ScholarBank@NUShttps://scholarbank.nus.edu.sgThe DSpace digital repository system captures, stores, indexes, preserves, and distributes digital research material.Thu, 28 Sep 2023 22:02:53 GMT2023-09-28T22:02:53Z5041- Procuring honest responses indirectlyhttps://scholarbank.nus.edu.sg/handle/10635/105310Title: Procuring honest responses indirectly
Authors: Chua, T.C.; Tsui, A.K.
Abstract: We propose a new method for obtaining truthful answers to sensitive survey questions indirectly from a dichotomous population. In the proposed method, each sample respondent is asked to report a numerical value chosen from two parallel set of multiple outcomes generated by arbitrary randomizing devices. We show that the proposed estimator of the population proportion is more efficient and general than those put forward by Warner (1965. J. Amer. Statist. Assoc. 60, 63-69), Franklin (1989. Survey Methodol. 15, 225-235) and Kuk (1990. Biometrika 77, 436-438). Intuitively, our approach creates a remarkably respondent-friendly environment for each interviewee to discharge honest answers without the fear of exposing one's identity. Two illustrative cases using the normal and exponential distributions as randomizing devices are performed to demonstrate the superiority of the proposed estimator. © 2000 Elsevier Science B.V.
Fri, 01 Sep 2000 00:00:00 GMThttps://scholarbank.nus.edu.sg/handle/10635/1053102000-09-01T00:00:00Z
- Linear transformation, its application in the paper industryhttps://scholarbank.nus.edu.sg/handle/10635/105199Title: Linear transformation, its application in the paper industry
Authors: Surya, P.I.; Chua, T.C.; Khoo, W.K.
Abstract: If one variable (y) can be expressed in another variable (y′) such that y′ = a0 + a1y, where a0 and a 1 are constants, we say that y is linearly transformed into y′. Many descriptive measures retain their interpretation through linear transformation. Not only that, if the relationship between y (dependent variable) and x (independent variable) can be expressed using linear regression, the value of regression coefficient of determination; R2, obtained will be exactly the same if y is replaced with y′. Care needs to be exercised when assigning a case as linear transformation, because it may appear to be; actually, it is not.
Thu, 01 Aug 2002 00:00:00 GMThttps://scholarbank.nus.edu.sg/handle/10635/1051992002-08-01T00:00:00Z
- The use of linear regression to estimate replacement ratio - An example from paper applicationshttps://scholarbank.nus.edu.sg/handle/10635/105507Title: The use of linear regression to estimate replacement ratio - An example from paper applications
Authors: Surya, P.I.; Chua, T.C.; Tan, E.K.; Chong, A.L.; Goh, L.
Abstract: Replacement ratio is a concept that attempts to answer the following question in a quantitative manner: How good is one product compared to the other for achieving a given level of performance? Even though this concept is not new and has been used widely in various disciplines, to date, there is no literature that discusses it fully. In this article, we address this void. As such, this work has several objectives: First, to show that replacement ratio is a useful concept if the aim is to compare two products by focusing on performance. Second, to demonstrate how to estimate replacement ratio using regression equation. Third, to interpret the meaning of the margin of error associated with the replacement ratio estimate. And finally, to discuss the limitations of the replacement ratio procedure. To achieve the above objectives, we used retention and drainage devices.
Sun, 01 May 2005 00:00:00 GMThttps://scholarbank.nus.edu.sg/handle/10635/1055072005-05-01T00:00:00Z
- Procuring honest responses indirectlyhttps://scholarbank.nus.edu.sg/handle/10635/20051Title: Procuring honest responses indirectly
Authors: Chua, T.C.; Tsui, A.K.
Sat, 01 Jan 2000 00:00:00 GMThttps://scholarbank.nus.edu.sg/handle/10635/200512000-01-01T00:00:00Z