2026 ASEE Annual Conference & Exposition

This methods/theory full paper describes the validation of PERMA-V(N), a modified framework for understanding well-being in engineering doctoral students. Engineering doctoral students (EDS) experience disproportionately high rates of mental health challenges, with depression and anxiety rates more than six times higher than the general population [1], [2]. These well-being challenges are particularly acute during high-stakes academic milestones like qualifying examinations (QEs), which serve as gatekeeping mechanisms that determine advancement to doctoral candidacy [3]. Existing well-being measurement frameworks focus primarily on deficit-based mental health indicators such as depression and anxiety, failing to capture the full spectrum of positive and negative experiences that characterize doctoral student well-being during critical academic transitions. Positive psychology frameworks like PERMA-V, which assesses Positive emotions, Engagement, Relationships, Meaning, Accomplishments, and Vitality [4], cannot distinguish between emotionally-neutral and emotionally-negative aspects of EDS experiences. This limitation is particularly problematic during QEs, where, for example, students experiencing high engagement with preparation activities that may be reported as neutral with no negative connotations, or may be stress-inducing.

This study presents the validation of PERMA-V(N), a modified framework that extends the established PERMA-V model by adding a Negativity dimension to better capture EDS’ experiences. PERMA-V(N) additionally expands PERMA-V's ‘Positive emotions’ construct into ‘Positivity’ which has subconstructs of ‘Positive emotions’, ‘Positive affect’, and emotionally neutral ‘Positive statements of evaluation or assessment of circumstances’.

Nine second-year biomedical EDS from a mid-sized private R1 university participated in multiple semi-structured interviews before and after their QE. Using a constant comparative approach, researchers applied PERMA-V(N) codes (Positivity, Engagement, Relationships, Meaning, Accomplishments, Vitality, and Negativity) to interview transcripts. Framework validation employed a three-stage coding process with two independent coders, and third adjudicating senior researcher. Inter-rater reliability was assessed using Cohen's kappa across pre-review, post-review, and post-discussion agreement phases. This multi-stage process allowed examination of how different levels of coder collaboration affect reliability when coding complex emotional data.

The framework demonstrated good to excellent reliability across most dimensions, with Cohen's kappa values ranging from 0.536 to 0.988. Negativity codes (n=188) appeared nearly twice as frequently as Positivity codes (n=104), empirically validating the documented stress of qualifying examinations and demonstrating the framework's sensitivity to negative experiences that traditional PERMA-V coding would miss. Cohen's kappa values for Negativity ranged from κ=0.732 (good agreement) to κ=1.000 (perfect agreement) across validation stages, with all disagreements resolved to perfect agreement following structured discussion. The framework successfully captured co-occurring positive and negative experiences within single statements, such as simultaneous frustration with advisor mentoring style and appreciation for learning opportunities, and distinguished between emotionally-neutral engagement descriptions and stress-laden engagement experiences.

PERMA-V(N) provides an improved framework, able to distinguish between positive, negative, and emotionally-neutral dimensions while maintaining strong inter-rater reliability. This improved framework has application for studies that depend upon recognizing that well-being encompasses both flourishing and distress states that can co-occur rather than purely exist as opposite poles of a single continuum.