Page 726 - ISC PROCEEDINGS 21.4

collaboration, and the pace of curricular adaptation to digital-economy demands.
3.3. Layer 2: Student capabilities
The second layer captures the tripartite structure of graduate competencies:
Human capital (HC): domain-specific knowledge, theoretical mastery, and applied
problem-solving abilities, corresponding to traditional human capital investments.
Digital and AI-Related Skills (DS): data literacy, proficiency with digital tools,
computational thinking, and the capacity to work with AI-enabled systems are
competencies increasingly essential in the digital economy.
Behavioural and soft skills (BS): communication, teamwork, creative thinking, work
discipline, and a growth-oriented learning mindset, reflecting the ‘generic skills’
component of CareerEDGE and ‘employability skills identified in cross-national research.
It is important to note the conceptual distinction between these sub-
components: Professional Skills (X₂ in the pilot) captures generic transferable
competencies, including communication, analytical thinking and work effectiveness, that
apply broadly across occupations. Digital and AI-Related Skills (DS), by contrast, refer
specifically to technological proficiencies required in the digital economy, including data
literacy, use of AI-enabled tools and computational thinking. In the pilot instrument, DS
competencies are embedded within the Professional Skills construct due to sample and
instrument limitations; future iterations should measure DS as a separate latent construct.
3.4. Layer 3: Behavioural and perceptual layer
This layer distinguishes the framework from purely skills-based models, capturing
behavioural and psychological processes that mediate between capabilities and outcomes:
Self-efficacy and career confidence: The degree to which students believe they can
successfully apply their competencies.
Perceived employer demand: Students' understanding of what employers actually
require, which may be distorted by information frictions, cognitive biases, or social
influences.
Career aspirations and risk attitudes: Shaped by social norms, family expectations,
peer effects, and present bias.
3.5. Layer 4: Outcomes
The final layer encompasses both self-perceived employability, which is students’
subjective assessment of their labour-market competitiveness, and actual early labour-
market outcomes such as employment status, job–qualification match, and earnings. The
pilot data address only the self-perceived dimension.
3.6. Theoretical propositions
The framework generates four testable propositions:
P1: Graduate employability is a function of the joint interaction among domain
knowledge, digital/AI skills, and behavioural competencies.
P2: The behavioural–perceptual layer mediates the relationship between objective
capabilities and self-perceived employability.
P3: Behavioural biases, particularly present bias and overconfidence or
underconfidence, lead to systematic underinvestment in skills with high long-term returns
among students in emerging economies.
P4: Institutional factors moderate the capability–employability relationship, with
stronger institutional support amplifying returns to individual capabilities.

725