Emotions come in various ways, such as subjective feelings, physiological responses, and facial expressions. Whether these components exhibit synchronised coordination during emotional episodes—referred to as emotional response coherence—has been central to a century-long debate about the nature of emotions and their underlying mechanisms. Despite agreement that emotions involve multiple interconnected components, hardly any studies have adopted a structural perspective to capture these interrelationships. This study examined coherence between subjective emotional experiences and physiological arousal, measured via skin conductance (SC), within an achievement context using a psychometric network approach. A total of 159 participants completed a numeric reasoning test, in which the item difficulty levels were manipulated to induce varying emotional responses. During the test, self-reports of discrete emotions (joy, pride, anger, anxiety, frustration, and boredom), valence, and arousal were collected alongside continuously recorded SC signals. Multi-level graphical vector auto-regressive (GVAR) models were employed to analyse multi-componential relationships at the intra-individual level. Findings revealed a multifaceted relationship between emotional experience and physiology, rather than a straightforward one-to-one correspondence. Significant associations were found between valence and SC measures, particularly SC response peak frequency and tonic level, while subjective arousal showed no significant links to SC measures. These results suggest that emotional response patterns may not be uniquely tied to specific emotions and may vary depending on context and measurement methods. As one of the first studies to examine emotional response coherence using a network approach, this research offers novel insights that may help reconcile long-standing theoretical divergences in emotion research.