Monte Carlo Methods Applied in Health Research
The root surface area of the tooth (RSA) is an important prognostic factor in the field of dentistry. The estimation of RSA from routine clinical data, such as tooth length (TL) and mesiodistal diameter of crown (MDC), is of interest because provides clinicians with information to decide objectively without additional costs. The aim of the paper is to determine the sample size for a regression analysis of RSA on TL and MDC using both power and parameter accuracy perspectives with Monte Carlo (MC) methods, as describe by Beaujean (2014).
A random sample of 5 lower second premolar teeth were scanned in X-ray microtomograph. The respective RSA were obtained through the planimetric method where the TL and MDC were measured on 1:1 photographs. The model of interest was defined as RSA = β0+ β1TL+ β2MDC, in accordance with the research question. The sample size was determined based on the model of interest and strength of the relations among the variables using the MC methods. The packages lavaan and simsem of R software were used to define the model and to run the simulations.
A sample size of 37 was calculated meeting the criteria for Monte Carlo data quality proposed by Muthén and Muthén (2002).
KeywordsRoot surface area Sample size Monte Carlo methods
- Beaujean, A. A. (2014). Sample size determination for regression models using Monte Carlo methods in R. Practical Assessment, Research & Evaluation, 19, 12.Google Scholar
- Forster, M. R. (2001). The new science of simplicity. In A. Zellner, H. A. Keuzenkamp, & M. McAleer (Eds.), Simplicity, inference and modelling: Keeping it sophisticatedly simpley (pp. 83–119). Cambridge: Cambridge University Press.Google Scholar
- Hair, J. F., et al. (2012). Multivariate data analysis (6th ed.). Upper Saddle River: Prentice-Hall Inc.Google Scholar
- Hays, W. L. (1973). Statistics for the social sciences (2nd ed.). New York: Holt, Rinehart & Winston.Google Scholar
- Nachtigall, C., Kroehne, U., Funke, F., & Steyer, R. (2003). (Why) should we use SEM? Pros and cons of structural equation modeling. Methods of Psychological Research Online, 8(2), 1–22.Google Scholar
- Pornprasertmanit, S., Miller, P., & Schoemann, A. (2012). simsem: SIMulated Structural Equation Modeling [Computer software].Google Scholar