Transactions of an international conference. Worldwide concern: better dental care for more people. An international collaborative study. How the oral and social measures were compared
Duppenthaler, J.L.
International Dental Journal 26(3): 314-316
1976
ISSN/ISBN: 0020-6539 PMID: 1067228 Document Number: 108874
This paper explains the statistical methods used to reach the conclusions regarding the relationship between selected social variables and oral health status reported in subsequent papers referring to specific countries. The procedure called 'stepwise regression' was employed in relation to each of the three clinical categories of data. The single social variable which explains the greatest variation in the oral data is identified as the most important predictor. Its influence is then removed by mathematical adjustment and the second and subsequent social variables are similarly identified until no further ones remain which are significant in explaining remaining variability in the oral data. The effect of utilization was always removed first regardless of its predictive power. The findings for each country were used to construct 'tree' diagrams on which the best and worst pathways are indicated to give a profile of those subjects having the fewest and the most decayed teeth.