Validation of results is an important phase in the organization of a researcher’s work. Libraries and the internet offer a number of sources for guidance with respect to conducting validation in a variety of fields. However, construction engineering and management (CEM) is an area for which such information is unavailable. CEM is an interdisciplinary field, comprised of a variety of subjects: human resources management, project planning, social sciences, etc. This broad range means that the choice of appropriate validation methodologies is critical for ensuring a high level of confidence in research outcomes. In other words, the selection of appropriate validation methodologies represents a significant challenge for CEM researchers. To assist civil engineering researchers as well as students undertaking master’s or doctoral CEM studies, this thesis therefore presents a comprehensive review of validation methodologies in this area. The validation methodologies commonly applied include experimental studies, observational studies, empirical studies, case studies, surveys, functional demonstration, and archival data analysis. The author randomly selected 365 papers based on three main perspectives: industry best practices in construction productivity, factors that affect labour productivity, and technologies for improving construction productivity. The validation methodologies that were applied in each category of studies were examined and recorded in analysis tables. Based on the analysis and discussion of the findings, the author summarized the final results, indicating such items as the highest percentage of a particular methodology employed in each category and the top categories in which that methodology was applied. The research also demonstrates a significant increasing trend in the use of functional demonstration over the past 34 years. As well, a comparison of the period from 1980 to 2009 with the period from 2010 to the present revealed a decrease in the number of papers that reported validation methodology that was unclear. These results were validated through analysis of variation (ANOVA) and least significant difference (LSD) analysis. Furthermore, the relationship between the degree of validation and the number of citations is explored. The study showed that the number of citations is positively related to the degree of validations in a specific category, based on the data acquired from the examination of articles in Constructability and Factors categories. However, based on the data acquired from the examination of articles in the year 2010, we failed to conclude that there existed significant difference between clear-validation group and unclear validation group at the 95 % confidence level.

Validation of results is an important phase in the organization of a researcher’s work. Libraries and the internet offer a number of sources for guidance with respect to conducting validation in a variety of fields. However, construction engineering and management (CEM) is an area for which such information is unavailable. CEM is an interdisciplinary field, comprised of a variety of subjects: human resources management, project planning, social sciences, etc. This broad range means that the choice of appropriate validation methodologies is critical for ensuring a high level of confidence in research outcomes. In other words, the selection of appropriate validation methodologies represents a significant challenge for CEM researchers. To assist civil engineering researchers as well as students undertaking master’s or doctoral CEM studies, this thesis therefore presents a comprehensive review of validation methodologies in this area. The validation methodologies commonly applied include experimental studies, observational studies, empirical studies, case studies, surveys, functional demonstration, and archival data analysis. The author randomly selected 365 papers based on three main perspectives: industry best practices in construction productivity, factors that affect labour productivity, and technologies for improving construction productivity. The validation methodologies that were applied in each category of studies were examined and recorded in analysis tables. Based on the analysis and discussion of the findings, the author summarized the final results, indicating such items as the highest percentage of a particular methodology employed in each category and the top categories in which that methodology was applied. The research also demonstrates a significant increasing trend in the use of functional demonstration over the past 34 years. As well, a comparison of the period from 1980 to 2009 with the period from 2010 to the present revealed a decrease in the number of papers that reported validation methodology that was unclear. These results were validated through analysis of variation (ANOVA) and least significant difference (LSD) analysis. Furthermore, the relationship between the degree of validation and the number of citations is explored. The study showed that the number of citations is positively related to the degree of validations in a specific category, based on the data acquired from the examination of articles in Constructability and Factors categories. However, based on the data acquired from the examination of articles in the year 2010, we failed to conclude that there existed significant difference between clear-validation group and unclear validation group at the 95 % confidence level.

本研究所使用違約機率（Probability of Default, 以下簡稱PD）的預測方法為Robust Logistic Regression（穩健羅吉斯迴歸），本研究發展且應用這個方法是基於下列兩個觀察：1. 極端值常常出現在橫剖面資料，而且對於實證結果往往有很大地影響，因而極端值必須要被謹慎處理。2. 當使用Logit Model（羅吉斯模型）估計違約率時，卻忽略極端值。試圖不讓資料中的極端值對估計結果產生重大的影響，進而提升預測的準確性，是本研究使用Logit Model並混合Robust Regression（穩健迴歸）的目的所在，而本研究是第一篇使用Robust Logistic Regression來進行PD預測的研究。 變數的選取上，本研究使用Z-SCORE模型中的變數，此外，在考慮公司的營收品質之下，亦針對公司的應收帳款週轉率而對相關變數做了調整。 本研究使用了一些信用風險模型效力驗證的方法來比較模型預測效力的優劣，本研究的實證結果為：針對樣本內資料，使用Robust Logistic Regression對於整個模型的預測效力的確有提升的效果；當營收品質成為模型變數的考量因素後，能讓模型有較高的預測效力。最後，本研究亦提出了一些重要的未來研究建議，以供後續的研究作為參考。 / The method implemented in PD calculation in this study is “Robust Logistic Regression”. We implement this method based on two reasons: 1. In panel data, outliers usually exist and they may seriously influence the empirical results. 2. In Logistic Model, outliers are not taken into consideration. The main purpose of implementing “Robust Logistic Regression” in this study is: eliminate the effects caused by the outliers in the data and improve the predictive ability. This study is the first study to implement “Robust Logistic Regression” in PD calculation. The same variables as those in Z-SCORE model are selected in this study. Furthermore, the quality of the revenue in a company is also considered. Therefore, we adjust the related variables with the company’s accounts receivable turnover ratio. Some validation methodologies for default risk models are used in this study. The empirical results of this study show that: In accordance with the in-sample data, implementing “Robust Logistic Regression” in PD calculation indeed improves the predictive ability. Besides, using the adjusted variables can also improve the predictive ability. In the end of this study, some important suggestions are given for the subsequent studies.