Frontiers | Employability, work readiness and job alignment among Qatari STEM graduates of international branch campuses: evidence from a graduate survey

Employability, work readiness and job alignment among Qatari STEM graduates of international branch campuses: evidence from a graduate survey

• Khalifa Fadol Alyafei ^*

• Mohammed Abdul-Hai Issa

• Qatar Finance and Business Academy (QFBA), Doha, Qatar

Article metrics

View details

Abstract

International branch campuses (IBCs) deliver foreign science, technology, engineering and mathematics (STEM) qualifications inside host economies, yet how graduates’ reported educational experiences relate to their professional outcomes has received little attention in the Gulf. This study examined employability, work readiness and job alignment among Qatari nationals who completed a bachelor’s degree at one of two United States IBCs in Qatar. A cross-sectional survey of 75 employed graduates, recruited purposively through professional networks and snowball referrals, was analysed using Fisher exact tests with exact odds ratios and effect sizes, together with parsimonious binary, ordinal and Firth penalised logistic regression with Wald intervals for coefficients and bootstrap percentile intervals for average predicted probabilities; responses to one open-ended question were categorised descriptively. Most graduates worked in roles related to their major (86.7%) and entered employment immediately after graduation (61.3%). Labour-market preparedness was associated with very high perceived employability (odds ratio 4.9, 95% CI 1.6–14.7) and with a very smooth transition to work (Firth odds ratio 13.9, 95% CI 3.5–54.8), and professional ability development was also associated with very high employability (odds ratio 6.5, 95% CI 1.9–22.4). Job alignment corresponded with smoother transitions, though imprecisely (odds ratio 8.6, 95% CI 1.0–73.0). Women were less likely than men to report being very well prepared (adjusted odds ratio 0.11, 95% CI 0.03–0.45); other gender comparisons were inconclusive. Graduates most often requested internships and practical exposure and locally grounded labour-market knowledge. The findings suggest that monitoring preparedness and occupational alignment alongside employment rates offers a more informative account of the professional experiences reported by graduates of international STEM programmes delivered locally.

1 Introduction

Graduate employability now sits at the centre of how universities are judged: what matters is less the credential awarded than the labour-market outcomes a programme enables (; ; ). Rapid technological and organisational change has sharpened this shift by reshaping the skills employers seek and the ways graduates must deploy them (; ). The question becomes more demanding still in economies that import foreign provision through international branch campuses (IBCs), because education designed elsewhere must produce graduates who can function in a local labour market with its own occupational structures, languages and institutions (; ). This study examines how internationally delivered STEM education relates to professional outcomes for Qatari bachelor’s graduates of two United States IBCs.

Employment status alone is a weak proxy for the deeper process through which education becomes effective workplace participation. Recent scholarship distinguishes employability from work readiness, and both from the quality of the education-to-work transition and the degree to which a role aligns with a graduate’s field of study (; ; ; ). A graduate can hold a job that does not use the capabilities a degree developed, can enter employment quickly yet find the transition difficult, or can acquire skills that never convert into perceived readiness (; ; ). These distinctions motivate our focus on preparedness, job alignment, perceived employability, transition smoothness and time to first employment as outcomes that reach beyond a binary of employed or not.

Little of this literature speaks directly to national graduates of Gulf IBCs. Doha hosts one of the region’s largest clusters of transnational provision, and Qatari nationals form a substantial share of IBC enrolments, yet systematic evidence on how these graduates fare in the local labour market remains thin (; ; ). The gap matters because the promise of an IBC rests precisely on the claim that internationally framed capabilities can be used locally.

The study makes three contributions. It provides quantitative evidence on the relationships among learning experience, institutional support, professional ability development, labour-market preparedness, job alignment, perceived employability and transition smoothness for this under-represented population. It applies a graduate-capital lens in a way that acknowledges this tradition already addresses the contextual mobilisation of educational resources; the specific contribution is to distinguish the international organisation of education from the international migration of the student by examining citizens who obtained a foreign qualification through a locally situated branch campus. This setting motivates the proposition that domestic location and citizenship need not make the relationship between internationally organised education and local professional practice automatic, since familiarity with a country is not the same as familiarity with its particular occupational, organisational and professional-language requirements (; ). This is offered as an interpretation of associations among separate self-reported dimensions; credential returns, a comparative institutional advantage, and the temporal or causal mechanisms of conversion are not established by the cross-sectional design. And it demonstrates a modelling strategy suited to a small, skewed graduate sample, relying on exact tests, effect sizes, penalised regression and robustness checks rather than models the data could not support (; ). Five research questions guide the analysis: what educational experiences, skills and employment outcomes graduates report (RQ1); how learning experience, university employment support and ability development are associated with preparedness (RQ2); how preparedness, ability development and job alignment are associated with perceived employability and transition smoothness (RQ3); whether outcomes differ by gender, institution and graduation cohort (RQ4); and what additional support graduates identify as necessary (RQ5).

2 Literature review, conceptual framework and context

2.1 Employability as more than skills possession

Employability is commonly defined as the achievements, understandings and personal attributes that make graduates more likely to gain and succeed in work (). The field has moved decisively beyond a possession view toward multidimensional and contextual accounts: the same skills yield different outcomes across individuals and settings, perceived employability and graduate identity matter alongside skill acquisition, and context shapes how resources convert into advantage (; ; ; ; ). A recent systematic review synthesises this multidimensional turn, showing that employability is enacted through transversal skills and authentic, collaborative learning rather than owned as a fixed stock (). Because graduate outcomes are also shaped by labour-market structure and institutional practice, employment rates alone can understate or overstate an institution’s contribution (; ). A study of IBC outcomes should therefore examine reported readiness and occupational fit alongside employment status, not only whether graduates are employed.

2.2 Graduate capital and the conversion problem

reframes employability as the accumulation and mobilisation of human, social, cultural, identity and psychological capitals, separating the acquisition of resources from their conversion into labour-market advantage. Later work extends the model to transitions, career ownership and the unequal conversion of graduate resources, increasingly foregrounding graduates’ own agency in deploying capital (; ). Research on international graduates is especially relevant here: show that such graduates must develop an agentic capital to decide how to mobilise their human, social and identity resources in a host labour market. Human capital theory supplies the underlying premise that education raises productive capability (), but current readings reject the assumption that capability automatically clears the market. We use human-capital conversion in this analytic sense, as a lens that foregrounds the translation step between educational capability and professional outcome rather than as a new grand theory.

2.3 Work readiness, career transition and perceived employability

A distinct strand separates work readiness from employability and from career adaptability, warning that interventions often measure one while claiming another (). Work readiness denotes the attributes that prepare graduates for early workplace performance; it overlaps with, but is not identical to, employability or the smoothness of transition (; ; ; ; ). Perceived employability, a graduate’s confidence in securing and maintaining suitable work, is an established construct that recent STEM research has examined by gender and field (; ). For an IBC population these distinctions carry particular weight: a graduate may feel employable in an international sense yet not be ready for the specific demands of a local market, which makes transition quality and readiness outcomes worth studying in their own right.

2.4 Practical experience, university-industry collaboration and alignment

Work-integrated learning, including internships, live projects and authentic assessment, is widely studied as an approach to employability development, although evidence of its causal effects on labour-market outcomes remains limited (; ; ; ; ). Review evidence points to transversal skills and sustained university-industry collaboration as the conditions under which employability skills are enacted in authentic settings, with internships and employer engagement recurring across contexts as high-value practices (). Recent curriculum-mapping research goes a step further, mapping curricula against the ESCO occupational taxonomy and job-vacancy signals to locate where curricular skill claims are, or are not, recognised in labour-market demand, and treating occupation-curriculum alignment as a measurable outcome (). That work informs our treatment of job alignment, the fit between a graduate’s role and field of study, as a substantive outcome rather than a detail of employment status, and it motivates employer involvement in curriculum review ().

2.5 International branch campuses and education at home

IBCs deliver a foreign institution’s programmes and credentials in a host country. Scholarship on the Gulf situates Doha among the regional centres of transnational higher education, where national students form a substantial share of enrolments and internationally framed study is valued as a locally delivered option (). Studies of IBC choice and of transnational graduates’ careers show that the value of such a degree depends on how well it connects to host-country requirements and credentials (; ; ). In the Gulf specifically, research on graduate preparedness in Qatar and on labour-market participation among nationals highlights gaps an imported curriculum does not automatically fill, including local labour-market knowledge and, in Qatar, professional Arabic competencies (; ). Outcomes evidence for national graduates of Gulf IBCs nonetheless remains scarce. That scarcity defines the space this study occupies: the education is international while the labour market is local, providing a context for examining whether graduates report support needs related to local professional practice.

2.6 Conceptual framework

The study applies graduate-capital theory to citizens obtaining foreign qualifications through locally situated branch campuses, distinguishing the international organisation of education from the necessity of relocating abroad to obtain the qualification. The conversion problem at the centre of graduate-capital theory, the translation of accumulated educational resources into labour-market advantage (; ), need not dissolve simply because geographic distance is removed. The branch-campus setting provides a distinct context for examining graduates’ reported preparedness and professional experiences, but does not isolate the influence of student migration or other contextual differences. A contextual proposition is that familiarity with the country need not eliminate occupational, organisational or professional-language support needs. The survey does not establish the origins of these needs or demonstrate that one source of friction has replaced another. The empirical contribution is evidence on associations and reported support needs within this setting, rather than identification of a distinct conversion mechanism.

Figure 1 organises the measured variables and the hypothesised associations considered in the analysis; it is not an estimated process model. Educational experience, comprising the overall learning experience and university employment support, is expected to be associated with professional ability development and labour-market preparedness. Preparedness and ability development, together with job alignment, are in turn expected to be associated with perceived employability and with the smoothness of the university-to-work transition. Current job alignment is an outcome in Model 5 and a covariate in Models 2 and 3; these statistical roles do not establish its temporal position relative to preparedness or transition. Arrows in the figure represent hypothesised conceptual associations rather than established causal effects; the framework guides variable selection and interpretation without asserting causal direction, which a cross-sectional design cannot establish.

2.7 Qatar and institutional context

Qatar’s long-term development framework, the Qatar National Vision 2030, sets out a transition toward a diversified, knowledge-based economy underpinned by a capable national workforce (). The current phase is guided by the Third National Development Strategy 2024–2030, which makes workforce nationalisation and the alignment of education with labour-market needs explicit priorities, including a higher share of Qatari nationals in the private and semi-private workforce (). These settings frame graduate employment for nationals as active workforce participation rather than credential acquisition alone.

Higher education provision in Qatar includes a cluster of IBCs in Education City alongside national institutions, with STEM prominent in the mix (). The two institutions studied here, Texas A&M University at Qatar and Carnegie Mellon University in Qatar, deliver United States engineering, computing and applied-science curricula to a predominantly local and regional student body. Both campuses opened in Qatar within approximately one year of each other, making institutional age an unlikely explanation for observed differences. Interpreting any campus contrast still requires caution: the institutions differ in disciplinary composition, programme orientation, graduate and gender composition, cohort profile and typical industry destinations, and a survey of this kind cannot separate those influences. The study was not designed as a ranking exercise, and institutional contrasts are treated as potentially reflecting programme mix and graduate composition rather than educational quality. The labour market itself is distinctive: employment for Qatari nationals is shaped by Qatarisation policy, a large public and energy sector, and a bilingual professional environment in which Arabic retains an important role alongside English (). These features provide context for interpreting the associations and support needs reported here; their roles were not directly tested.

3 Materials and methods

3.1 Study design

The study used a quantitative cross-sectional graduate survey supplemented by descriptive categorisation of responses to one open-ended question; the latter was not treated as an independent qualitative strand. Associational language is used throughout: estimates describe relationships observed in the data and do not establish that one variable caused another

3.2 Participants and recruitment

The target population comprised Qatari nationals who had completed a bachelor’s degree in engineering, computing or another applied STEM field at Texas A&M University at Qatar or Carnegie Mellon University in Qatar; postgraduate qualifications were outside the study frame. Graduation cohorts ranged from before 2015 to 2023 and later. Responses were collected through Microsoft Forms between 29 April and 26 September 2025.

Participants were recruited using purposive sampling through the researchers’ LinkedIn professional networks, supplemented by snowball referrals. Recruitment targeted employed Qatari nationals who had completed a bachelor’s degree in a STEM discipline at Carnegie Mellon University in Qatar or Texas A&M University at Qatar. During recruitment, the researchers assessed prospective participants against these eligibility criteria using their professional knowledge of the individuals and the information recorded on their LinkedIn profiles, including educational qualifications, institution, graduation status and employment history. This constituted profile-based eligibility screening rather than authentication against degree certificates, university records or official nationality documents. Seventy-six submissions were received, of which one test entry was excluded during data screening, leaving 75 respondents for analysis. The findings describe these surveyed respondents and are not estimates for all graduates of the two campuses or for Qatari STEM graduates nationally.

3.3 Survey development and administration

The questionnaire was developed from the employability and graduate-capital literature and administered through Microsoft Forms in English. Before data collection it was reviewed by experts and piloted with a small group to check the clarity and relevance of items. The instrument combined single-choice items on learning experience, labour-market preparedness, perceived employability, transition smoothness and professional ability development; two multiple-response items on skills gained and valuable knowledge; and one open-ended question on additional support required; its required or optional setting at collection could not be independently verified from the retained records. No psychometric validation is claimed and no internal-consistency coefficient is reported, which would not be appropriate for unrelated single-item and multiple-response measures. These items provide global evaluations rather than comprehensive measures of the underlying constructs, and their measurement error cannot be estimated from internal consistency. Retrospective evaluations may also be influenced by current employment experience, and collecting predictors and outcomes through the same questionnaire may contribute shared response tendencies whose influence cannot be separated from substantive associations in this design.

3.4 Measures and coding

Outcomes and predictors were coded as documented in the variable codebook. Labour-market preparedness was coded as very well prepared versus moderately prepared; perceived employability was retained as ordinal (moderate, high, very high) and also dichotomised at very high; transition smoothness was retained as ordinal (difficult, smooth, very smooth) and dichotomised at very smooth. Job alignment was respondents’ yes/no assessment of whether their current role was related to their STEM major; occupations were not independently classified using an external occupational taxonomy, and ambiguous or cross-functional roles were not separately adjudicated. Time to first job distinguished immediate from later entry. Professional ability development was coded as greatly versus moderately enhanced, learning experience as excellent versus good or lower, and university support as received versus not. A skill-breadth count was computed from the number of selected skills, and graduation cohort was grouped into before 2015, 2015–2018 and 2019 onward.

Table 1 presents the item wording, observed response options and analytical coding for each measure

Construct	Item (English)	Response options (observed)	Coding used in analysis
Overall learning experience	How would you rate your overall learning experience at the IBC?	Excellent; Good; Average	Binary: Excellent = 1 vs. Good/Average = 0
University employment support	Did your university support you in securing a job after graduation?	Yes; No	Binary: Yes = 1 vs. No = 0
Professional ability development	To what extent did the IBC experience enhance your talents and professional abilities?	Greatly enhanced; Moderately enhanced	Binary: Greatly enhanced = 1 vs. Moderately/less = 0
Labour-market preparedness	How prepared did you feel for the local labour market after graduation?	Very well prepared; Moderately prepared	Binary: Very well prepared = 1 vs. Moderately/less = 0
Perceived employability	How would you evaluate your employability after graduating from the IBC STEM program?	Very high; High; Moderate	Ordinal (Moderate < High < Very high); binary Very high = 1 vs. else = 0
Transition smoothness	How smooth was your transition from university to the workplace?	Very smooth; Smooth; Difficult	Ordinal (Difficult < Smooth < Very smooth); binary Very smooth = 1 vs. else = 0
Job alignment	Is your current role related to your STEM major?	Yes; No	Binary: Yes (STEM-aligned) = 1 vs. No = 0
Time to first employment	How long did it take you to secure your first job after graduation?	Immediately; Within 6 months; Within 1 year	Binary: Immediately = 1 vs. later = 0
Skills gained (multiple response)	Which skills did you gain most from studying at the IBC? (Select all that apply)	Seven predefined categories plus free-text write-ins	Skill-breadth = count of selections (with and without write-ins); each category = 1
Valuable knowledge areas (multiple response)	Which knowledge areas were most valuable in your current job? (Select all that apply)	Four predefined categories plus free-text write-ins	Each category = 1; practical industry experience is a valued knowledge area, not internship completion
Gender	Gender	Female; Male	Female = 1 vs. Male = 0
Institution	University Attended	CMUQ; TAMUQ	CMUQ = 1 vs. TAMUQ = 0
Graduation cohort	Graduation Year	Before 2015; 2015–2018; 2019–2022; 2023 onwards	Grouped Before 2015/2015–2018/2019 onward; Before 2015 = 1 vs. later (Model 5)

Measurement of constructs: item wording, observed response options, and coding applied. The central constructs are single categorical self-report items for which no psychometric validation or internal-consistency coefficient is claimed

3.5 Statistical analysis

Descriptive statistics summarised participant characteristics, employment outcomes and educational experience; for multiple-response items, percentages were calculated from all 75 respondents and do not sum to 100%. Bivariate associations were tested with Fisher’s exact tests, reported with exact conditional odds ratios, 95% confidence intervals and Cramer’s V, following standard procedures for categorical data (); Mann–Whitney U and Kruskal–Wallis tests handled ordinal comparisons. Because several models were estimated, a supplementary multiple-comparison assessment is reported in the Supplementary Materials, applying Holm and Benjamini–Hochberg control within each model with every coefficient included; p-values in the main text are nominal and 95% confidence intervals are pointwise, and the associations that remain significant under false-discovery-rate control are identified there. The bivariate associations in Table 5 are treated as preliminary and exploratory, and variance inflation factors were computed for the multivariable models to assess multicollinearity.

Multivariable models were deliberately parsimonious given the sample size and event counts. Model 1 examined being very well prepared in relation to ability development, learning experience and university support, with gender and institution as controls. The binary specifications compare the highest response category with all lower observed categories; these thresholds were selected after inspection of the response distributions and discard information, so ordinal specifications were retained as sensitivity analyses. Model 2 examined very high perceived employability in relation to preparedness, ability development and job alignment, with binary logistic regression as the primary specification and an ordinal model as sensitivity. Model 3 examined a very smooth transition using Firth penalised logistic regression, which corrects sparse-data bias and separation in small samples (; ). Two secondary models assessed immediate employment and job alignment; in the alignment model, graduation cohort was collapsed into before 2015 versus 2015 onward because only ten graduates held non-aligned roles. Cohort was otherwise examined through subgroup and sensitivity analyses, with age tested separately given its overlap with cohort. The three primary model specifications were defined from the research questions before the results were examined; the dichotomisation thresholds, the collapsing of sparse categories and the estimator choices were decided after inspecting the data and are described as such.

The revised analyses and reproduction workflow were executed in Python 3.11.15 using statsmodels, SciPy, NumPy, pandas and Matplotlib, with package versions recorded in the supplementary environment manifest, and the Firth estimates were validated independently using the R package brglm2 (adjusted-score estimation with mean bias reduction, equivalent to the Jeffreys-prior penalty); coefficient estimates agreed to two decimal places, and inference used Wald intervals from the model information at the bias-reduced estimates. For the two sparse Firth models, profile penalised-likelihood confidence intervals were additionally computed as a sensitivity check (Supplementary Materials), and the substantive conclusions were unchanged. For each figure, 5,000 bootstrap samples were attempted; models were refitted within each sample and predictions averaged over its covariate distribution, and percentile intervals were calculated from 4,558 accepted fits for Figure 2, 4,877 for Figure 3 and 4,991 for Figure 4. Rejection criteria and reasons, and the penalised sensitivity analyses for Figures 2, 3, are documented in the Supplementary Materials; the intervals are recomputations during revision, and unchanged point estimates do not imply unchanged intervals. Full diagnostics and robustness analyses appear in the Supplementary Materials. A complete, runnable reproduction package (the coding library, the full analysis pipeline, the pinned software environment, and the bootstrap, multiplicity and diagnostic records) is provided as Supplementary Materials; it reproduces every reported table, model and figure from a clean session and halts if any headline statistic fails to reproduce at its reporting precision.

3.6 Open-ended responses

Responses to the open question on additional support were categorised from the verbatim text using an explicit codebook and a documented respondent-level coding matrix. Because the original matrix linking responses to the submitted categories could not be located, the responses were recoded during revision, and the revised analysis replaces the originally submitted open-question results. The initial categorisation was assisted by artificial intelligence and reviewed by the author. As no independent second coder was involved, this component is treated as descriptive content categorisation, and no intercoder reliability coefficient is reported. Percentages were computed among 73 eligible responses, excluding one blank response and one response consisting solely of a full stop from the 75 analytical respondents. Explicit statements that no additional support was required were retained as meaningful responses; blank responses were treated as missing. One response expressing positive appraisal without specifying a support request remained in the denominator but received no substantive category code. The codebook and aggregate frequencies are provided as Supplementary Material; the participant-level coding matrix is available upon reasonable request, subject to institutional approval and the conditions of participant consent. No identifiable job titles or verbatim quotations are reported.

3.7 Data integrity

All results derive from the original raw survey file of 76 responses. A documented integrity check identified the first response as a test entry, based on a test value in the industry field, a non-genuine job title and invalid text in two open-ended fields, and it was removed. No duplicate records, identifiers or timestamp anomalies were found; all closed-ended items were complete. One response to the open question was blank and another consisted solely of a full stop. Both were excluded from the content categorisation, while both respondents were retained in all quantitative analyses. The final analytical dataset contained 75 retained responses after removal of the test entry, and no synthetic, generated, duplicated or augmented data were used at any stage. Distinct recorded submissions and the absence of duplicate timestamps do not by themselves establish statistical independence or that no repeat participation occurred.

3.8 Ethics

The study received institutional approval for data collection from the Dean of Academic Affairs, Qatar Finance and Business Academy (QFBA), in academic partnership with Northumbria University, on 24 September 2024. Participation was voluntary; electronic informed consent was presented as the first item in Microsoft Forms, and only respondents who agreed proceeded to the questionnaire. The questionnaire did not collect names or other direct identifiers. Because participants were recruited from the researchers’ professional networks, their individual records are treated as confidential rather than fully anonymous. The de-identified analytical dataset and coding matrix are not publicly available. Requests for access should be directed to the corresponding author and will be considered subject to institutional approval and the conditions of participant consent.

4 Results

4.1 Participant and employment profile

The 75 graduates were split between Carnegie Mellon University in Qatar (53.3%) and Texas A&M University at Qatar (46.7%); 56.0% were women. Graduation cohorts were evenly spread: 32.0% before 2015, 32.0% in 2015–2018 and 36.0% from 2019 onward. Most graduates worked in roles related to their STEM major (86.7%), the largest sectors being oil and gas (33.3%), finance and banking (18.7%) and information technology (17.3%). A majority entered employment immediately after graduation (61.3%), with a further 37.3% within six months. On the perceptual outcomes, 56.0% reported being very well prepared for the labour market, 57.3% rated their employability as very high, and 40.0% described their transition as very smooth, most of the remainder calling it smooth (Tables 2, 3).

Characteristic	Category	n	%
Age	20–24 years	12	16.0
25–29 years	23	30.7
30–34 years	22	29.3
35 + years	18	24.0
Gender	Female	42	56.0
Male	33	44.0
University	Carnegie Mellon University in Qatar (CMUQ)	40	53.3
Texas A&M University at Qatar (TAMUQ)	35	46.7
Cohort	Before 2015	24	32.0
2015–2018	24	32.0
2019 onward	27	36.0

Participant characteristics (n = 75)

Percentages are based on 75 respondents; totals may differ from 100% because of rounding

Outcome	Category	n	%
Sector	Oil & Gas	25	33.3
Information Technology	13	17.3
Finance & Banking	14	18.7
Education	8	10.7
Health & Healthcare	6	8.0
Aviation	5	6.7
Government & Public sector	2	2.7
Other	2	2.7
Role related to major	Yes	65	86.7
No	10	13.3
Time to first job	Immediately	46	61.3
Within 6 months	28	37.3
Within 1 year	1	1.3
Preparedness	Very well prepared	42	56.0
Moderately prepared	33	44.0
Perceived employability	Very high	43	57.3
High	28	37.3
Moderate	4	5.3
Transition	Very smooth	30	40.0
Smooth	40	53.3
Difficult	5	6.7

Employment profile and graduate outcomes (n = 75)

Percentages are based on 75 respondents; totals may differ from 100% because of rounding

4.2 Educational experience, skills and knowledge

Graduates evaluated their education positively. Four in five rated the overall learning experience as excellent (80.0%), two thirds reported receiving university employment support (66.7%), and 70.7% felt their professional abilities had been greatly enhanced. The most frequently reported skills were communication (81.3%), self-learning (78.7%), analytical thinking (77.3%) and technical expertise (70.7%), followed by teamwork (58.7%), project management (53.3%) and leadership (36.0%); on average graduates selected 4.6 skills. The knowledge resources valued most were research and problem-solving (64.0%), practical industry experience (60.0%), subject-specific knowledge (57.3%) and business and market understanding (42.7%) (Table 4). The pattern indicates strong self-reported acquisition of transferable capabilities alongside a lower emphasis on leadership and business understanding.

Domain	Category	n	%
Learning experience	Excellent	60	80.0
Good	14	18.7
Average	1	1.3
University support	Yes	50	66.7
No	25	33.3
Ability development	Greatly enhanced	53	70.7
Moderately enhanced	22	29.3
Skills gaineda
Technical expertise	53	70.7
Analytical thinking	58	77.3
Project management	40	53.3
Communication	61	81.3
Teamwork	44	58.7
Leadership	27	36.0
Self-learning	59	78.7
Valued knowledgea
Subject knowledge	43	57.3
Practical experience	45	60.0
Research/problem-solving	48	64.0
Business/market	32	42.7

Educational experience, skills and knowledge re

a

Multiple-response items. Percentages use all 75 respondents as the denominator and do not sum to 100%

4.3 Factors associated with labour-market preparedness

Table 5 reports the bivariate associations. An excellent learning experience was associated with greatly enhanced ability development (OR 11.7, 95% CI 2.8–60.3; Cramer’s V 0.48). Ability development and university support were each associated with being very well prepared (OR 4.1, 95% CI 1.3–14.2 and OR 3.4, 95% CI 1.1–10.7), as was an excellent learning experience (OR 7.2, 95% CI 1.7–44.3). Preparedness in turn was associated with very high employability (OR 6.2, 95% CI 2.1–20.0) and with a very smooth transition (OR 17.2, 95% CI 4.3–102.9; V 0.56), and greatly enhanced ability with very high employability (OR 8.3, 95% CI 2.4–34.3). Job alignment corresponded with a very smooth transition (Fisher p = 0.044), but only one non-aligned graduate reported a very smooth transition, so the exact interval is very wide (OR 7.1, 95% CI 0.9–328.5) and the estimate imprecise.

Association	OR	95% CI	Fisher p	Cramer’s V
Excellent learning/greatly enhanced ability	11.7	2.83–60.33	<0.001	0.48
University support/very well prepared	3.39	1.14–10.74	0.025	0.28
Greatly enhanced ability/very well prepared	4.08	1.29–14.16	0.010	0.31
Very well prepared/very high employability	6.22	2.09–20.04	<0.001	0.43
Very well prepared/very smooth transition	17.22	4.3–102.86	<0.001	0.56
Job aligned/very smooth transition	7.1	0.89–328.51	0.044	0.24
Greatly enhanced ability/very high employability	8.33	2.41–34.33	<0.001	0.45
Female/very well prepared	0.36	0.12–1.03	0.039	0.24

Bivariate educational, institutional and demographic associations: fisher exact tests, exact conditional odds ratios and confidence intervals

OR, odds ratio; CI, confidence interval. Two-sided Fisher p-values and exact confidence intervals use related but distinct procedures and may differ at the significance boundary. Only one respondent in a non-aligned job reported a very smooth transition. The association labels do not indicate established causal effects

In the multivariable preparedness model (adjusted Model 1, Table 6), the estimates for ability development, learning experience and university support were imprecise, with confidence intervals spanning one; the adjusted estimates were uncertain, and the analysis does not establish why their intervals were wide. The model’s apparent AUC was 0.804 and McFadden’s pseudo-R² was 0.252; these are in-sample fit summaries, not evidence of external predictive validity. Women were less likely than men to report being very well prepared (adjusted OR 0.11, 95% CI 0.03–0.45), an association stable under leave-one-out refitting (OR range 0.06–0.12) and Firth penalisation (OR 0.14, p = 0.004). The Carnegie Mellon indicator showed a higher adjusted probability that did not reach significance (OR 3.1, 95% CI 0.8–11.9) and is not read as a difference in educational quality. Figure 2 shows predicted probabilities of 0.22 for graduates reporting neither an excellent learning experience nor greatly enhanced ability, against 0.67 for those reporting both. Variance inflation factors for the Model 1 predictors were all below two (learning experience 1.82, ability development 1.36, university support 1.49, gender 1.10, institution 1.32), and did not indicate marked collinearity within this predictor set.

Model/predictor	OR	95% CI	p
Model 1: Very well prepared
Ability development (greatly)	2.71	0.67–10.87	0.160
Learning experience (excellent)	4.18	0.65–27.14	0.133
University support	1.83	0.45–7.39	0.398
Female	0.11	0.03–0.45	0.002
CMUQ (vs. TAMUQ)	3.13	0.83–11.85	0.093
[n = 75, events = 42, AUC = 0.804, pseudo-R2 = 0.252]
Model 2: Very high perceived employability
Very well prepared	4.92	1.64–14.72	0.004
Ability development (greatly)	6.5	1.89–22.36	0.003
Job aligned	0.72	0.14–3.61	0.693
[n = 75, events = 43, AUC = 0.792, pseudo-R2 = 0.236]
Model 3: Very smooth transition (Firth)
Very well prepared	13.91	3.53–54.83	<0.001
Ability development (greatly)	7.57	1.48–38.85	0.015
Job aligned	8.63	1.02–72.97	0.048
[Firth penalised; n = 75, events = 30]

Primary regression models 1–3 (odds ratios). Model 3 uses Firth penalised estimation

OR, odds ratio; CI, confidence interval; AUC, area under the receiver operating characteristic curve. Outcomes and reference categories are defined in Section 3.4. AUC describes apparent discrimination within the sample

4.4 Perceived employability

In Model 2, preparedness (OR 4.9, 95% CI 1.6–14.7) and ability development (OR 6.5, 95% CI 1.9–22.4) were each associated with very high employability, while the job-alignment estimate was imprecise (OR 0.72, 95% CI 0.14–3.61), providing insufficient evidence to determine the direction or magnitude of its association; discrimination was good (AUC 0.79). The binary model is primary; an ordinal sensitivity model is reported below and in the Supplementary Materials. Average predicted probabilities (Figure 3) were 0.71 (95% bootstrap percentile CI 0.56–0.85) for the very-well-prepared scenario and 0.39 (0.22–0.57) for the comparison scenario. Estimates were essentially unchanged with or without job alignment, and campus-specific sensitivity analyses produced directionally similar estimates with wide confidence intervals owing to the reduced subsample sizes (Supplementary Materials). The ordinal sensitivity model gave preparedness OR 5.43 (95% CI 1.85–15.93), ability development OR 6.34 (1.98–20.27) and job alignment OR 0.63 (0.13–3.00). The binary model for at-least-high versus moderate employability did not converge under the specified fitting procedure; only four respondents selected moderate employability, so the intended threshold-based proportional-odds assessment could not be completed reliably and the proportional-odds assumption was not established.

4.5 Transition smoothness

Model 3, estimated with Firth penalised regression, found preparedness (OR 13.9, 95% CI 3.5–54.8), ability development (OR 7.6, 95% CI 1.5–38.8) and job alignment (OR 8.6, 95% CI 1.0–73.0) each associated with a very smooth transition. Job alignment showed a positive association, although the estimate was imprecise and close to the significance boundary. Ordinary logistic estimates were larger and less stable, the expected behaviour when penalisation corrects sparse data, and adding institution left the core estimates unchanged (Supplementary Materials). Predicted probabilities from the penalised model (Figure 4) ranged from 0.02 for graduates who were neither very well prepared nor in an aligned role to 0.64 for those who were both; among the four model scenarios, setting both preparedness and current alignment to their higher categories produced the highest average predicted probability, and this ordering does not establish interaction or a benefit from changing either characteristic. Current-role alignment was measured at the time of the survey and cannot be assumed to represent alignment when respondents first entered employment, so the association is interpreted descriptively. In the ordinal transition sensitivity model, preparedness and ability development retained positive estimates (OR 11.9 and OR 5.4), whereas alignment was more uncertain (OR 3.17, 95% CI 0.74–13.68); the alignment finding was therefore sensitive to outcome specification. In the supplementary multiple-comparison assessment of all 18 model coefficients, preparedness in Model 3 remained below the adjusted 0.05 threshold under both the within-model and pooled Benjamini-Hochberg assessments, whereas ability development and job alignment remained below that threshold within Model 3 but not in the pooled assessment.

4.6 Secondary analyses

Two secondary models appear in Table 7. For immediate employment, skill breadth was associated with entering work straight after graduation (OR 1.5 per additional skill, 95% CI 1.0–2.2); this association was robust to how skill breadth was counted, giving essentially the same estimate whether free-text additions were included (OR 1.53, 95% CI 1.04–2.23) or only the seven predefined skill categories were counted (OR 1.54, 95% CI 1.05–2.27). Ability development and the Carnegie Mellon indicator were of borderline significance in opposing directions, and the university-support estimate was below one but uncertain (OR 0.46, 95% CI 0.13–1.56). In the Firth job-alignment model, graduating before 2015 was associated with lower odds of currently holding a STEM-aligned role (OR 0.17, 95% CI 0.04–0.75), and selecting practical industry experience as a knowledge area valued in the current role — not a record of completing an internship — showed a positive but imprecisely estimated association (OR 4.7, 95% CI 0.9–24.3). Descriptively, 71% of pre-2015 graduates were in aligned roles against 96% and 93% of the later cohorts. Because this comparison involves sparse cells (three of six expected counts below five), it was assessed with the Fisher-Freeman-Halton exact test rather than the chi-square approximation (exact p = 0.039); the cohort difference remains significant at the 0.05 level, though the exact test is less extreme than the chi-square approximation would suggest. This pattern may reflect later career mobility into managerial or cross-functional roles, although cross-sectional data cannot distinguish progression from earlier mismatch.

Model/predictor	OR	95% CI	p
Model 4: Immediate employment
Ability development (greatly)	3.52	0.95–13.11	0.060
Skill breadth (per skill)	1.53	1.04–2.23	0.029
University support	0.46	0.13–1.56	0.211
CMUQ (vs. TAMUQ)	0.32	0.1–1.05	0.060
[n = 75, events = 46, AUC = 0.732, pseudo-R2 = 0.134]
Model 5: Job alignment (Firth)
Practical industry experience selected as valuable knowledge	4.72	0.92–24.28	0.063
Technical expertise	2.53	0.55–11.64	0.233
Graduated before 2015	0.17	0.04–0.75	0.019
[Firth penalised; n = 75, aligned = 65]

Secondary models 4–5: immediate employment and job alignment (odds ratios). Model 5 uses Firth penalised estimation

OR, odds ratio; CI, confidence interval; AUC, area under the receiver operating characteristic curve. Outcomes and reference categories are defined in Section 3.4. AUC describes apparent discrimination within the sample

Perceived employability did not differ significantly by gender or institution in ordinal comparisons (Mann–Whitney p = 0.81 and 0.38) or across cohorts (Kruskal–Wallis p = 0.61). The clearest evidence of a gender association in this sample concerned reported preparedness; the other comparisons do not establish equivalent outcomes. An exploratory covariate-adjusted mediation analysis of the ability-employability association is reported in the Supplementary Materials; its confidence interval included zero, so it provides no conclusive evidence of mediation and is not carried into the main findings. Because all measures are cross-sectional and self-reported, this calculation cannot establish temporal ordering or a mediation mechanism; it is retained in the Supplementary Materials for transparency only and is not part of the evidential argument.

4.7 Additional support identified by graduates

Of the 75 analytical respondents, 73 provided an interpretable answer to the open question. One blank response and one response consisting solely of a full stop were excluded from the content categorisation, while both respondents were retained in all quantitative analyses. The responses were categorised from the verbatim text during the revision using an explicit codebook and a documented respondent-level coding matrix available upon reasonable request, subject to institutional approval and the conditions of participant consent. Multiple categories could be assigned to a response, so the percentages do not sum to 100%. The most frequently identified need was internships and practical exposure (n = 30, 41.1%), followed by local labour-market and government knowledge (n = 11, 15.1%). Ten respondents (13.7%) indicated that no additional support was required. A further cluster concerned interpersonal and leadership development (n = 7, 9.6%), professional Arabic for dealing with local and government entities (n = 5, 6.8%) and workplace conduct and transition preparation (n = 5, 6.8%). Smaller numbers mentioned employer and industry engagement (n = 4, 5.5%), project and business management (n = 4, 5.5%) and other specified support, including access to other local universities, community engagement and research skills (n = 4, 5.5%), followed by career preparation and job-search support (n = 3, 4.1%), industry-specific technical skills (n = 2, 2.7%), professional networking (n = 2, 2.7%) and emerging technologies including AI (n = 1, 1.4%). One further eligible response expressed positive appraisal without specifying a support request and remained in the percentage denominator. These revised frequencies replace the open-question results reported in the submitted version, in which employer and industry engagement had appeared as a leading request; under the revised codebook it is no longer a leading category. Because one researcher categorised the responses, with artificial-intelligence assistance for the initial pass and no independent second coder, the frequencies are a descriptive summary; taken together they point toward practical, locally grounded preparation as the priority graduates identified.

5 Discussion

5.1 Educational evaluations and perceived preparedness are distinct measures

The results are consistent with a conversion reading of employability; this reading is an interpretation of associations among self-reported measures, not a demonstrated causal process. An excellent learning experience was associated with ability development and preparedness, yet none of the three educational variables dominated once modelled together, and it was preparedness rather than educational evaluation that carried the association with downstream outcomes. The pattern aligns with recent work distinguishing employability from work readiness and cautioning against conflating them (; ), and with graduate-capital accounts that separate acquiring capability from converting it (; ). Ability development behaved similarly: it was associated with preparedness, employability and transition quality, and retained an association with employability alongside preparedness, echoing evidence that applied capability and identity resources, rather than credentials alone, underpin employability and early careers (; ; ; ) and review findings on transversal skills and authentic, collaborative learning (; ; ). For an IBC population, positive educational evaluations did not invariably coincide with the highest reported preparedness category; what the data cannot establish is direction, so ability development is read as corresponding with, rather than producing, the outcomes examined.

5.2 Preparedness as a recurring correlate of outcomes

Preparedness was the most consistent correlate of both very high employability and a very smooth transition, with large predicted-probability contrasts. This fits accounts in which readiness concerns perceived preparation for workplace demands, rather than demonstrating that capability has been converted into performance (). The direction was consistent across the reported specifications, penalisation and campuses, but the small non-probability sample and the uncertainty of the estimates limit the inference, which remains an association rather than a demonstrated effect. Its practical reading, developed below, is that support addressing graduates’ reported preparedness warrants prospective evaluation.

5.3 Job alignment and transition quality

Job alignment corresponded with a very smooth transition and, combined with preparedness, with the highest predicted probability of one, while its association with perceived employability remained inconclusive and its transition estimate was imprecise. The alignment estimates differed across the two outcome models, but this study does not establish that alignment is associated with one outcome and unrelated to the other, in line with recent work treating occupation-curriculum alignment as a substantive outcome (). The cohort pattern, in which earlier graduates were less often in aligned roles, may reflect later career mobility, although the design cannot separate mobility from earlier mismatch.

5.4 Local conversion of international education

Internships and practical exposure were the most frequently coded support need, followed by local labour-market and government knowledge and leadership and interpersonal development; five respondents requested professional Arabic support. These categories describe requests in this sample and do not test whether the proposed support would improve employment outcomes. The requests echo review evidence on practical experience and employer collaboration (), curriculum-labour-market alignment research (), and Gulf findings on graduate preparedness and labour-market participation among nationals (; ). The reported language and local-knowledge requests are consistent with research emphasising socialisation and contextual resources in graduate employability (); the present study contributes descriptive evidence from citizens educated at locally situated branch campuses ().

5.5 Gender differences in reported preparedness

Women were less likely to report being very well prepared, while perceived employability and transition outcomes did not significantly differ by gender. The finding concerns self-reported preparedness at the point of entry and should not be read as weaker capability, as evidence of discrimination, or as a general national gender gap. It resonates with STEM research showing gendered differences in early confidence and preparedness that do not always correspond to differences in attainment or later outcomes (; ; ; ). The appropriate institutional response is equitable access to practical exposure, mentoring and career preparation, framed as equal opportunity rather than remediation.

6 Practical implications

Each implication follows from a reported association and is offered as a plausible priority, not a proven intervention. Because preparedness was the most consistent correlate of employability and transition, and internships and practical exposure formed the most frequently coded support category, structured and assessed internships and live employer projects are the clearest starting point (; ; ). Broadening applied skill development, with particular attention to project management and leadership, the least frequently reported skills, is motivated by the skill-breadth association, though this does not establish that expanding training would shorten time to employment.

Institutions could also track job alignment as a distinct indicator alongside employment rate, following graduates at one, three and five years to separate entry alignment from later career movement, and involve employers in periodic curriculum review to keep provision connected to labour-market demand (). Embedding local labour-market, regulatory and government awareness, and offering professional Arabic for technical and administrative settings, responds directly to what graduates asked for and aligns with national priorities (). Finally, because women were less likely to report feeling very well prepared while the perceived-employability comparison remained inconclusive, equitable access to practical exposure, mentoring and early employer contact is warranted; differences in programme composition, occupational exposure or interpretation of the response scale could contribute to this association, which the present study does not distinguish (; ).

7 Limitations and future research

Several limitations shape interpretation. The sample was modest (n = 75) and drawn from two institutions, so estimates are imprecise and cannot be generalised to all Qatari STEM bachelor’s graduates or presented as nationally representative. The cross-sectional design does not permit causal inference; the associations are consistent with the conversion framework but do not establish direction. Outcomes were self-reported and several were single items, which invites recall bias, common-method bias and ceiling effects, and some outcome categories were sparse, producing wide confidence intervals, particularly for job alignment. Recruitment through the researchers’ professional networks and referrals may have favoured graduates who were visible within those networks and willing to participate, and eligibility was screened against professional knowledge and profile information rather than official documentary authentication. Because recruitment focused on employed graduates, the study cannot estimate employment or unemployment rates, describe the experiences of graduates who remained unemployed, or establish an employment advantage associated with an international branch-campus qualification. Institutional comparisons may reflect programme and cohort composition rather than quality, and the open-ended responses were categorised by a single researcher and are therefore descriptive.

Future research should follow graduates longitudinally across the early career, draw on larger multi-institution samples, incorporate objective employment indicators and employer perspectives, and test the framework with validated multi-item measures of preparedness, alignment and employability ()

8 Conclusion

This study examined how internationally delivered STEM education relates to professional outcomes for Qatari bachelor’s graduates of two United States branch campuses in Qatar. Among the employed respondents surveyed, most reported roles related to their major and immediate entry into employment, alongside positive educational experiences.‏ yet the analysis indicates that labour-market preparedness, professional ability development and occupational alignment were the factors most closely associated with perceived employability and the quality of the transition to work. The observed gender association with reported preparedness, and a consistent graduate demand for internships and practical exposure and locally grounded knowledge, indicate where branch campuses might add value.

For institutions and policymakers, the message is measurement as much as provision. Monitoring labour-market preparedness and occupational alignment alongside employment rates may provide a more informative assessment of how international STEM education is converted into locally effective professional participation

Statements

Data availability statement

The datasets presented in this article are not readily available because of ethical and confidentiality considerations. Anonymized data may be made available by the corresponding author upon reasonable request, subject to obtaining the necessary ethical approvals. Requests to access the datasets should be directed to KA, k.alyafei@qfba.edu.qa

Ethics statement

The study received institutional approval for data collection from the Dean of Academic Affairs at Qatar Finance and Business Academy (QFBA), in academic partnership with Northumbria University, on 24 September 2024. The study was conducted in accordance with applicable local legislation and institutional requirements. Participation was voluntary, and electronic informed consent was obtained from all participants before they proceeded to the questionnaire.

Funding

The author(s) declared that financial support was not received for this work and/or its publication

Acknowledgments

The authors thank the graduates who participated in the survey

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. The authors used ChatGPT (OpenAI) and Claude (Anthropic) to assist with data visualisation, analysis code development and checking, categorisation of open-ended responses, and manuscript drafting and language editing. All use was supervised by the authors, who reviewed and verified the outputs and take full responsibility for the analyses, interpretations and final manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/feduc.2026.1993591/full#supplementary-material

References

Summary

Keywords

graduate employability, human capital conversion, international branch campuses, job alignment, Qatar, STEM graduates, work readiness

Citation

Alyafei KF and Issa MA-H (2026) Employability, work readiness and job alignment among Qatari STEM graduates of international branch campuses: evidence from a graduate survey. Front. Educ. 11:1993591. doi: 10.3389/feduc.2026.1993591

Received

09 September 2026

Revised

08 October 2026

Accepted

18 October 2026

Published

29 October 2026

Volume

11 – 2026

Edited by

Rose Murray, University of Bristol, United Kingdom

Reviewed by

Willson Gustiawan, Politeknik Negeri Padang, Indonesia

Steve Springer Laryea, Kumasi Polytechnic, Ghana

Updates

Copyright

© 2026 Alyafei and Issa

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher

---

Source: www.frontiersin.org