For academic researchers and postgraduate students — from a first thesis to a documented analysis.
01 — Sample Size Calculation
Know exactly how many participants your study needs Determine the number of participants needed to detect a meaningful effect with statistical confidence, so you neither waste resources nor end up underpowered.
Power analysis for RCTs, cross-sectional, case-control and cohort designs Sample size for hypothesis tests, confidence intervals and equivalence / non-inferiority Attrition adjustments, clustered designs and multiple comparison corrections Written justification paragraph you can paste into a protocol or thesis Software: G*Power, PASS, nQuery, R (pwr, WebPower, simr)
02 — Survey Analysis
Turn questionnaire responses into validated, publishable insight From instrument design to psychometric modelling — reliability, factor structure and the full measurement story behind your questionnaire.
Questionnaire design consultation, cognitive testing and pilot analysis Reliability: Cronbach's alpha, McDonald's omega, test-retest, ICC Exploratory (EFA) and Confirmatory (CFA) factor analysis Structural equation modelling, path analysis and measurement invariance Software: SPSS, R (lavaan, psych), Mplus, Jamovi
03 — Clinical Biostatistics
End-to-end statistics for clinical and health research From raw clinical data to publication-ready outputs, with the assumption checks and sensitivity analyses reviewers ask for.
Descriptive and inferential analysis for every data type Parametric and non-parametric testing matched to the study design Regression modelling: linear, logistic, Poisson, ordinal and multinomial Model diagnostics, assumption checking, outlier detection and sensitivity analysis Software: R, SPSS, STATA, Python
04 — Statistical Modeling
Advanced models for complex research questions Longitudinal, clustered and time-to-event data handled with the right model — and explained in language your committee or reviewers will follow.
Mixed-effects models (LMM, GLMM) for longitudinal and clustered data Survival analysis: Kaplan-Meier, Cox regression, competing risks, parametric models Generalized Estimating Equations (GEE) for repeated measures Machine learning for research: random forests, gradient boosting, SVM with interpretability Software: R (lme4, survival, tidymodels), STATA, Python (scikit-learn)
05 — Meta-Analysis
Synthesise evidence across studies, rigorously Everything from literature screening to publication bias assessment, with forest plots and heterogeneity diagnostics ready for submission.
Systematic search strategy development and study selection support Effect size extraction and transformation (OR, RR, SMD, Hedges' g, correlation) Fixed-effects, random-effects (DL, REML) and hierarchical Bayesian models Heterogeneity, meta-regression, subgroup analysis and publication bias diagnostics Software: R (metafor, meta), RevMan, CMA
06 — Data Visualization
Publication-ready figures that read at a glance Charts and diagrams built to journal specification — correct sizing, typography and colour, exported at the resolution the publisher wants.
Publication figures in R (ggplot2), Python (matplotlib/seaborn) and GraphPad Prism Custom colour palettes, typography and layouts matching journal guidelines Multi-panel figures with shared legends, highlighting and annotations Forest plots, Kaplan-Meier curves, volcano plots, heatmaps and correlation matrices Software: R (ggplot2), Python (matplotlib, seaborn), GraphPad Prism
07 — Data Cleaning & Validation
Start every analysis with data you can defend Audit, clean and validate your dataset before any model is fitted, with a documented trail of every decision made along the way.
Raw data auditing: missingness patterns, duplicate detection, logic checks Missing data handling with multiple imputation (MICE, Amelia) and sensitivity analysis Outlier identification using statistical and visual diagnostics Standardisation, variable coding, transformation and full cleaning documentation Software: R (tidyverse, mice), SPSS, Python (pandas)
