本研究主要目的是在發展一套適用於國小的數學乘法系列診斷測驗題庫，並希望研究者設計的「解題歷程、題目類型雙向度建構模式」，能有效的建構本測驗；另外，為了要使測驗的內容能隨不同的需要而有不同的組合，並提供多種測驗組合以方便教師使用；以垂直等化連結各式測驗，來擴充題庫內容；最後將測驗相關內容放在網際網路上，以加速測驗使用的流通，並透過分享來獲得使用者的意見。
本研究先根據文獻探討的結果，以Polya（1945）提出的四段解題歷程為縱座標，並以Bell（1989）依題目性質將乘法應用題分成八類為橫座標，綜合為「解題歷程、題目類型雙向度建構表」，並據此架構編擬五式「國小數學乘法系列診斷測驗」。
玆將研究結果陳述如下：
一、從試題反應模式適合度考驗結果來看，顯示組成題庫的各式測驗的反應資料，大致符合二參數IRT模式。全測驗之鑑別力（a）的平均值為.5432，難度（b）之平均值為 -.5864，全測驗的x2值為791.2(df=761, P=.2172)。
二、學生數學能力值（Θ）在性別與學校有交互作用，男生在數學乘法系列診斷測驗學生能力值顯著高於女生的數學能力值。
三、在試題鑑別度上歷程與類型的交互作用未達顯著水準，不同歷程的試題鑑別度有顯著差異，事後考驗顯示瞭解問題的試題鑑別度顯著高於擬定策略的試題鑑別度。不同類型的試題鑑別度有顯著差異，事後考驗顯示各類型間的鑑別度均未有顯著差異。在試題難度上歷程與類型的交互作用未達顯著之水準，不同歷程的試題難度有顯著差異，事後考驗顯示擬定策略的試題難度顯著高於回顧、執行、瞭解問題的試題難度。不同類型的試題難度有顯著的差異，事後考驗顯示各類型間的試題難度均未有顯著差異。
四、學校甲、乙、丙的學生和數學能力值與月考成績均得到中度以上相關，顯示國小數學乘法系列診斷測驗對數學科情形有相當程度的預測力。
五、本研究的編製之數學乘法系列診斷測驗的解題歷程部份之因素分析模式的驗證，結果x2為14855.27，df為12238，達.01之統計顯著水準，無法支持數學乘法系列診斷測驗中四組解題歷程的想法；惟輻合效度尚稱理想，但區別效度不甚理想。另外，數學乘法系列診斷測驗的題目類型部份之因素分析模式的驗證，結果x2為15022.83，df為12374，達.01之統計顯著水準，無法支持數學乘法系列診斷測驗中八種類型題目的想法；惟輻合效度尚稱理想，但區別效度不甚理想。
六、網路版「國小數學乘法系列診斷測驗」的首頁網址在：（http://epnet2.ecpsy.ntnu.edu.tw/~mathdiag），其內容共分七部份：(一)線上評量；(二)試題內容；(三)線上問卷；(四)理論介紹；(五)使用說明；(六)公共論壇；(七)意見欄。
最後研究者根據上述的研究結果，提出幾項建議，供未來研究之參考。

The purposes of this study were to develop an item bank on mathematical diagnosis for elementart school, to validate the item bank by using "solve process by item type framework", and to make it on internet by WWW method.
The items included in the item bank were divided into five forms. 16 commin items were used in two consecutive forms for linkage purpose. The data collected from target samples were analyzed by vertical equating techiniques using BIL OG-MG. The confirmatory factor analysis model were used to validate the content of item bank.
The major findings of this study were as follow:
1.For the IRT modeling, from the goodness of fit index, the response data were mostly fit in two parameters IRT response model.
2.For the difference in mathmatical ability score (Θ), the interaction effect between sex and school reaches significance level. Male's score were significantly higher than female's.
3.Regarding the items' discrimation power(a), the interaction effect between solve proces and item type wasn' t significant. There were significant main effect for both solve process in a paramenter. From the plst hoc, The a parameters of understanding to problems were higher than the items of devising plan.
4.For the items' difficulty level(b), the interaction effect between solve process and item type wasn' t significant as a parameter. There were significant main effect for solve process, in the mean time, the results from post hoc showed that the items of divising plan are harder than the items of understanding to problems, carrying out plan and looling back.
5.Mathematic ability scores are significantly plsitive correlated with the schools mathematics achievement examination.
6.The response data wasn' t fitted well to 'sove process by item type framework' model. However, the result indicated acceptable convergent validity for both solve process and item type model.
7.A computerized version of the item bank was made and the homepage was located in httpd://epnet2.ecpsy.ntnu.edu.tw/~mathdiag.
Baes on the results of this study, the researcher made suggestions in regaed to the research and future studies on the subject.