大專教師選用教科書考量因素之研究－以行銷管理用書為例

本研究嘗試針對大專教師選用行銷教科書這個實際的消費者決策，運用行銷領域常見的方法與工具，來發掘出可以實際運用的行銷資訊。實際運用的方法/工具包括多重屬性決策模式、重要性–績效分析、以及根據屬性重要性進行集群分析所衍生的市場區隔等。經以文獻探討和深度訪談發掘重要的教科書屬性，再以網路問卷取得291位受訪者提供的資料，並執行適當的統計分析之後，主要研究發現包括下列幾點： 1.本研究所納入的17個屬性都獲得一定程度的重視，其中最受重視的5個屬性依序為內文流暢易讀、章節架構合理、收錄台灣本土案例、附有教學研討個案、和深度與廣度的調和。 2.多重屬性模式確實可以解釋使用意願，但其解釋能力並不穩定，而且極可能受到光環效應所干擾。 3.重要性–績效分析確實可以導出應該優先改善和應該繼續維持的屬性，但競爭者品牌(教科書版本)之間可能高度雷同。 4.可根據屬性重要性導出具有統計效度的兩個市場區隔，但兩個區隔在17個屬性的重視程度上呈現相同的差異，因此區隔結果很可能只是反映了受訪者的反應風格，而不是對各屬性的實際重視程度。上述研究發現可供業者實際應用在新教科書開發決策，而其他幾個發現則顯示有必要進行若干修正複現研究，藉以確認既有的方法/工具是否需要進一步調整。

關鍵字

多重屬性模式；光環效應；重要性－績效分析；集群分析；市場區隔

並列摘要

This research tries to employ some marketing methods/tools to investigate the applicable information in a real world context, i.e. the choice of marketing textbooks in universities/colleges. The methods/tools employed in this research include the multiattribute model of decision, the importance-performance analysis (IPA), and the attribute-based, cluster-wised market segmentation. A literature review and an in-depth interview were conducted first to explore the possible attributes of marketing textbook. Next, a web survey was conducted and 291 usable questionnaires were collected. The analytic results of these questionnaires suggest that: 1.The 17 textbook attributes in this research are perceived to be somewhat important at least. The 5 most important attributes are fluency of texts, reasonable chapters, localized cases, teaching cases, and appropriate compromise between depth and width. 2.Although the multiattribute models are successful in explaining the choice intentions, the explaining powers are unstable and possible to be contaminated by the halo effect. 3.Although the importance-performance analyses are successful in classifying the attributes into improvement and maintenance sectors, the classification results of competing brand (i.e. various textbooks) are possible to be highly similar. 4.Although the clustering procedure successfully reaches a two segments solution, the differences among the 17 attribute weights between these two segments are homogeneous. This result suggests that the segment solution is highly possible to be a reflection of response styles, instead of the perceived attribute weights. The first finding is directly applicable in textbook development decision. The other 3 findings indicate that some modified replication researches are needed to explore whether the modification of existing methods/tools are necessary or not.