本研究以80位四年級兒童為對象，探討討論與提供實物的真實情境，對兒童解決生活數學問題與類比遷移解題的影響，以及分析兒童討論解題的策略與錯誤類型。研究結果發現，同時提供討論與實物的真實情境，兒童解決問題的表現最好。兒童在單獨提供實物或討論，或兩者並具的情境，於解題後經過三週的時間，類比遷移解題表現優於無實物無討論情境組的兒童。在解題策略分析上，兒童多數使用乘、除以及混合加、減運算，或等量連加策略，而且使用一一點數或跳數策略仍可見。在好與差解題者的類比遷移解題與錯誤類型分析上，好解題者雖然解題表現優於差解題者，而且解題錯誤次數低於差解題者，但是仍有購買過多或不足的錯誤。未能買得剛剛好與提取不相關的訊息解題，是兒童解題錯誤比例較高的類型。針對研究結果與發現，本文並提出相關的教學與研究建議。

Eighty fourth-grade children worked in mixed-ability dyads and were randomly assigned to four experimental conditions: with and without real-world settings and verbal interaction with peers. Dyads of children were asked to solve two daily mathematical problems in the first phase. Three weeks later, the children solved three daily mathematical problems as delayed transfer tasks. Dialogues between pairs of children during problem-solving were analyzed. Results show that verbal interaction accompanied by real-world settings is the most effective way to improve children’s performance in solving problems. Either verbal interaction or real-world setting helps children’s analogical transfer compared with solving problems individually without any real-world settings. Moreover, children tend to mix addition, multiplication and counting to solve problems through verbal interaction with peers in the performance of tasks. Accuracy at picking the best buys as well as selecting essential information from problem contexts were the major difficulties in children’s problem-solving. In the conclusion of this study, the author offered some suggestions in teaching and research of related topics.