使用資訊科技學習數學：以網路同儕互評為例

Abstract

網路同儕互評是以資訊科技輔助學生學習的方式之一種，其理論基礎包含社會建構、社會學習及認知衝突情境等，本研究旨在分析不同性別與高低數學程度學生進行網路同儕互評後之數學學習成就、數學態度、與互評態度。由於研究對象為國中常態班學生，學習科目為數學，本研究亦肩負檢驗互評系統是否適用於常態班，國中學生是否適合進行同儕互評，數學科如何進行同儕互評，在網路上實施有何考量等多重目的。 以中部某國中二年級60名學生為研究對象，選定一次函數圖形為教學單元。研究採準實驗研究法有前後測設計，實驗處理為資訊科技輔助數學教學及網路同儕互評。教師首先對全體學生施以資訊科技輔助數學教學，並要求學生在一網路同儕互評系統上繳交作業並進行互評。學生群以性別及先前數學程度分為四組：高程度男組、高程度女組、低程度男組、低程度女組。每位同學評論四份(分別來自四組)作業，每份作業也被四人(分屬四組)評論。教學前先施以第一次數學成就測驗和數學態度問卷，再進行教學與互評活動，第一次作業繳交完，實施第二次成就測驗和數學態度問卷。進行兩輪互評後，實施第三次成就測驗、數學態度和互評態度問卷。 本研究發現如下：國中生對數學作業進行互評，其評分與教師評量有一致性。男女生數學成就得分並無不同，成就測驗的施測次數與數學程度間有交互作用，高程度組的數學成就在三次施測中越來越高，低程度組的進展比較受到侷限。程度高者的自評、互評、及教師評分均高於程度低者，性別間並無差異，所有組的第二次作業評分比第一次明顯進步。數學態度與第二、三次數學成就間均有顯著正相關。學生對互評的三種活動：觀摩、評論及回饋進行獲益度評估，所有學生都有中上的獲益度評估，不同性別與數學程度組間無顯著差異。而且多數同學支持本網路同儕互評活動，高達72%學生肯定本互評教學活動，願意參加下一次活動。 最後本研究提出常態班版本的互評系統規劃、同儕互評單元選擇考量、互評作業題目設計原則、數學科網路同儕互評作業的設計原則、對不同程度者的評論行為進行分析，也為實際施行網路同儕互評教學活動提出建言。

Networked peer assessment is a way to use computer technology in enhancing students’ learning. Its implementation is based on several theories, such as social construction of knowledge, social cognitive learning and cognitive conflict. This research was to investigate whether junior high school students, with different gender and previous math levels, reacted differently in a Web-based peer assessment while learning mathematics. The author also hopes to know whether peer assessment facilitated through Internet is adequate for junior high students in ordinary classes (ability grouping was not in use, so individual difference seems huge) when subject matter was mathematics. Sixty second-graders at a junior high school in central Taiwan served as subjects. They were asked to learn linear function. A quasi-experiment design was adopted with math instruction assisted with computer technology and networked peer assessment as the treatment. Students were assigned to four groups: boy with high previous math learning, girl with high previous math learning, boy with low previous math learning, girl with low previous math learning. Each student had to review four assignments, each selected from one of the four groups. Any assignment was also reviewed by four students, each from one of the groups. In pretest, students had to answer a math achievement test and a math learning attitude scale. Immediately after instruction and the completion of the 1st round assignment, they had to fill in, again, a math achievement test and the math learning attitude scale. Finally after two-rounds of peer assessment, a posttest including math achievement, math learning attitude, and perception about networked peer assessment was measured. The result showed that, first, student’s rating about peers’ assignments was highly correlated with those of teachers’. Second, about performances in math achievement tests, though gender difference was not significant, the 2-way interaction of test-times (pre-, middle, and post tests) and previous math learning levels (high and low) was significant. In other words, previously high math achievers performed better and better in pre-, middle, and post tests, while low achievers gained in the beginning but then lost momentum. Regarding rating of assignment, gender caused no difference, either did any interaction. However, previous high math achievers did much better in self, peer, and teacher ratings than the low achievers, and the second round ratings were always higher than the first round ratings. Math learning attitude, total score and all subscales, were correlated with math achievement in second and third times. When students subjectively judged gains from various activities of peer assessment, e.g., modeling, review, and reading feedback for further improvement, they responded positively no matter their previous levels of math learning or gender. Furthermore, students replied that the whole instructional activities were interesting and challenging, and about 72% of subjects would like to participate again, if the next lesson will be taught in the same format. Finally, this study offers several implementation suggestions: principles of designing a networked peer assessment system suitable for junior high students with no ability grouping the selection consideration of math materials, principles of designing peer assessment tasks – especially math tasks, and reflection about the significance of high and low achievers participate in peer assessment.