% (Very Sat. 46.2% + Sat. 53.8%)	94.1%	76.5% | 76.5%
2016	25	100% (Very Sat. 53.9% + Sat. 46.1%)	82.8%	64.1% | 68.8%	Consolidation & expansion	Problem-solving; communication (written & oral); leadership; critical thinking; data analysis; networking with colleagues
2017	28	100% (Very Sat. 52.9% + Sat. 47.1%)	83.6%	59.7% | 56.7%
2018	23	100% (Very Sat. 46.2% + Sat. 53.8%)	91.3%	84.8% | 93.5%
2019	25	100% (Very Sat. 53.3% + Sat. 46.7%)	85.7%	85.7% | 85.7%
2020	38	86.7% (Very Sat. 46.7% + Sat. 40%)	59.5%	59.4% | 59.4%	Pandemic disruptions	Self-organisation; digital literacy; resilience; proactive communication
2021	34	83.3% (Very Sat. 50% + Sat. 33.3%)	76.3%	71.1% | 76.3%
2022	28	83.3% (Very Sat. 33.3% + Sat. 52.4%)	66.7%	66.6% | 80%
2023	28	100% (Very Sat. 72.7% + Sat. 27.3%)	100%	93.8% | 100%	Post-pandemic adjustment	Interpersonal skills; persuasion; reflective practice; adaptability; professional values

Yet even within this positive picture, the early years’ data also revealed structural tension. Some partners noted that the internship would be more productive if both organisations and students had clearer mutual understanding of roles and expectations before placement began, a recurring theme that would inform later programme development. They also requested greater involvement in candidate selection, signals that partner engagement, while broadly positive, already required active institutional management (Partner Survey, 2012 and 2013). One commented that “the internship schedule was too tight, making it difficult for students to complete their assigned tasks fully” (Partner Surveys and Comments, 2014 and 2015).

As well, student feedback from the earliest cohorts repeatedly pointed to ambiguity around task scope and placement choice. In 2012, students raised concerns about an “unfair and non-transparent internship system” and noted that “the requirements of different internship partners are not clear enough,” generating uncertainty in selecting a suitable internship that align well with students’ preferences (Student Comment, 2012). These concerns reflected a typical expectation-alignment problem, in which students were uncertain not about workload per se, but about whether their placement would offer a coherent learning experience relevant to their programme of study. Student survey data are consistent with this picture. Ratings of programme organisation and of the degree’s knowledge relevance to placement both remained modest at best (Table 2), reflecting the uneven institutional support of the formative phase.

In fact, even within these early years, there were uneven experiences across placements. Some students described their organisation as “a good place to adopt what I have learnt from the lecture into the working environment,” explicitly crediting both the placement design and academic supervisor support for helping them connect coursework to workplace practice. These positive cases were not evenly shared, however. The contrast between well-managed placements and those offering little supervisory structure or role clarity foreshadowed what would become a persistent equity concern throughout the decade.

A second recurring theme from students was the challenge of translating social science learning into workplace outputs. Several students explicitly challenged the requirement to link internship work with academic content when the placement itself did not naturally demonstrate disciplinary connections. A 2012 student described how placement mismatch made it “quite difficult to link my internship with my course.” Similar remarks appeared in 2014–15, with students noting that coursework requirements were “difficult to relate theories to the internship duty,” and that communication between the institution and certain host organisations was “not enough” (Student Comments, 2012–2015). This highlights a structural dilemma. Unlike professional placements, internships in social sciences programmes do not build on a pre-existing vocational consensus about what students should learn and do. Theory-practice integration therefore becomes an institutional objective that must be actively promoted, not an automatic outcome of placement.

Because the internship was credit-bearing from the outset, students were also alert to fairness in assessment. Feedback from 2012 called for evaluation forms to be filled by supervisor and made visible to students, and raised concerns about “unfair payment” and uneven subsidies across placements. Partners, too, questioned whether grading could be equitable given variation in supervisors and work environments, and suggested that, if grades were required, they should rest on shared outputs assessed against a common academic standard (Student Survey and Comment, 2012).

These concerns prompted early institutional responses. Programme Committee meetings in 2014 and 2015 formally discussed transport and meal allowance disparities and asked faculty management to explore possible subsidies, an early recognition that equity in participation conditions mattered as much as equity in assessment (Programme Committee Minutes, 2014–2015). In fact, the question of paid versus unpaid placements added a further dimension. The programme’s predominantly NGO and civil society’s partnership base meant that stipends were limited and unevenly available, a structural feature that became more salient as cohort sizes grew and the range of placements widened.

Taken together, the formative phase is best characterised as one of institutional emergence. The infrastructure for tripartite collaboration existed but had not yet been routineised, leaving learning outcomes heavily dependent on placement-specific circumstances rather than consistent programme-level support. Two thematic concerns were most salient in this phase: expectation alignment between students, partners and the programme, and theory-practice integration, which could not be assumed given the non-vocational nature of social sciences placements. In Kolb’s (1984) terms, the formative phase disrupted the learning cycle at its first stage. Without a coherent institutional framework for placement, many students could not readily translate concrete experience into structured reflection, let alone progress toward abstract conceptualisation. The core skills students reported, such as task management, basic teamwork, report writing, reflect this early-stage character (Table 2).

Consolidation and expansion (2016–2019): Skill articulation and identity formation

The consolidation phase marked a shift in how students experienced and understood their internships. As programme infrastructure matured, with pre-internship workshops, standardised evaluation instruments, and more structured placement-matching protocols introduced in response to earlier partner feedback, students increasingly viewed the internship as an opportunity for transferable skill development and career exploration rather than an ambiguous workplace exposure. By 2018, partners were recommending joint mid-term reviews involving academic supervisors and students together, recognising that standardised forms alone were insufficient without structured three-way conversation to calibrate expectations and provide early support (Partner Survey, 2018).

Partner satisfaction data was consistently positive. From 2017 to 2019, all surveyed organisations rated student performance as very satisfied or satisfied, and confirmed willingness to offer placements the following years (Table 2; Partner Surveys, 2017–2019; Programme Committee Minutes, 2019). Partners described students as “willing to learn,” possessing “good teamwork spirit,” and able to complete tasks with minimum supervision. Several noted that students “are able to apply their theoretical knowledge in a practical way”, a notable contrast with the expectation-alignment and theory-practice difficulties as evident in the formative years (Partner Comments, 2017–2019).

Student survey data pointed in the same direction. By 2018, near-universal agreement was recorded that organisations provided genuine learning opportunities (97.8%) and that the internship enabled knowledge and skills consolidation (above 85% for 2018–19) (Table 2; Student Survey, 2018). Qualitative responses also showed a shift in how students described their learning. Rather than listing generic competencies, they narrated development in specific, situated terms. One intern recounted that the placement “taught me how to write business emails appropriately and effectively present my ideas in face-to-face meetings with my supervisor”; another described learning to “work according to the directives from above and collaborate across many teams and departments” in a large NGO and government contexts (Student Comments, 2018). These accounts suggest that students were becoming more capable of naming and evaluating what they had learned, the kind of reflective articulation that is particularly valued in non-professional graduate employment contexts (O'Dwyer et al., 2023). In Kolb’s (1984) terms, this points to a more consistent progression from concrete experience toward reflective observation, a meaningful advance on the formative years where such movement had been uneven across placements. The expanded skill repertoire recorded in Table 2 for this phase (e.g., problem-solving, critical thinking, data analysis) points to the more differentiated self-awareness that consolidation made possible.

A second pattern is students’ use of the internship as a reality test for career identity. Survey data show that the share of respondents reporting a clear career preference increased following internship completion (Student Surveys, 2017–2019), and qualitative responses reflected both affirmation and critical self-examination. One intern described the workplace as “a big theater of persuasion,” defining employability as the capacity to convince employers, colleagues, and oneself through communication, adaptability, and resilience: “We not only need to convince others… sometimes, we need to persuade and motivate ourselves, too.” Another intern acknowledged being “afraid of proactively voicing out my thoughts” and noted that “the reality in the work field might not be so romantic” (Student Comments, 2017–2019). These reflections illustrate that internships in social sciences can generate critical self-knowledge rather than mere affirmation of one’s capacity. It is consistent with Holmes's (2013) graduate employability model, in which students are not simply acquiring competencies but actively constructing a professional self-narrative and positioning themselves in relation to organisational norms and career expectations.

Institutional development during this phase also showed a maturing feedback pattern. A 2017 university-level graduate review provided independent corroboration, with graduates identifying “cooperation and teamwork,” “social interaction skills,” and “positive personality” as their three highest-rated post-graduation competencies, a profile consistent with what internship experience in non-professional programmes is designed to cultivate (Programme Committee Minutes, 2017–2018). Acting on student and partner feedback, the programme also introduced a technical report-writing workshop after these skill gaps were flagged, an example of the responsive, iterative development characteristic of the consolidation phase (Partner and Student Comments, 2018).

Despite overall consolidation, critical voices persisted. Students continued to raise concerns about task relevance, workload intensity, and the unequal physical demands of certain placements (Student Comments, 2016–2019). Financial equity remained unresolved. Requests for on-the-job allowances, first raised at committee level in 2014, were still outstanding. Some placements in partners with outdoor activities required “strong physical ability,” raising questions about whether students with health limitations or those assigned to more demanding sites faced different learning conditions (Student Surveys and Comments, 2016–2019). Nonetheless, the consolidation phase is most clearly distinguished by the salience of skill development as a thematic concern, in which students were not only completing placements but articulating, reflecting on, and in some cases critically evaluating what they had learned. These concerns are a reminder that because placements are heterogeneous and not anchored to a professional accreditation framework, active institutional mediation of expectation alignment and learning conditions remains necessary even during years of relative organisational stability.

Pandemic disruption (2020–2022): Isolation, and unequal learning conditions

The 2020–2022 period brought the programme’s latent structural vulnerabilities most sharply into view. In Summer 2020, interns were expected to complete placements across 38 partner organisations as planned. However, COVID-19 social distancing restrictions in Hong Kong meant that several partners requiring outdoor or face-to-face operations, including social service providers, environmental education centres, and elderly care organisations, could not sustain standard arrangements. Four students at one community-facing partner completed 150–160 hours of on-site internship before the site closed, far below the expected 200 hours of service hours. Others transitioned abruptly to work-from-home (WFH) arrangements not originally envisaged in their placement agreements (Programme Internship Review, September 2020).

A formal internal review conducted in September 2020 identified five operational challenges, including the inapplicability of WFH for environment- and interaction-dependent placements, premature (or short-noticed) termination of some placements; high supervisor turnover at certain organisations; large variation in required service hours; and, most critically, reduced communication between academic supervisors and student interns. The review noted that “almost all interns had not consulted with their academic supervisor in advance as they experienced the pandemic disruptions,” suggesting that the supervisory relationships of the programme had partially broken down (Programme Internship Review, September 2020).

Student survey and qualitative feedback corroborates this account. Students repeatedly described WFH arrangements as creating a “total absence of working atmosphere,” with reduced interaction as the defining limitation of their experience. Job tasks were frequently characterised as “less challenging and repetitive,” with specific examples including routine data entry, phone calling, and stocktaking. A minority of students used strongly negative terms in open-ended responses, including “worst learning experience” and “wasting time and efforts” (Student Surveys and Comments, 2020–2021). While not representative of the cohort as a whole, these constitute important negative cases signalling dissatisfaction severe enough to warrant sustained institutional attention, particularly where they coincide with accounts of repetitive task allocation and minimal supervisory contact.

Partner feedback points in the same direction. They noted that “learning and supervision become harder without a physical interface,” and that even regular online meetings “cannot fully facilitate coaching of interns.” One partner observed that interns “must come equipped with basic skills” such as managing email and prioritising tasks in order to function effectively under remote conditions, a demand that implicitly disadvantages students with less prior workplace exposure (Partner Survey and Comment, 2020). In Kolb’s (1984) terms, COVID-19 did not simply reduce placement hours. It disrupted the relational conditions, notably supervisory contact, physical co-presence, informal feedback, through which reflective observation and active experimentation ordinarily occur, leaving many students without the necessary institutional supports to move beyond the first stage of the learning cycle.

The pandemic also reinforced structural inequalities as in earlier phases. WFH capacity was itself unevenly distributed. Students from lower-income households or smaller living spaces faced greater difficulties maintaining productive working environments, while those placed at organisations with stronger digital infrastructure fared considerably better than those in smaller and resource-constrained partners with limited remote working capacity. The disruption of placement stipends at some partners also had direct financial consequences for students who had relied on them (Student Surveys and Comments, 2020–2021).

Faculty reflections from this period also shed light on the particular challenges facing students during this time. A colleague observed that students “experienced social isolation during pandemic, thus having fewer opportunities to work closely with others” (Programme Committee Minutes, 2021). This informed the programme’s decision to strengthen pre-internship preparation. The September 2020 review proposed mandatory pre-placement meetings between academic supervisors and interns, and skills-readiness checks at the end of each pre-internship workshop (Programme Internship Review, September 2020). Together, these adjustments reflected the programme’s capacity to learn from difficulty and respond institutionally. Notably, the skills students reported have shifted toward self-organisation, digital communication, and resilience (Table 2), reflecting the adaptive demands of remote and disrupted placement.

As a result of these disruptions, partner satisfaction fell to approximately 83–87% (Table 2), and student ratings of internship organisation and the relevance of degree knowledge to placement tasks declined noticeably relative to the consolidation phase. Of the five thematic concerns, supervision and assessment was the most salient in this phase. The pandemic did not simply reduce learning opportunities, it made the quality and continuity of supervisory relationships the decisive issue in determining whether meaningful learning occurred at all. The declines underlined how dependent the tripartite model is on regular and sustained interaction among all three parties.

Post-pandemic adjustment (2023): Institutional learning and persistent tensions

By 2023, survey indicators suggest partial but substantive recovery. Programme organisation was rated satisfactory or above by 82.8% of student respondents, up markedly from the pandemic-era low of 59.5%, and 92.2% reported having learned appropriate professional values and behaviours during their placement (Student Survey, 2023). Partner satisfaction returned to near pre-pandemic levels, with 100% of responding supervisors expressing satisfaction (Table 2).

This recovery is most visible in student accounts of how the programme itself operated. One intern described the departmental arrangements in notably specific terms: “The department provided me with a vast network of partner organizations… arranged briefing sessions… [and] keep tracking the progress of my work at the agency… a faculty member, as an academic supervisor, would visit my workplace to discuss in detail and evaluate my performance together with my on-site supervisor” (Student Comment, 2023). This account suggests how tripartite coordination, involving pre-placement briefing, ongoing monitoring, and dual-party evaluation, had been present in earlier years but became more explicitly articulated and consistently applied in the post-pandemic phase. Recovery, in other words, was achieved through institutionalised routines rather than through any assumption that placements would deliver comparable learning on their own. In Kolb’s (1984) terms, these practices represent a deliberate effort to support all four stages of the experiential learning cycle, rather than leaving any stage to emerge from the placement encounter alone.

Partners’ comments in 2023 also conveyed appreciation for students’ enthusiasm and communication skills, while flagging gaps in “basic knowledge and skills” and requesting more practical preparation before placement begins (Partner Survey and Comment, 2023). This recurring tension between positive relational performance and weaker practical preparation is structurally characteristic of internships embedded in non-professional degrees, reflecting the ongoing difficulty of translating social science curricula into diverse workplace contexts rather than any simple deficit on the part of students.

Student reflections from the post-pandemic cohort engaged this tension with some sophistication. Several described learning to navigate the gap between the workplace expectations and the demands that organisational life actually makes. As one intern concluded: “We not only need to convince others to believe in us… sometimes, we need to persuade and motivate ourselves, too” (Student Comment, 2023). The re-emergence of this kind of identity-construction language in 2023, largely absent during the pandemic, points to a resumption of the process Holmes (2013) describes, in which students position themselves relative to professional communities and begin negotiating the value of their academic formation in workplace settings. The 2023 skill profile, from interpersonal skills, persuasion, reflective practice, adaptability, and professional values (Table 2), is the most sophisticated recorded across the cohorts, consistent with the identity-formation emphasis of this phase.

Placement variability, nonetheless, remains a persistent concern. A non-trivial minority expressed neutral or dissatisfied responses on key learning outcome items, including 17.2% neutral and 4.7% dissatisfied on knowledge and skills consolidation, indicating that some students continue to experience weaker learning alignment or less supportive supervisory environments (Student Survey, 2023). As one faculty member observed: “I always remind them that different institutions have different strengths and that they should do their best, whether they are sent to a small or big organization” (Programme Committee Minutes, 2023). While it is the programme’s aspiration to ensure equitable engagement across all placements, the data suggest that learning outcomes continue to vary. Of the five thematic concerns, institutional mediation was the most salient in this phase. The recovery was not self-generating, but an outcome of deliberate, structured programme intervention across briefing, supervision, and assessment.

Shifting salience of thematic categories across four phases

The first and analytically distinctive finding is that the five thematic categories identified through grounded coding did not operate with uniform prominence across the internship’s four phases under analysis. Rather than treating expectation alignment, skill development, theory-practice integration, supervision and assessment, and institutional mediation as parallel and stable dimensions of internship experience over the years, the longitudinal data reveal a shifting hierarchy of salience in which each category becomes most prominent when the conditions that generate it are most acute. Table 3 shows this variation and offers an interpretive note for each theme.

OPEN IN VIEWER

Table 3.

Relative salience of five thematic categories across four phases.

Thematic category	Phase 1 2012-15	Phase 2 2016-19	Phase 3 2020-22	Phase 4 2023+	Note on shifting salience
Expectation alignment	•••	••	•	•	Dominant at outset when programme lacked structured matching and pre-placement briefing; progressively managed through institutional routines but never fully resolved across heterogeneous placements.
Skill development	•	•••	•	••	Most prominent in consolidation when students recognise specific competency gains; partially suppressed during pandemic by remote and repetitive task conditions; re-emerges in recovery.
Theory-practice integration	••	••	•	••	Acute early when disciplinary connections are ambiguous; recedes mid-period as placement quality improves; returns as persistent challenge after the pandemic where partner critiques of knowledge application recur.
Supervision & assessment	•	••	•••	••	Most salient during pandemic when face-to-face arrangements became very challenging, exposing supervisory fragility; assessment fairness concerns run through all periods but peak in crisis.
Institutional mediation	•	••	••	•••	Grows more prominent as the programme learns; reaches highest salience in pandemic (reactive) and post-pandemic (proactive) phases. The clearest marker of programme maturation.

The pattern in Table 3 carries several implications of broader relevance to internship design and management. First, the progressive rise of institutional mediation as a salience category across all four phases, from relatively background in the formative phase to dominant after the pandemic, is not simply a story of programme improvement. It reflects a structural feature of internships in non-professional degrees. Because there is no pre-existing vocational consensus about what students should learn and do in the workplace, the institution must actively construct and maintain that alignment, through briefing, matching, supervision, and assessment, rather than assuming it will emerge from placement participation alone. These works do not become less important as the programme matures. They become more visible, more differentiated, and more indispensable.

Second, the pandemic period’s sharp increase of supervision and assessment salience is best understood not as an anomaly but as a stress test that exposed fragility in supervisory relationships that were functioning adequately under normal conditions, but proved fragile once sustained interaction among all three parties was absent. Third, the recurrence of theory-practice integration as a concern in both the formative and post-pandemic phases, despite a decade of programme development between them, suggests that this tension is intrinsic to social science placements rather than a problem that institutional learning can solve once and for all.

These patterns resonate with Kolb’s (1984) experiential learning framework. In the formative phase, the salience of expectation alignment can be traced to the disruptions at the concrete experience stage of the learning cycle, in which students could not engage in productive reflective observation because the terms of the experience itself were unclear. The prominence of skill development in the consolidation phase points to a more consistent progression from concrete experience toward reflective observation and, in some cases, the beginnings of abstract conceptualisation amongst students, as they became more capable of identifying, situating and reflecting on what they had learned. The pandemic disrupted the conditions through which reflective observation and active experimentation would ordinarily occur, leaving many students without the institutional support needed to move beyond the first stage of the cycle. The post-pandemic phase’s renewed attention to institutional mediation represents a deliberate effort to re-establish all four stages on more explicit and solid basis.

On the other hand, Holmes’ (2013) graduate employability model offers a complementary lens. The progression from the formative to consolidation phases demonstrates a movement from emergent possibilities to consolidated identities. Students in later cohorts were not merely acquiring skills but actively constructing a professional self-narrative, positioning themselves relative to organisational norms and career expectations in ways that earlier cohorts had not yet learned to do. As Table 2 shows, the shift in self-reported core skills across phases, from basic task management and communication in the formative years to problem-solving, critical thinking, and networking in consolidation, and interpersonal skills, persuasion, and reflective practice by 2023, captures this trajectory in concrete terms.

Internship as a tripartite learning and feedback process

A second key finding is that the internship functions not only as a site where students develop employability capabilities through cycles of learning, doing, and reflecting, but also as a feedback process through which all three parties recalibrate their expectations and practices. The “Learning, Doing, and Reflecting” dynamic of experiential learning (Lim and Mustafa, 2013; Simons et al., 2012) was not unidirectional. The dynamics have been mutually constitutive. Students adjusted self-understanding and career expectations; employers updated their models of graduate preparedness; and faculty revised briefing protocols, assessment rubrics, and partner-matching criteria in response to evidence from all three parties. Over time, this iterative exchange revealed what worked, and what did not, in internship practice. Furthermore, students’ reflections emphasise growing self-understanding and initiative, what Holmes (2013) describes as the consolidation of an employability identity through interaction with workplace others, while employer comments and programme records document the ongoing work of aligning expectations, supporting reflection, and monitoring supervision quality, a form of institutional mediation that the workplace learning literature identifies as central to effective learning in practice settings (Billett, 2011).

Internship outcomes are best understood as jointly produced, a view supported by studies of work-integrated learning which consistently find placement quality, supervisory support, and institutional coordination to be stronger predictors of learning than duration of placement alone (Lim and Mustafa, 2013; Scholz et al., 2004). Employability gains are most consistent where experience is supported by institutional mediation and workplace supervision rather than assumed to emerge spontaneously. The pandemic underscores this by exposing the limits of remote supervision, a finding consistent with evidence from programmes that shifted online during 2020–2022, where the absence of situated interaction weakened conditions for skill acquisition and identity formation (Billett, 2011; Smith and Green, 2021), and a structural question that extends well beyond the COVID period for any programme that depends on sustained relational contact among students, supervisors, and faculty.

Internships and inequality of learning experience

A third implication is that credit-bearing internships in non-professional degrees can reproduce or amplify educational inequality, even when overall satisfaction rates are high. Because placements are heterogeneous, and not anchored to a shared professional framework, students’ learning opportunities depend heavily on the quality of task design, supervision, and workplace integration in specific partner organisations. This creates situations in which some students gain meaningful responsibilities, sustained mentoring, and rich professional socialisation, while others encounter repetitive tasks, limited feedback, or weak supervisory support. When internships carry academic credit, this variation of experiences translate into unequal learning outcomes under formally curricular conditions, particularly where students also perceive uneven evaluation standards across partners (Frenette, 2013; Perlin, 2011).

Moreover, our analysis also points to resource inequality embedded in students’ participation. Student concerns about unpaid placements, and incidental expenses indicate that internship is not cost-neutral. Those with fewer financial resources may face greater strain, reduced flexibility in placement choice, or the need to maintain paid employment alongside their internship, with consequences not only for wellbeing but for learning, since constrained students may have less time for the reflective and networking activities that convert placement experience into employability capital (Shade and Jacobson, 2015). This variation is shaped largely by partner organisation resources rather than any programme rules, meaning that the financial burden falls unevenly on students with little control by the programme. To mitigate these challenges, early and clear information about financial support at the matching stage, and more consistent assessment standards across placement sites, are important steps toward preventing a well-intentioned credit-bearing scheme from inadvertently favouring students with greater economic and social capital.