過去探討學科補習的研究，其關注焦點主要著重在學科補習階層化與其效益這兩方面。但這些研究在探討階層化與效益時，存在著兩大缺口，一是探討學科補習階層化的研究並沒有考量家庭社經地位對學科補習參與影響中可能存在的中介變項，二是探討學科補習效益的研究存在著高估學科補習效益的可能性。本文在檢討過去探討學科補習階層化與效益的研究後，參考Wisconsin 模型，運用「臺灣教育長期追蹤資料庫」資料，以更適當的研究架構，對上述問題重新探討。研究結果顯示：一、家庭社經地位愈高，父母教育期望與子女自我教育期望愈高，進而提高學科補習參與；二、控制父母教育期望與子女自我教育期望後，學科補習參與對於學業成績先升後降的非直線影響下降許多。由於過去探討學科補習效益的研究很可能高估學科補習的效益，本研究的發現有助於打破學科補習高效益的迷思。

Previous studies of cram schooling have focused primarily on stratification and efficacy. However， they have fallen short of the depth of their discussions. Concerning stratification， previous studies have not considered possible intervening variables caused by socioeconomic status of family that might influence children's attendance at cram schools. Moreover， these studies have perhaps overestimated the efficacy of cram schools. This research focused on two unresolved issues regarding the stratification and efficacy of cram schools， taking the Wisconsin model into account and using data collected from the Taiwan Education Panel Survey， to develop a well-defined research framework to reinvestigate the research questions above. It was found that higher socioeconomic status of a family led to higher educational expectations of parents and children. This in turn led to an increase in children's attendance at cram schools. The non-linear influence， first ascending and then descending， of student attendance at cram schools on academic achievement， decreased greatly after the variables of educational expectations of parents and children were held constant. Because previous studies probably overestimated the efficacy of cram schooling， this study may help dispel the myths surrounding the efficacy of cram schools.