本研究目的旨在修訂學童復原力量表，並測試其個人背景、自覺健康狀況、運動習慣、自覺學業成績對復原力得分的影響，期望未來能提供學童復原力教育介入之參考。本研究方法分成二部分進行：（一）量表修訂：首先以Grotberg復原力模式為理論基礎，參考現有之復原力量表及相關文獻擬定量表初稿，並進行專家審查。接著，以台中市某國小四、五、六年級共100人為預試樣本，進行項目分析及再測信度。（二）量表測試：以分層隨機抽樣方式選取台中市七個區九所公立國小四、五、六年級共865人為研究樣本，採用內在一致性信度、因素分析、描述性統計、積差相關分析、t檢定、單因子變異數分析及多元迴歸分析等統計方法，進行資料分析。
根據統計分析結果，本研究獲致以下的結論：

一、量表修訂：本量表總題數為67題、內含三個分量表（外在資源、內外控、社會技巧）。信度部分：內在一致性Cronbach’s α值介於0.69~0.92之間，再測信度γ值介於0.56~0.84之間，達p<0.01之顯著水準。顯示本量表具有良好的信度。效度部分：每個分量表經因素分析後可分成4至5個因素，與量表編製架構大致相符。內在一致性效度方面呈現分量表與總量表間有中度至高度的相關（γ=0.35~0.90），均達p<0.01之顯著水準。顯示本量表有不錯的效度。
二、量表測試：（一）復原力量表得分：四年級學童在復原力總量表、外在資源量表、社會技巧量表的平均得分均高於五年級及六年級，唯內外控量表方面呈現六年級學童平均得分高於五年級及四年級；另發現女性學童無論在總量表或分量表部分得分皆優於男性（二）與國小學童復原力量表總分呈顯著差異的變項有：性別、年齡、父母親工作現況、自覺健康狀況、運動型態與自覺學業成績等六個變項。而父母婚姻狀況及父母教育程度則未達顯著差異。將以上顯著變項投入多元迴歸分析，結果亦呈顯著差異（F= 16.35, p<0.001），約可解釋24％的學童復原力。

OBJECTIVE-The purpose of this study is to reedit the resilience scale and to test it’s relationships among the personal background, perceived health status, exercise pattern and academic achievement for elementary school students, which can be the reference of resilience educational program in elementary school in the future.  RESEARCH DESIGN AND METHODS-(1) Scale reediting: Based on the resilience model by Grotberg, current resilience scale and correlated literature, and the expert content validity, “Resilience Scale for Elementary School Students”(RSE) was developed. One hundred students (4th, 5th and 6th grade) from an elementary school in Taichung were chosen to do the testing as the sample for the pilot study. This scale were analyzed by using item analysis, and test-retest reliability. (2) Scale reediting: By stratified random sampling, main study sample were 865 elementary school students (4th, 5th and 6th grade) from 9 schools, which located in 7 different districts in Taichung City. This data was analyzed by using consistency reliability, factor analysis, descriptive statistics, Pearson’s correlation, t-test, One-Way ANOVA, and Multiple Regression analysis.
RESULTS-The major findings of this study were summarized as follow:

1. Scale-reediting: There are sixty seven items after reediting the RSE, and the RSE consists 3 subscales: external resources, locus of control and social skills. (1) Reliability: The Cronbach α value of this scale is between 0.69 and 0.91 which shows a good internal consistency. The coefficient of test-retest reliability is between 0.56 and 0.84 which is closed to the significant level of 0.01. This scale shows a great reliability. (2)Validity: Four or Five main elements were chosen from the factor analysis in each subscale. All the main factors fit the conceptual framework. The pearson’s correlation(γ) value were between 0.35~0.90 in each subscales and whole scale, which is closed to the significant level of 0.01., means the scale has a good internal consistence validity. From the listed outcome we can tell that the scale shows a good validity.

2.Scale-testing：(1) The score of resilience scale: 4th grader have the higher average than 5th and 6th grader in” resilience scale”, “external resource scale” and “social skills scale”. One the other hand, 6th grader have the higher average in “ locus of control scale”. Comparing to boys, girls showed the higher average in each resilience scale and subscales. (2) Significant correlations were found among the student’s sex, age, parental work, perceived health status, exercise pattern, and perceived academic achievement. No Significant correlations were parental education level and marital status. In multiple regression, above variables showed significant different (F= 16.35, p<0.001)with the RSE, and explained the 24%.