Biostatistics & Statistical Modeling Portfolio

Author: Amira Mandour
Biostatistician | Clinical Trials & Statistical Modeling

2023-01-20

Welcome to My Biostatistics Portfolio

I am Amira Mandour, a biostatistician with expertise in data analysis, statistical modeling, data visualization, and R programming. I specialize in using R and RMarkdown to automate reports and create dynamic, reproducible analysis workflows. In this portfolio, you’ll find a diverse set of projects involving complex statistical analysis and data visualization, showcasing various statistical techniques from exploratory data analysis and hypothesis testing to predictive modeling and survival analysis. Each project applies a unique set of statistical tests and methodologies.

Portfolio Projects

Below are examples of formal reports I created for projects’ statistical analysis.

Each with a brief description and a link to the full analysis.

Project 1: Knowledge, Attitude, and Perception of Public Towards Participation in COVID-19 Clinical Trials: A Cross-Sectional Study

Tools used: R, R Markdown, ggplot2, ggpubr, corrplot, dplyr, tidyr, gtsummary, psych, finalfit.

Key Techniques Used: Cronbach’s alpha for reliability, KAP scoring system, descriptive statistics, Correlation analysis, logistic regression, regression modeling, data visualization with ggplot2 and corrplot.

Project: View Project

Code: View Code on GitHub

Project 2: Diabetes Risk Factor Modeling: A Logistic Regression Approach to Identifying Risk Factors

Tools Used: R, R Markdown, ggplot2, plotly, naniar, gt, kableExtra, lmtest, ROCR, qqplot, performance, glm, dplyr, tidyr, caret, broom.

Key Techniques Used: Odds ratios, Logistic regression, multivariable analysis, variance inflation factor (VIF), residual diagnostics (residuals vs. fitted values, normal Q-Q plot, scale-location plot), ROC curve analysis, model comparison, and visualizations (forest plot, logistic regression curves).

Project: View Project

Code: View Code on GitHub

Project 3: Safety Outcomes From a Randomized Controlled Trial: Comparative Analysis of Vital-Sign–Related Adverse Events

Tools Used: dplyr, stringr, rstatix, tidyverse, ggpubr, Hmisc, labelled, sjlabelled, finalfit, gtsummary, forcats, psych, reshape2, ggplot2, knitr, kableExtra, geepack, MESS, emmeans, ggeffects, sjPlot, insight.

Key Techniques Used: Longitudinal data analysis, Generalized Estimating Equations (GEE), normality testing, sensitivity analysis, model comparison, and visualizations (time-series plots).

Project: View Project

Code: View Code on GitHub

Project 4: Modeling Risk Factors for Myocardial Infarction Using Conditional Logistic Regression: Insights from a Matched Case–Control Study

Tools Used: gtsummary, dplyr, survival, gt, ggplot2, ggtext.

Key Techniques Used: Odds ratio estimation, Conditional logistic regression, matched case-control study design, model fit testing (Likelihood ratio test, Wald test, Score test), pseudo R², forest plot visualization, and predicted risk modeling.

Project: View Project

Code: View Code on GitHub

Project 5: Effect of Cognition-Targeted vs Symptom-Targeted Exercise on MS Fatigue

Tools Used: geepack, MESS, emmeans, ggplot2, dplyr, gt, gtsummary.

Key Techniques Used: Generalized Estimating Equations (GEE), Repeated measures analysis, longitudinal data analysis, interaction modeling, outcome comparison between groups, and visualization (time-series plots).

Project: View Project

Code: View Code on GitHub

Project 6: Evaluating Risk of Fractures Among Patients: A Poisson Regression Analysis of Drug Effects

Tools Used: ggplot2, dplyr, gtsummary, ggpubr, gt, glm.

Key Techniques Used: Count data analysis, Poisson regression, incidence rate ratio (IRR) estimation, model comparison by treatment type, controlling for age, and visualizations with ggplot2.

Project: View Project

Code: View Code on GitHub

Project 7: Repeated Measures ANOVA: Effect of Time on improving Self-Esteem levels Using R

Tools Used: datarium, rstatix, tidyverse, ggpubr, survival, psych, emmeans, ggplot2, dplyr, gtsummary, ggboxplot, ggqqplot, ggline.

Key Techniques Used: Repeated-measures ANOVA, Mauchly’s test for sphericity, Bonferroni-corrected post-hoc comparisons, effect size (η²), and visualizations (mean plots, boxplots).

Project: View Project

Code: View Code on GitHub

Project 8: From “The WORST” to “THE BEST”: Modeling Students’ Attitudes Toward Studying Statistics Using Ordinal Logistic Regression

Tools Used: ordinal, gt, gtsummary, broom.mixed, performance, dplyr.

Key Techniques Used: Handling ordinal outcomes, Ordinal logistic regression (Proportional Odds Model), likelihood ratio test, model comparison, AIC and log-likelihood, pseudo R² (Nagelkerke), proportional odds assumption test, odds ratios with 95% CI, Interpretation of ordered categories, significance testing (α = 0.05), and visualizations (bar plots, box plots).

Project: View Project

Code: View Code on GitHub

Project 9: Survival Analysis and Prognostic Modeling in Lung Cancer

Tools Used: survival, survfit, coxph, dplyr, tibble, knitr, kableExtra, survminer.

Key Techniques Used: subgroup analysis (gender differences), Kaplan-Meier survival analysis, log-rank test, Cox proportional hazards model, hazard ratios with 95% confidence intervals, testing of proportional hazards assumption (Schoenfeld residuals), stratified Cox model, Progression-Free Survival (PFS) analysis, and visualizations (survival curves, PFS curves).

Project: View Project

Code: View Code on GitHub

Project 10: Cox Proportional Hazards Regression Analysis in Lung Cancer

Tools Used: survival, survfit, coxph, dplyr, gt, gtsummary, tibble, knitr, kableExtra, survminer.

Key Techniques Used: Hazard ratios with 95% confidence intervals, Cox proportional hazards regression, testing of proportional hazards assumption (Schoenfeld residuals), stratified Cox model, model evaluation (C-index, likelihood ratio test, Wald test, score test), and visualizations (survival curves, Schoenfeld residuals plots).

Project: View Project

Code: View Code on GitHub

Skills & Expertise

Contact Information

You can reach me through the following channels:

Feel free to reach out for any questions or project inquiries!