本研究以認知負荷的觀點，探討不同教學順序之工作範例對「學習成效」與「認知負荷」的影響，其中教學順序方式包括「先看範例再練習（W-P）」及「先自己解題再看範例（P-W）」兩種。本研究採實驗研究法，研究的教材內容為國小幾何教材中之面積覆蓋。教材依認知負荷理論的教學設計原則設計，特別是工作範例原則。研究對象為國小四年級104名學生，將其隨機分配為「先看範例再練習組」與「先自己解題再看範例組」。研究工具包含兩節數位教材及認知負荷量表。實驗過程是先進行前測，再將學生分兩組學習兩節的數位教材課程，一週後進行後測，再隔一週進行延後測，在學習活動後及測驗階段後均進行認知負荷的問卷調查。研究結果發現如下：一、後測及延後測時，P-W組的學習成效顯著優於W-P組。二、在認知負荷之花費心力面向，P-W組在數位教材第一節及延後測時與W-P組有顯著差異，且P-W組花費心力多於W-P組。三、在面積覆蓋題型中，被覆蓋物是方形的通過率高於三角形；覆蓋物以4平方公分正方形通過率最高，0.5平方公分的等腰直角三角形最困難。

The purpose of this study was to design instructional material based on the cognitive load theory, and to investigate the consequences of learning achievement and cognitive load caused by different instructional orders. The instructional orders used in this study included the “worked example-problem solving (W-P)” and “problem solving-worked example (P-W)” orders. This study adopted an experimental design. The content was the “area covering” in the elementary school geometry. The teaching material was designed by applying the instructional principles of cognitive load theory, especially the principles of worked-examples. The participants were 104 fourth graders of an elementary school. They were randomly assigned to either the “worked example-problem solving” (W-P) or the "problem solving-worked example” (P-W) group. The research instruments included two multimedia lessons and a cognitive load questionnaire. A pretest was taken prior to dividing the participants into groups. Then two multimedia lessons were given, followed by a post-test one week later, and a delay post-test after another week. The cognitive load questionnaire was given after each learning lesson and test. The results were as follows: 1. The learning achievement of P-W group was much better than W-P group in both the post-test and delay post-test. 2. The P-W, however, perceived higher level of mental effort than the W-P and reached a significant difference in the first lesson and the delay post-test. 3. In the area covering problems, the students did better in squares than in triangles; they did best in 4 cm2 squares and worst in 0.5 cm2 triangles.