학동의 치면별 우식경험영구치면수 증가 예측변수선택에 관한 계량적 연구
Prediction Variable Selection of DMFS Increments
문혁수,
소속 상세정보
문혁수 ( ) - 서울대학교
KMID : 0355419930170010013
Abstract
For the purpose of selecting variables for developing the DMFS prediction model to improve oral health. Clinical , microbiological and questionnair data were collected from 6-year-old school children, 1991. For this purpose, one thousand
first-grade
elementary school children were selected for longitudinal study. Oral examination for selected subjects were conducted in 1991 and 1992 by one trained examiner to get one year increments of DMFS.
The one year increments of DMFS were used as a respondent variable. Clinical, microbiological, dietary and sociodemographic variables were used prediction variable. Using the variable selection criteria; residual mean square(MSEp2 ) and total
squared
error(Cp). Prediction variables were selected according to smooth surface DMFS and pit & fissure. The results obatained were as follows:
1. Fluorosis indes, S. mutans, frequencies of snack intake between breakfast and lunch, and employment status of mother were selected as predication variable for both smooth surface and pit & fissure DMFS incremets prediction.
2. S. mutans, fluorosis index and frequencies of snack intake between breakfast and lunch were selected as prediction variable for pit & fissure DMFS increments predication.
3. Fluorosis index, water fluoridation, sex, habit of snack and S. mutans were selected as prediction variable for smooth surface DMFS increments prediction.
4. prediction ability of smooth surface DMFS increments prediction model and pit & fissure DMFS increments prediction model were higher than those of both smooth surface and pit & fissure DMFS increments prediction model.
키워드
원문 및 링크아웃 정보
등재저널 정보