螞蟻演算法應用於測驗選題之研究

早期電腦尚未普及的年代，命題者採人工出題的方式，依其偏好及想法來完成測驗卷的命題，在整體測驗的相關事宜上，從出題、施測、成績計算與統計，都相當的耗時費力，尤其是全國的大考，必須考慮到測驗的公正與公平性，往往造成試務單位很大的負擔與壓力。隨著電腦科技的快速發展，在命題與施測的流程上，經由電腦化處理，一切變得非常的容易簡單，效率提高，成本降低。現在大部分的教師或是測驗中心早已開始使用電腦資料庫來管理測驗試題，並隨時可從資料庫中挑選出需要的試題製成測驗卷。測驗一直是評估學習者學習成效的重要一環，透過良善的測驗選題機制所製作出的試卷，不僅可以協助施測者評估學習者的學習狀態，而且有利於診斷出學習者在學習過程中所發生的問題，因此，如何選取適當的試題以產生符合需求的試卷，是一個非常重要的課題。而測驗的目的與受測者樣本分布的不同，在試題的組合上會有不同的需求。以「試題反應理論」為主要架構，加上施測者對不同施測目的的需要，已有不少學者投入測驗選題方式的研究中，所以施測者如何由題庫中針對預期的測驗目的與對象，結合高效能的選題方式以組成一份有效需求的試卷是本論文主要的研究方向。本研究以「螞蟻演算法」結合「試題反應理論」進行測驗選題評估，結果顯示運用「螞蟻演算法」在有限制條件的測驗選題上是一個可行的方法。以「螞蟻演算法」所計算出的測驗誤差遠低於傳統的「隨機法」，對於測驗選題方面的研究，更提供了另一項較為精確的方法，也擴展了「螞蟻演算法」可以運用的範圍，對於日後在測驗選題的領域上，「螞蟻演算法」可提供一個更有效率的演算技術，使得測驗品質更能符合施測者的施測需求。

關鍵字

試題反應理論；測驗選題；螞蟻演算法

並列摘要

Testing has always been an essential link of evaluating learners’ learning effectiveness. A test based on the well-designed testing mechanism can not only assist the evaluator to assess the learners’ learning status, but also be helpful in diagnosing the problems the learners encounter during the learning process. Therefore, choosing an appropriate test to meet the needs has become a very important issue. The aim of the test and the distribution of samples affect the combination of testing questions. Item Response Theory being used as the main structure and the evaluators’ needs according to different goals of evaluation, quite some scholars have been researching the choice of testing questions. The main direction of this thesis is how evaluators choose from data base with the aim and the result shows that Ant Algorithm is eligible in tests of limited conditions. The error calculated from Ant Algorithm is far less than the traditional Random method. A more accurate mathematical calculations is provided for the research of choosing testing questions, and the application of Ant Algorithm has also been broadened. Ant Algorithm has provided a more efficient mathematical calculation technique in the field of choosing testing questions, thus, the quality of testing will better meet the demand of evaluators.

並列關鍵字

Select Test Items ； Item Response Theory ； Ant Algorithm

參考文獻

[19] Dorigo, M., Maniezzo, V. & Colorni, A. (1996). “ The Ant System: Optimization by a colony of cooperating agents,” IEEE Transactions on Systems, Man, and Cybernetics – Part B, Vol. 26, No. 1:pp.1-41.

[21] Dorigo, M. & L.M. Gambardella (1997), “Ant colonies for the traveling salesman problem,” BioSystems, Vol. 43, pp.73-81

[22] Dorigo, M., Caro, G. D., Gambardella, L. M., (1999), “Antalgorithms for distributed discrete optimization,” Artificial Life,Vol.5(2), pp.137–172.

[24] Dorigo, M., Bonabeau, E., theraulaz, G., (2000), “Ant algorithms andstigmergy,” Future Generation computer Systems 16 , pp.889-914.

[29] Hambelton, R.K., Swaminathan, H., & Rogers, H. J. (1991). “Fundamentals of Item Respone Theory,” California, NP: Sage.

國際替代計量

螞蟻演算法應用於測驗選題之研究

全文下載

主題瀏覽