Browsing by Author "Huang, GH"

Now showing 1 - 6 of 6

Building an identifiable latent class model with covariate effects on underlying and measured variables
(2004-03-01)
Huang, GH
;
Bandeen-Roche, K
;
統計學研究所
;
Institute of Statistics
In recent years, latent class models have proven useful for analyzing relationships between measured multiple indicators and covariates of interest. Such models summarize shared features of the multiple indicators as an underlying categorical variable, and the indicators' substantive associations with predictors are built directly and indirectly in unique model parameters. In this paper, we provide a detailed study on the theory and application of building models that allow mediated relationships between primary predictors and latent class membership, but that also allow direct effects of secondary covariates on the indicators themselves. Theory for model identification is developed. We detail an Expectation-Maximization algorithm for parameter estimation, standard error calculation, and convergent properties. Comparison of the proposed model with models underlying existing latent class modeling software is provided. A detailed analysis of how visual impairments affect older persons' functioning requiring distance vision is used for illustration.
Selecting the number of classes under latent class regression: A factor analytic analogue
(2005-06-01)
Huang, GH
;
統計學研究所
;
Institute of Statistics
Recently, the regression extension of latent class analysis (RLCA) model has received much attention in the field of medical research. The basic RLCA model summarizes shared features of measured multiple indicators as an underlying categorical variable and incorporates the covariate information in modeling both latent class membership and multiple indicators themselves. To reduce complexity and enhance interpretability, one usually fixes the number of classes in a given RLCA. Often, goodness of fit methods comparing various estimated models are used as a criterion to select the number of classes. In this paper, we propose a new method that is based on an analogous method used in factor analysis and does not require repeated fitting. Two ideas with application to many settings other than ours are synthesized in deriving the method: a connection between latent class models and factor analysis, and techniques of covariate marginalization and elimination. A Monte Carlo simulation study is presented to evaluate the behavior of the selection procedure and compare to alternative approaches. Data from a study of how measured visual impairments affect older persons' functioning are used for illustration.
Self-rated health among young people with type 1 diabetes in relation to risk factors in a longitudinal study
(2004-02-15)
Huang, GH
;
Palta, M
;
Allen, C
;
LeCaire, T
;
D'Alessio, D
Care for type 1 diabetes mellitus should both alleviate the physical complications of the disease and improve overall quality of life. The Wisconsin Diabetes Registry, comprising a population-based cohort that is followed longitudinally from diagnosis of type 1 diabetes, provided the authors with a unique opportunity to examine quality of life in children, adolescents, and young adults with type 1 diabetes and its relation to both sociodemographic and clinical risk factors. The authors analyzed data from 1987-2002 with a mean of 10.2 years' duration of diabetes (n = 569). They used ordinal-scaled self-rated global health as a measure of quality of life. In this paper, they propose a random-effects model for drawing inferences on individuals regarding the relation of longitudinally measured quality of life to multiple risk factors. Results showed that male gender, higher parental socioeconomic level, younger age at diabetes diagnosis, shorter diabetes duration, no hospitalization in the preceding 6 months, lower glycosylated hemoglobin level, and questionnaire responses by a person other than the subject were independently associated with better reported health. The authors found that individuals varied in their reported health even after adjustment for all identified risk factors. This could imply either that there is variability in perception even with the same health status or that there are independent unmeasured risk factors for poor health in persons with type 1 diabetes.
Synthesis and characterization of light-emitting H-bonded complexes and polymers containing bis(pyridyl) emitting acceptors
(2006-01-24)
Lin, HC
;
Tsai, CM
;
Huang, GH
;
Tao, YT
;
材料科學與工程學系
A series of bis(pyridyl) acceptor emitters are synthesized and the LC and photophysical (PL and EL) properties of their supramolecular structures, including H-bonded polymer networks and trimers, are explored. Unique mesomorphic properties may be introduced into these H-bonded complexes and polymers containing nonmesogenic acceptor emitters. In addition, the emission properties of bis(pyridyl) acceptor emitters can be adjusted by their surrounding nonphotoluminescent proton donors. Compared with pure bis(pyridyl) acceptor emitters, red-shifts of PL emission wavelengths were observed in most of the H-bonded complexes and polymers. Because of the higher acidity value of m-PC(10)BA (16) and its stronger H-bonded effect in H-bonded polymer complexes, the largest red-shift of 93 nm in lambda(max) value of PL has occurred in the fully H-bonded polymer network p-PBBBP-OC8 (8)/m-PC(10)BA (16) in contrast to pure emitter p-PBBBP-OC8 (8). A very narrow full width at half-maximum (fwhm) value (38 nm) of EL spectra can be obtained in the device of H-bonded complexes blended with PVK, and a higher brightness of EL is produced at an appropriate annealing temperature.
Synthesis and characterization of light-emitting oligo(p-phenylene-vinylene)s and polymeric derivatives containing three- and five-conjugated phenylene rings
(2006-01-15)
Lin, HC
;
Tsai, CM
;
Huang, GH
;
Lin, JM
;
材料科學與工程學系
Several series of light-emitting oligo(p-phenylene-vinylene)s (BIII and BV series containing three- and five-conjugated phenylene rings) with various side groups and end groups attached to the cores were synthesized and characterized. The analogous PBV polymers, derived from the BV series, were also synthesized and investigated. Blue and greenish light emissions were observed in the photoluminescence (PL) and electroluminescence (EL) spectra of the blend and pure films with these pi-conjugated structures. In contrast to the three-conjugated ring oligomers, the five-conjugated ring derivatives (oligomers and polymers) had larger maximum emission wavelength values of PL and EL emissions. Mesomorphism was introduced into the BV series by the replacement of three-conjugated rings (BIII series) with five-conjugated phenyl cores (BV series). The liquid-crystalline properties of the BV series with end groups (on both end rings) were better than those of analogous BV-OC8 without end groups. Polarized PL emissions were obtained by the alignment of liquid-crystalline phase in rubbing cells. Upon heating, different PL emission wavelengths and intensities were observed in various phases. Not only the solubility and thermal properties but also the PL and EL properties could be effectively adjusted by the attachment of flexible alkoxy groups either on the central rings or on both end rings of the conjugated cores. (c) 2005 Wiley Periodicals, Inc.
Treatment of high strength hexamine-containing wastewater by Electro-Fenton method
(1999-02-01)
Chou, SS
;
Huang, YH
;
Lee, SN
;
Huang, GH
;
Huang, CP
A novel Electro-Fenton (EF-Fere) method, applied H2O2 and electrogenerated ferrous ion, was investigated for treating the hexamine-containing wastewater. The performance of Fe2+ generation in the electrolytic system was first evaluated, including the factors of the cathode material, initial pH, initial ferric concentration (Fe-i), and current density. When initial pH exceeded 2.5, the current efficiency dramatically decreased, which was due to the formation of Fe(OH)(3). Between 3000 and 10,000 mg/l of Fe-i, the initial current efficiency of Fe2+ generation was almost constant (85-87%), which dropped sharply to 39% at 1000 mg/l. In EF-Fere experiments, the COD removal efficiency attained above 94% after 5 h of reaction. The relationship between the temperature, dissolved oxygen, and COD was discussed. The changes in hexamine and its oxidation intermediates (methanol, formaldahyde, formate, ammonium and nitrate) during the reaction were also investigated. Three additional experiments using H2O2/Fe2+, H2O2/Fe3+, and direct electrolysis were also conducted to treat the hexamine-containing wastewater for comparison. The results showed that the EF-Fere method was the most efficient. (C) 1998 Elsevier Science Ltd. All rights reserved.