Agnes Matope

Statistician/Mathematical Biologist

I hold a Master’s degree in biostatistics with distinction (University Hasselt, Belgium (2015-2017)) and a Bachelor of Science degree (University of Malawi, Chancellor College (2013), a major in Statistics and Mathematics). During this period, I have been highly involved in project research work and data analysis from different therapeutic areas. I am well versed in clinical trials and health research with specialisation in longitudinal data analysis, multi-level/clustered data analysis, survival data analysis and dose response modelling. Through my MSc. thesis titled: Robust Methods in Dose Response Analysis I have developed skills in analysing dose-response data considering the order restrictions in the dose-response modelling relationship. This involved fitting isotonic/antitonic regression models for simple order alternative trends and unimodal partial order alternatives also known as umbrella profiles for the non-monotone profiles. Estimation was conducted using both the median and mean fit estimates. On the other hand, classification of compounds to different dose-response relationship trends was conducted using the two-stage Order Restricted Information Criterion Clustering algorithm (ORICC). Inference was conducted using the Likelihood Ratio Test (LRT) based on permutations tests using the median and mean fit as well as, Multiple Contrasts Tests (MCT). Outliers were detected using a t-test with a bonferroni adjustment procedure based on the studentized deleted residuals from a one-way Analysis of Variance (ANOVA) model that was fitted for each compound as well as, graphical techniques (scatter plots). Primary software used were R and SAS.

I am well versed in advanced data modelling using the following statistical software: SAS, Stata, SPSS, R-package, WINBUGS/OPENBUGS, Latex for writing pdf documents and Microsoft Office Applications (Word, Excel, PowerPoint, Publisher, Outlook).


Medical statistics: I have experience in employing the different statistical methods to medicine and health/clinical sciences which involves designing and analysing of medical research studies. I have experience in fitting statistical models using linear models for continuous response using Analysis of Variance (ANOVA) and Regression; multi-category logit models using the continuation-ratio logit model, proportional odds model, cumulative logit models and adjacent-category logit models for ordinal response data; Poisson regression models for count data and logistic regression models for binary data. I am also well versed in fitting nonlinear models and models for repeated measurements. In addition, I have experience in conducting Bayesian modelling for linear and non-linear models, and conducting inference using bootstrap (parametric, semi-parametric and non-parametric) and permutations tests.

Models for longitudinal/multi-level/clustered data: I am well versed in analysing longitudinal or clustered data using finite mixture models, and linear and non-linear mixed models for a continuous response and Genialized Linear Mixed Models (GLMM) and Generalized Estimating Equations (GEE) for a categorical response. I also have experience in conducting the analysis in the presence of missing data using Weighted GEE and Multiple Imputation as well as, doing a sensitivity analysis.

Models for Survival data analysis: I am well knowledgeable in analysing univariable time to event data using the Kaplan-Meier estimator, fitting multivariable time-to-event regression models using Cox Proportional Hazard models and Accelerated Failure Time (AFT) models as well as, fitting frailty models for correlated survival data. I also have experience in Bayesian analysis for time-time-event using the frailty model.