Page 729 - ISC PROCEEDINGS 21.4

Table 5. Simple regressions of employer-demand alignment on individual predictors
Predictor β t R²
Knowledge alone 0.856 29.93 0.902
Skills alone 0.873 32.25 0.915
Attitude alone 1.265 17.48 0.759
Source: Author
The finding that Attitude has the lowest individual explanatory power yet remains
significant in the multiple regression is consistent with the framework's proposition that
attitudinal and behavioural dimensions contribute independently beyond knowledge and
skills.
The high inter-predictor correlations (r = 0.891–0.961) indicate substantial
multicollinearity in the OLS model. Variance Inflation Factors (VIF) confirm this: VIF(X₁) =
14.41, VIF(X₂) = 16.35, and VIF(X₃) = 9.25. The values for Knowledge and Skills exceed the
commonly cited threshold of 10, indicating that their individual coefficients should be
interpreted with caution, as standard errors are inflated and it is difficult to isolate the
unique contribution of each predictor. Nonetheless, the overall model fit (R² = 0.932)
remains valid and the joint significance of the three predictors is not affected by
multicollinearity. This limitation reinforces the call for future studies to employ PLS-SEM
or confirmatory factor analysis to properly separate the latent constructs and test the
mediating mechanisms proposed in P2 and P3.
4.6. Gender differences
Exploratory t-tests revealed no statistically significant gender differences in any
construct (p > 0.10 for all). Female students scored slightly (but not significantly) higher
on all four dimensions.
5. Discussion
5.1. Alignment with the Framework
The pilot results provide preliminary support for several framework propositions.
The finding that knowledge, skills, and attitude jointly predict self-perceived employer-
demand alignment (P1) corroborates the multidimensional nature of employability. The
lower mean and coefficient for Attitude is instructive. In line with the behavioural layer of
the framework (P2), this suggests that attitudinal and psychological factors, while
contributing to employability, may be the dimension where students perceive the
greatest gap. This aligns with Savickas's (1997) career adaptability concept and Dacre Pool
and Sewell's (2007) emphasis on self-efficacy as a higher-order employability component.
The very high inter-construct correlations support the framework's layered,
interactive structure: students who invest in knowledge acquisition also tend to develop
stronger professional skills and more positive attitudes, creating a mutually reinforcing
cycle. This synergistic pattern over a purely additive model reinforces the need for holistic
rather than piecemeal approaches to employability development.
5.2. Behavioural economics implications
The wide dispersion of scores (approximately 31% of students scoring below 2.5 on
Knowledge and Skills) suggests substantial heterogeneity in students' human capital
investments, potentially reflecting present bias, which is the tendency to discount future
benefits of skills development. The moderate mean scores (M ≈ 3.0) suggest neither
widespread overconfidence nor extreme underconfidence, which is encouraging from a
behavioural perspective.
However, the pilot design does not allow direct testing of P3 and P4 concerning

728