Research
One question, asked in three settings: why does commitment so often fail to convert into execution, and what conditions close that gap?
Across strands I look for asymmetry — cases where the factors that produce a stated intention are not the factors that produce the corresponding behaviour. Methodologically this pushes toward designs that can represent asymmetry directly: configurational analysis alongside regression, moderated mediation alongside main effects, latent profiles alongside variable-centred models.
Careers and human capital
The largest strand. How do researchers and technical workers convert learning opportunity, organizational support, and psychological resources into employability and innovative behaviour — and where does that conversion break down?
Recent work examines how career development support transforms personal, social, and organizational resources into research passion among STEM researchers; how mid-career scientists sustain employability under resource threat; and how psychological safety shapes retraining intentions during digital transformation.
Methods: PLS-SEM, multi-group analysis, moderated mediation, latent profile and K-means clustering
Scale-up and venture finance
Why do ventures that clear early validation stall at scale, and what do investors actually weigh when the technology is deep and the horizon is long?
Work in this strand covers corporate venture capital decision criteria for deep-tech startups, the relationship between financing difficulty and growth aspiration among turnover-scaleup firms, and regional disparities in technology innovation capability.
Methods: fsQCA, importance–performance analysis, propensity score matching, network analysis
Green and digital transition
What makes climate and energy commitments credible, and when does digital transformation help or hinder environmental outcomes?
Current work examines whether digital transformation impedes reuse innovation with environmental benefits, how venture firms’ social value orientation relates to ESG management and performance, and carbon disclosure quality across differing regulatory regimes.
Methods: two-way fixed effects, fsQCA, vignette experiments
A note on method
Unless a design supports causal identification, findings are reported in associational terms. Sample restrictions, excluded observations, and construct decisions are documented explicitly rather than left to the reader to infer.