隨著電腦及統計模型的發展、原本用於稅賦評估工具的大量估價逐漸受到重視。美國更利用完善的國家層級資料，將大量估價改良為自動估價模型（Automated Valuation Model, AVM）。由於估值估算快速、成本低廉及龐大的資料庫等優點，受到美國私部門的歡迎。另一方面，不動產市場仍有存在著個別估價的客觀性的質疑，而估價行為學也證實了個別估價有偏離估價程序的可能。市場上存在著兩者的支持者，但由於兩者特性迥然不同，甚難比較。 本研究以一致的資料庫及衡量標準，將交易價格做為市場價值表徵，以估值的準確性作為衡量兩者的準則，找出兩者的特性及適用範圍。實證結果顯示當勘估標的數量達到一定程度，可忽略不動產的異質性時，個別估價與大量估價準確性並沒有顯著的差異。但個別估價準確性的分配較大量估價集中、且偏誤程度也較低。個別估價較不受不動產特性的影響，適用的範圍較廣；而大量估價較易受到不動產特性影響，適用範圍視資料庫的內容而定。另一方面，個別估價的估價認知具有時間性，表示不動產估價師需更新估價認知，以保持高度的估值準確性。此說明了國內不動產市場的變化快速及國內不動產估價師的專業能力。 / With the development of computer and statistical model, the mass appraisal as assessment tool has been paid attention gradually. Americans utilize the national data to develop Automated Valuation Models (AVMs).For its advantages of faster valuation , less cost and huge database, mass appraisal is used as assessment but also appraisal tool . On the other hand, the real estate market stills have query with objectivity of individual appraisers. The study of appraisers’ behavior suggest that the possibility of deviate from standard procedure of appraising. In a matter of fact, the real state market is filled with supporters of appraisers and mass appraisals. Because of these different characteristic, it’s difficult to compare with. This research will trade selling-prices to seek for the market value form with unanimous database and criterion, regard accuracy of valuation as and weigh the criterion of the two, and find out the characteristic of the two and scope of application .Although the result showed that the accuracy is no difference between appraiser and mass appraisal significantly when the amount of real estates is large enough to neglect heterogeneity. Compare with mass appraisal, the distribution of appraiser’s accuracy is more centralized and the standard deviation is smaller. The accuracy of appraiser is steady, and the scope of application is relatively wide. But the accuracy of mass appraisal is affected by the characteristic of real estate, and the scope of application depends on content of the database.On the other hand, the appraisal cognition has timeliness, it shows appraiser needs to upgrade the appraise cognition and the market condition , in order to keep the valuation accuracy of the height. This has stated the fast-change market and domestic the professional ability of appraisers.

國內近年來有許多文獻在進行特徵價格模型預測時，避免樣本中存在異常點會造成模型估計值產生偏差，會使用統計軟體進行異常點檢測，但皆是直接將檢測出的異常點刪除，未加以著墨探究這些異常點的特徵結構、成因及特色等。因此，本研究透過統計檢定方法，探討刪除異常點前後整體樣本的特徵結構變化，並以個別估價觀點加以探討住宅交易樣本異常點的成因與特色，藉此歸納出實價登錄資料未揭露的重要特徵，以及迴歸模型搜尋疑似申報不實案件之可行性。 透過敘述統計及樣本結構差異檢定結果發現，異常樣本的離散程度相對原始樣本與正常樣本較大，且經過刪除異常點的正常樣本特徵結構差異程度縮小；異常點的形成可能受到區位變數無法反映實際情況及樣本群聚程度影響，也可能因模型未納入某些重要的特徵變數，而使隱含該變數的樣本被判斷為異常點；異常樣本與正常樣本的成交總價、土地坪數、建物坪數、總樓層、所在樓層及屋齡等變數平均數、變異數及中位數有顯著差異。 藉由個案分析結果歸納，可能因異常個案的住宅屬性存在整幢大樓住商混合使用、特殊鄰居、附屬建物占比過高、高總價豪宅產品、都更效益、增建效益、裝潢效益、約定專用空間效益、樓層高度挑高、獨特視野景觀或特殊區位條件；外部環境存在鄰近嫌惡設施或迎毗設施；交易情況存在買方身分特殊之影響，但受限於實價登錄未要求登載並揭露這些特徵，故模型未考量這些因素對價格的影響，使得模型可能將隱含這些特徵的樣本判斷為異常點，並進而影響模型預測結果。另外也發現，實價登錄資料存在登載錯誤及價格申報不實的情況，且可能被模型判斷為異常點。 / Many literatures use statistics-way to detect outliers in preventing any extreme deviation in hedonic price model prediction. Nevertheless, deleting the outliers instead of investigation into the structures, causes and features. Hence, this thesis studies the feature structures variation of the sample before and after deleting the outliers and with the valuations by appraisers’ perspective to inquire into the factors and features of the outliers in residential transactions. Thereby to summarize the significant features that are not disclosed by real price registration and feasibility in searching the possible false declaration of price by regression. Through descriptive statistics and sample structural difference parametric and nonparametric test shows the discreteness level of singular (outliers only) samples is greater than the primary (outliers including) and normal (outliers deleting) samples and the feature structure variation lessened after deleting the outliers in normal samples. The formation of outliers may be influenced by location variable not able to reflect actual circumstances and level of clustering in samples. Maybe some significant variables are not subsumed into the model, which leads to the judgement of samples with this variable to be outliers. The mean, variance and median in total traded price, land size, building size, total floors, exact floor and house age of singular samples are notably different with normal ones. With the analysis of cases, the possible reasons may be residential and commercial mixed-use in building, peculiar neighbors, high proportion of accessory building, luxury houses, urban renewal benefits, building addition benefits, interior decoration benefits, agreed space benefits, high-ceiling benefits, unique view or location, YIMBY and NIMBY property in environment and special relationship between the buyer and seller. Nevertheless, due to the nondisclosure of these features in real price registration that the model does not take these into consideration. That leads to the judgement of samples with these features as outliers and affects the model prediction. Also the registration error and false declaration in price may also be judged as outliers.