Selecting Alternative Dispute Resolution (ADR) methods in construction is often influenced more by contractual defaults and practitioner experience than by systematic evaluation of dispute and project characteristics, creating a gap between theoretically suitable and practical approaches. Building on prior identified structural, behavioral, and institutional drivers of this gap, this study examines how practitioner decision-making varies across dispute contexts and whether observable patterns reflect underlying structured preferences. A scenario-based survey of construction professionals in Kenya (n = 15) was conducted to elicit ranked ADR preferences across three contrasting dispute conditions: time-sensitive, technically complex, and relationship-sensitive. The study is based on ordinal ranking data and applies a non-parametric analytical approach. The results reveal clear context-dependent variation in ADR selection. Negotiation is preferred in time-sensitive disputes (mean = 5.47), expert determination in technically complex disputes (mean = 5.00), and negotiation (mean = 5.67) and mediation (mean = 4.60) in relationship-sensitive cases. These patterns suggest that practitioners implicitly weight decision criteria, including time efficiency, technical certainty, and relationship preservation. Spearman's rank correlation results indicate variation in preference structures across scenarios, reinforcing the context-dependent nature of ADR selection. This supports conceptualizing ADR selection as a multi-criteria decision problem and highlights the potential of the Multi-Criteria Decision Analysis (MCDA) as a preliminary framework for developing context-sensitive decision-support tools. However, the findings are limited by a small sample size, a simplified scenario structure, and the lack of direct measurement of decision-making. Despite these limitations, the study provides a foundation for further empirical validation and model development.