當受試者看到一個立體物的平面式直角投射圖並詢以這些圖是否來自於同一個立 體物時，他們傾向以建構立體物的三度空間心像模型的方式來解答。在本研究中我們檢驗（ 1 ）在何種情況下比較有可能建構立體的心像模型； （ 2 ）所建構的心像模型的本質為何 ；及（ 3 ）男女在建構立體心像模型的能力上有無差異。 實驗中，受試者依序解答兩項作 業。首先是直角相容作業：在此作業中，受試者需要判斷在兩張連續畫面上出現的三張直角 投射平面圖，是否代表同一個立體物體。第二項作業為未預警之再認作業：在此作業中，受 試者被要求指認螢幕上所呈現的立體圖形是否在先前之相容作業中曾經出現。當受試者正確 解答相容作業且亦能在後續的再認作業中指出與相容作業中對應之立體圖形為曾經出現的物 體 -- 此項列聯機率被視為受試者利用平面訊息建構立體心像模型的證據。結果顯示，（ 1 ）對於低複雜度的物體，受試者較易以建構立體模型為解題策略；而對高複雜度的物體，受 試者則易傾向以分析特徵和比對的策略解題； （ 2 ）當所呈現的已知平面圖中包括物體的 正前方投射圖時，受試者較能建構立體模型，反之則否； 及（ 3 ）雖然在相容作業上男生 的表現優於女生，但是沒有證據顯示二者在建構立體心像模型的能力上有差異。這些結果有 助於澄清在何種情境中較易建構立體心像模型以及由平面訊息所建構之立體心像模型的特性 為何等問題。

　　　　　When subjects were presented with 2-D orthographic projections of a 3-D object and asked to judge whether or not the projected views were compatible with one another in the sense of representing an identical object, they tended to solve the problem by constructing a 3-D mental model of the object out of the 2-D displays. In the present study, we examined (a) the conditions under which 3-D mental models were more or less likely constructed; (b) the nature of constructed mental model, and (c) sex differences in the ability of 3-D model construction. Subjects were asked to perform two tasks in sequence. In the first, orthographic compatibility task, they were asked to judge whether or not the three orthographic views of a block object (made of small cubes) presented in two consecutive frames represented an identical object. In the second, incidental recognition task, subjects were asked, without warning, to determine if the orthographic projection of the isometric drawing of a 3-D object appeared earlier in the compatibility task. The contingent probability of correctly recognizing a 3-D object given its compability problem was also solved correctly was taken as the evidence that subjects solved the compatibility problems by constructing 3-D models from 2-D displays. Results show that (a) subjects were more likely to use a construction strategy when solving problems that were relatively simple and were more likely to adopt an analytic, feature-matching strategy when solving problems that were relatively complex, and (b) subjects were more successful at constructing 3-D models when the first two views provided in the compatibility task contained an orthographic projection of the front side of the object than when they did not, and (c) although males tended to outperform females in solving compatibility problems, no evidence for male superiority in constructing 3-D mental models was found. The implications of these findings were discussed in terms of the conditions under which 3-D mental models are likely to be constructed as well as the characteristics of a 3-D model once it is constructed.