考慮集合在消費者選擇過程中扮演著簡化機制，已獲得學者的驗證支持。從資訊處理或學習理論均可解釋考慮集合形成的原因與過程，也有部分學者以經濟學的邊際分析方法，說明消費者對某一品牌是否納入考慮集合；藉由其預期邊際效用與搜尋成本和考慮成本的評估比較，如此可探討考慮集合的組成與其最適考慮品牌數目，以及其決定因素，本文即是以此一研究方向加以進行。 首先，本文以Hauser & Wernerfelt ( 1990 )一文為基礎，將Hauser & Wernerfelt概念化之決策成本，加以修正，使消費者考慮集合的成本結構更具體化。其次，倘若消費者選擇過程之考慮集合存在，也即並非單階段之選擇模式，則以一般之Logit或OLS方法估計會造成偏誤，進而影響預測選擇機率之結果。本文認為以選擇性偏誤之實證模式，可修正上述的估計偏誤，並可避免I.I.A的限制。最後，經由分析後本文提出七個命題，從廠商與消費者的立場，說明考慮集合的行銷涵義。

Numerous previous researches have supported that consideration set is the simplified mechanism of the consumer choice process. Both information processing and learning theory can explain the reason and process of the consideration set formation. Some scholars also use the marginal analysis of economics to explain if one brand should be included in the consumer's consideration set or not. Comparing the product expected marginal utilities with its search cost and consideration cost, we will explore the composition of consideration set and the determinants and the optimal number of brands of the consideration set. This is just the direction of this paper. First, based on the study of Hauser & Wernerfelt ( 1990 ). we refine the conceptual decision cost and enable the cost structure of consumer's consideration set to be more concrete. Second, if consideration set, not the single-phase choice model, exists in the consumer's choice process, then the traditional Logit or OLS model will results in the biased estimation and cause the incorrect choice probability. We convince the empirical model of selective bias can correct that estimation bias and avoid the restraint of I.I.A. Finally, we submit seven propositions, and explain the marketing implication of the consideration set in the view of point of firms and consumers.