Publication:

大學體育術科學測體適能篩選模式之建構

對於各大學體育系所而言，招收具備各項體適能均衡發展的新生是首要的任務。本文嘗試使用多變量統計技術以建構一體適能篩選模式，以協助體育術科學測中心篩選出具各項體適能均衡發展的學生。由於此分析是針對參與術科學測的高中生，因此樣本選用大學入學體育術科考試考生體適能測驗資料，其測驗項目包括：60公尺立姿快跑、反覆側步、一分鐘屈膝仰臥起坐、立定連續三次跳、1600公尺跑走等五項測驗，其分數的計算是由各項分數加總而成。

模式建構的程序如下：首先，由國立台灣體育大學學測中心取得93年度大學入學術科考試體適能測驗資料，依據此測驗資料，將所有樣本進行群組分析(形成不同體適能的組別)。針對各體適能群組，並利用區別模式的建立，以作為欲報考各大專院校體育系所的高中職學生的參考。最後再驗證篩選模式的正確率。另外再利用主成份分析建構一體適能總指標評比模式，根據此總指標可評比受測學生的優劣差異，作為體適能篩選的基礎。

Recruiting a potential freshman of the well-developed physical fitness is the important mission for the department physical education of the university. This research is to construct a fitness-selecting model based on the multivariate statistic techniques. By using this model, we can choose the freshman of the well-developed physical fitness from the senior high school who participated in the athletic-technique test. Due to this model is aimed at students of senior high school ,the samples were chosen from the participator of the athletics-technique test. The evaluated items of physical fitness included: standing running、side step、1-minute flexed-leg sit-up test、standing tritle jump、1600m run.

The fitness model was constructed as follow: First, we obtained the samples from the university athletic-technique test center in National Taiwan Sport University. The sample data included the whole participators of athletic-technique test in 2004. Second, we used the cluster analysis to group the data and used the discriminant analysis to construct the fitness model. Finally, identifing the accuracy of the model. Utilizing principal component analysis constructs physical fitness evaluating model. According to this total index, the evaluation could be to the good and bad difference of measuring the student.