臺灣博碩士論文加值系統

目前系統分析方法論可分成兩大類:結構化分析(Structured
Analysis)及物件導向分析(Object-Oriented Analysis)。兩類方法論各
有其優缺點﹐亦有些研究提出整合性方法論的模式﹐但對於比較兩種方法
論的實證研究還是很少。因此﹐本研究以實驗室實驗法(Laboratory
Experiment)比較結構化分析與物件導向分析兩種方法，主要的研究目的
在探討(1)何者具有較高的學習績效，以模式化過程中的一致性及最後結
果的語法完整性及正確性作為衡量標準﹔(2)何者具有較高的系統需求表
達效能，以模式化過程的語意一致性及最後結果的語意正確性與完整性作
為衡量標準。實驗是以資訊管理學系二年級學生為實驗對象﹐隨機分成兩
組，一組學習結構化分析方法論﹔另一組學習物件導向分析方法論﹐經過
十週訓練之後，兩週時間讓學生實際分析電話選課系統，然後請二位熟知
系統分析方法方面的專家及一位熟知業務領域的專家進行評分工作。
研究結果顯示在學習績效上﹐結構化分析比較容易達到語法完整性﹐至於
兩種系統分析方法的學習難易程度﹐並沒有顯著差異。在需求表達效能上
﹐結構化分析較優於物件導向方法﹐尤其在模式的語意完整性構面上。本
研究結果在學術上可以提供系統分俘與設計課程參考﹐對初學者而言﹐結
構化分析語法比較容易學得完整﹔而在實務上可以提供從事系統分析工作
者在選擇分析工具之參考﹐結構化分析工具還是比較成熟及完備﹐而且比
較容易表達完整的語意。

There are two major classification of system analysis:
structured analysisand objected-oriented analysis. Each
methodology has its advantages and disadvantages.
Integrated methodology model was suggested in some studies,
however, there are few empirical studies about comparison with
the two methodologies. Therefore, our study use Laboratory
Experiment to make comparison of two system analysis
methodology. The two major purposes of this study are : (1)
which is better on learning performance ? (2) which is
better on system requirement representation effectiveness?
Result of this experiment indicate that structured analysis is
easier to achieve syntactical consistent on the learning
performance, and there''s no difference between two
methodology. On the requirement representation
performance, structured analysis is better than object-oriented
analysis, especially on the semantic completeness of
modeling. This research result could provide principles of
course designing of system analysis and system design. For
novice, structured analysis is easy to learn completely. In
addition, it could provide suggestion for analyst in practice
when selecting system analysis tool. For tool selection , the
structured analysis is still better maturely and completely,
and is easier to present semantic completely.