Philip K.

I turn data into statistical results you can defend, whether the audience is a dissertation committee, a journal reviewer, a manager, or an auditor. If you have a dataset, a research question, or a decision that depends on numbers, I take it from raw files through analysis and validation to results you can explain and justify. WHEN CLIENTS BRING ME IN 🔹 You are unsure which statistical method fits. I assess your research question, data structure, and study design, select an appropriate method, check its assumptions, and explain why it is suitable. I work with SPSS, R, Python, and Stata. 🔹 Your analysis will be examined or challenged. I provide statistical analysis for dissertations, journal research, clinical trials, and other studies requiring defensible methodology. Deliverables include statistical interpretations, APA 7 tables and charts, reproducible code or syntax, and explanations to support your presentation or response to reviewers. 🔹 You need to evaluate experimental results or estimate effects. I work with hypothesis testing, regression, ANOVA, experimental data, and Bayesian analysis. My project experience includes developing an R/JAGS script for hierarchical Bayesian meta-analysis of marketing A/B tests. 🔹 You also need business analytics. I work with SQL, financial and operational data, Power BI and Tableau dashboards, forecasting, and predictive modelling. PROOF OF WORK 🔹 Kenya Revenue Authority, Domestic Taxes (May 2021–May 2023): Statistician. Conducted quarterly data-quality checks, record validation, and missing-data assessments across 50,000+ records. Investigated reporting discrepancies totalling KES 2.3 million for resolution. 🔹 Research and statistical analysis: Developed statistical analysis and data management plans for 21 research projects, covering sampling strategies, methods, validation procedures and reporting requirements. 🔹 Clinical research: Supported data management and statistical activities across 7 healthcare randomised controlled trials and factorial experiments, including instrument design, protocol-compliant data collection and hypothesis testing. 🔹 Bayesian analysis: Developed an R/JAGS hierarchical Bayesian meta-analysis script for marketing A/B testing, modelling pooled effects and between-test heterogeneity with posterior estimates and credible intervals. 🔹 Upwork track record: Top Rated, 100% Job Success Score, and 5.0 client rating. 🔹 Education: BSc Statistics; MSc Data Science . WHAT CLIENTS SAY “Philip again came through for me when I required help with formatting my statistical analysis charts in the most current APA, 7th edition style. ... He understands statistics and can help with plausible explanations for your data. ... He presents his ideas in a mature manner and provides solid justifications for them.” “After many weeks of meetings with my committee members, I requested five extra bar charts to corroborate the conclusions of my statistical study, and Philip delivered them once again. ... He communicates effectively and provides excellent reasoning to back his responses.” “He takes the time to describe the statistical analysis methodology when necessary. ... He went above and beyond by providing me with Key Notes for my upcoming presentation.” HOW I PROTECT YOUR RESULTS 🔹 I do not choose a statistical test simply because it is commonly used. I assess whether it fits the research question, study design, and data. 🔹 I examine relevant assumptions and limitations, including sample-size adequacy, multicollinearity, inconsistent variable definitions, and data leakage where predictive modelling is involved. 🔹 I distinguish statistical findings from conclusions the data cannot support. If a method or dataset cannot answer the question reliably, I explain why. 🔹 I provide reproducible code or syntax, clear interpretations, and supporting notes so that your analysis can be reviewed, rerun, and explained. START HERE Send your research question, a brief description of your dataset, and your deadline. I will review the scope, recommend an appropriate analytical approach, identify potential limitations, and provide a clear quote.