作者: Ivan Sever , Eva Klaric , Zrinka Tarle
DOI: 10.1007/S00784-015-1600-7
关键词:
摘要: Dental microhardness experiments are influenced by unobserved factors related to the varying tooth characteristics that affect measurement reproducibility. This paper explores appropriate analytical tools for modeling different sources of variability reduce biases encountered and increase validity studies. The enamel human third molars was measured Vickers diamond. effects five bleaching agents—10, 16, 30 % carbamide peroxide, 25 38 % hydrogen peroxide—were examined, as well effect artificial saliva amorphous calcium phosphate. To account both between- within-tooth heterogeneity in evaluating treatment effects, statistical analysis performed mixed-effects framework, which also included weighting procedure adjust confounding. results were compared those standard ANOVA model usually applied. weighted produced parameter estimates magnitude significance than model. former more intuitive, with precise better fit. Confounding could seriously bias study outcomes, highlighting need robust procedures dental research reliability. presented framework is flexible informative existing techniques may improve quality inference research. Reported be misleading if underlying measurements not taken into account. confidence outcomes increased applying presented.