國內高等教育快速擴張，學生來源議題受到關注並投入研究，但受限於資料全面取得不易，僅能就單一大學的分析，而無法做到同時期校際間之比較研究。本研究應用臺閩地區2000 年戶口及住宅普查資料，設定篩選條件擷取大學生，並分析其求學與來源地點，結果發現：居住學校宿舍的學生所分佈的村里，與實際學校的分佈相當吻合，可辨識學生所就讀的學校。但卻發現普查資料不合理的現象：1. 學生求學地點有308 個鄉鎮市，遠多於實際學校分佈的69 個；2. 有
56%的學生來自於該校所在地相同的鄉鎮；3. 有77%的居住學校宿舍學生，其五年前仍居住該宿舍。造成此類資料異常的可能原因有：受訪者對大學和專科的教育程度認知差異、對求學地點的誤解以及普查採用留置填表方式。建議相關之普查作業能夠：1. 明確區分及定義大學和專科的教育程度；2. 實施留置填表時，應加以明確說明，並請學生確實填寫；3. 針對本研究發現的異常樣態訂定檢誤機制。

With the rapid expansion of higher education in Taiwan, the distribution of enrolled students has received much attention and been the focus of many studies, yet most results were limited due to the difficulty of collecting data spanning the same period from each university. This research draws on the 2000 census data of Taiwan’s population for analysis of the distributions of enrolled university students who satisfied the conditions. The results showed the amount and distribution of boarding students’ boroughs and those of the universities matched quite well. However, the study also uncovered irregularities. According to the census data, the number of towns where students attended schools was 308, far more than the actual number of 69; about 56% of students chose to attend schools in their home towns, and about 77% of boarding students lived in the same dormitories as they did five years ago. Possible reasons behind these significant differences include differences in the interviewees’ understanding of educational levels, misunderstandings of the location of the schools attended, and census survey methods. We suggest that for the future census, education levels should be defined precisely, items on the census form have to be explained
clearly, students should be asked to fill in the questionnaires carefully, and a double-check mechanism should be established to prevent potential mistakes discovered in this study.