In engineering design, designers make sequential information acquisition decisions such as selecting designs for performance evaluation, selecting information sources, and deciding when to stop design exploration. While there is significant literature on normative models for making these decisions, there is a lack of knowledge about how human designers make these decisions and which strategies they use. This knowledge is important to identify the sources of inefficiencies for improving the design process. To address this gap, the primary research objective of this study is to identify models that provide the best description of a designer's information acquisition decisions when multiple information sources are present and the total budget is limited. We conduct a controlled experiment to gather empirical observations of subjects' decisions, and conditional on the experimental data, we perform Bayesian model comparison on various rational (normative) and heuristic decision models. Th control variables in the experiment are the amount of fixed budget and the monetary incentive proportional to the saved budget. The results indicate that the subjects' decisions are closer to the heuristic models than the rational models in general, and for some types of decisions, the fixed budget and the type of incentive affect their decisions. For example, the subjects employ simple heuristics at low fixed budget but rational judgments at larger fixed budgets for choosing between uncertain information sources. The prediction accuracy of a rational expected improvement-based model representing the decision to stop increases when there is an incentive to save the budget.