Accounting for Measurement Error in Covariates in the Context of ANCOVA Using Maximum Likelihood Estimation
Wilson, Ana. (2022-05). Accounting for Measurement Error in Covariates in the Context of ANCOVA Using Maximum Likelihood Estimation. Theses and Dissertations Collection, University of Idaho Library Digital Collections. https://www.lib.uidaho.edu/digital/etd/items/wilson_idaho_0089n_12367.html
- Title:
- Accounting for Measurement Error in Covariates in the Context of ANCOVA Using Maximum Likelihood Estimation
- Author:
- Wilson, Ana
- Date:
- 2022-05
- Program:
- Mathematics & Statistical Sci
- Subject Category:
- Statistics
- Abstract:
-
Analysis of covariance (ANCOVA) is a common statistical model. An implicit assumption of ANCOVA is that the covariate is measured without error. However, in many applications, there is covariate measurement error. In this case, the estimates produced by classic ANCOVA methods can include bias, causing predictions and inferences to be inaccurate. This thesis uses monte carlo simulation to examine the effectiveness of an alternative model in estimating the parameters associated with ANCOVA. This model is shown to be effective in accounting for covariate measurement error in the case where there aretwo treatment groups.
- Description:
- masters, M.S., Mathematics & Statistical Sci -- University of Idaho - College of Graduate Studies, 2022-05
- Major Professor:
- Johnson, Timothy
- Committee:
- Fu, Audrey; Williams, Chris; Abo, Hirotachi
- Defense Date:
- 2022-05
- Identifier:
- Wilson_idaho_0089N_12367
- Type:
- Text
- Format Original:
- Format:
- application/pdf
- Rights:
- In Copyright - Educational Use Permitted. For more information, please contact University of Idaho Library Special Collections and Archives Department at libspec@uidaho.edu.
- Standardized Rights:
- http://rightsstatements.org/vocab/InC-EDU/1.0/