<![CDATA[Web application performance has emerged as a critical determinant of digital business success, where load time reductions demonstrably correlate with user conversion rates. Nevertheless, the prevalence of JavaScript frameworks in contemporary development landscapes introduces a fundamental trade-off between developmental convenience and inherent performance overhead. This challenge is compounded by the divergent architectural paradigms adopted by frameworks, distinguishing between Virtual DOM (VDOM) and compile-time approaches. While optimization techniques exist, the question remains whether their efficacy is universal or moderated by framework architectural characteristics. This study analyzes the interaction effects between these two factors on user-centric performance metrics through a two-way factorial experimental design, developing two identical news portal websites using VDOM and compile-time frameworks. Five optimization strategies were implemented sequentially as experimental scenarios. Statistical analysis employed Aligned Rank Transform (ART) ANOVA and ART-C post-hoc testing on FCP, LCP, and TBT metrics. Findings indicate a consistent superiority of compile-time architecture over VDOM in FCP (p-value < 0,05), absent statistically significant optimization strategy influence (p-value > 0,05). Conversely, significant interaction effects emerged within LCP and TBT metrics (p-value < 0,05). The research concludes that optimization strategy effectiveness regarding LCP and TBT is contextual, contingent upon the architectural characteristics of the framework employed]]>
Copyrights © 2026
