R Data Science and Statistical Modelling Training Course

Course overview

Organisations hold customer, operational, financial and digital data that could improve decisions, but analysts often spend too long preparing inconsistent files, producing unrepeatable reports or applying models without clear validation. This course addresses the practical gap between knowing basic R syntax and delivering defensible analytical work. Participants learn to turn raw business data into reproducible analyses, statistical models and decision-ready visualisations that colleagues can inspect, rerun and use.

The course develops an end-to-end R workflow using RStudio, tidyverse and tidymodels. Participants import and profile data, reshape tables, handle missing values and outliers, and create publication-quality graphics with ggplot2. They apply exploratory data analysis, hypothesis testing, linear and logistic regression, model diagnostics, resampling and performance evaluation. The programme also covers feature engineering, model comparison, reproducible reporting and clear communication of uncertainty, assumptions and limitations to non-technical stakeholders.

Teaching combines instructor demonstrations with guided coding labs based on realistic business datasets, including customer retention, demand and operational performance scenarios. Each participant builds an R project with documented scripts, a data-preparation pipeline, exploratory visualisations, a validated predictive or explanatory model, and a Quarto report presenting findings and recommendations. This tangible portfolio piece gives attendees a reusable structure for applying R data science methods to their own workplace data.

Course objectives

By the end of this course, participants will be able to:

• Build reproducible R projects using scripts, package management conventions and structured project folders
• Import, clean and reshape tabular data with dplyr, tidyr and explicit data-quality checks
• Create exploratory visualisations with ggplot2 to identify trends, segments, outliers and relationships
• Apply hypothesis tests and confidence intervals to answer defined business questions
• Fit and interpret linear and logistic regression models in R
• Evaluate model assumptions, residual diagnostics, cross-validation results and classification performance metrics
• Engineer features and compare candidate models using tidymodels workflows
• Produce a Quarto analytical report containing code, visualisations, model findings and recommendations

Benefits of attending

Target competencies

Who should attend

• Data Analysts — who need to replace manual spreadsheet analysis with reproducible R workflows
• Business Intelligence Analysts — who need to investigate drivers behind dashboard trends using statistical methods
• Data Scientists — who need stronger foundations in model validation, interpretation and R project structure
• Research Analysts — who need to analyse survey, operational or observational data with defensible statistical evidence
• Digital Analysts — who need to model conversion, churn or campaign outcomes from behavioural data
• Technical Consultants — who need to present R-based evidence and recommendations to client stakeholders

Requirements and prerequisites

Participants should be comfortable working with tables of data and understand basic statistical concepts including averages, variation, correlation and sampling. Prior exposure to R is required: attendees should be able to create objects, run a script, use vectors and data frames, and install or load packages. Experience with RStudio and basic functions such as read.csv(), filter() or summarise() is helpful. Participants need a laptop capable of running R and RStudio; installation guidance should be completed before the course. Advanced programming, calculus, machine learning experience and prior use of tidymodels are not required.

Training methodology

The programme is delivered through short instructor-led explanations followed by live RStudio demonstrations and progressively more independent coding labs. Participants work with realistic datasets rather than isolated syntax examples, using tidyverse pipelines to prepare data and tidymodels workflows to build and evaluate models. Facilitated reviews compare modelling choices, diagnostic evidence and stakeholder-facing conclusions. Daily exercises produce components of a single analytical project, and the final session includes peer review and an application-planning workshop to identify a suitable workplace dataset, question and reporting audience.

Course outline

Day 1: Reproducible R Data Analysis Foundations

• RStudio projects, scripts, console workflow and working-directory control
• R objects, vectors, factors, lists and data-frame structures
• Importing CSV and Excel data with readr and readxl
• Data profiling with glimpse(), summary() and targeted validation checks
• Tidy data principles and reshaping with pivot_longer() and pivot_wider()
• Filtering, grouping and aggregation with dplyr verbs
• Missing-value, duplicate-record and data-type treatment strategies

Workshop: Participants create a structured R project and produce a cleaned, documented customer-transactions dataset from inconsistent source files.

Day 2: Exploration, Visualisation and Statistical Evidence

• Exploratory data analysis questions and analytical framing
• Distribution analysis using histograms, density plots and box plots
• Segment comparison with grouped summaries and faceted graphics
• Relationship analysis using scatterplots, correlation and contingency tables
• ggplot2 aesthetics, layers, scales, themes and annotation
• Confidence intervals and practical interpretation of sampling uncertainty
• t-tests, chi-squared tests and non-parametric test selection

Workshop: Participants investigate variation in customer spend and retention, producing an annotated visual evidence pack and hypothesis-test summary.

Day 3: Regression Models for Business Questions

• Defining response variables, predictors and model hypotheses
• Simple and multiple linear regression with lm()
• Coefficient interpretation, effect sizes and confidence intervals
• Categorical predictors, interaction terms and reference categories
• Logistic regression for binary outcomes with glm()
• Odds ratios, predicted probabilities and classification thresholds
• Residual analysis, multicollinearity and influential-observation diagnostics

Workshop: Participants build and diagnose regression models explaining order value and predicting customer churn, then draft a plain-language findings note.

Day 4: Predictive Modelling and Validation with tidymodels

• Training, testing and resampling strategies for supervised learning
• Feature engineering with recipes in tidymodels
• Workflow construction for preprocessing and model fitting
• Cross-validation and hyperparameter tuning concepts
• Regression metrics including RMSE, MAE and R-squared
• Classification metrics including ROC AUC, precision, recall and confusion matrices
• Comparing baseline, regression and tree-based candidate models

Workshop: Participants create a tidymodels workflow, cross-validate alternative churn models and select a model using agreed performance criteria.

Day 5: Communicating and Deploying Analytical Findings

• Translating analytical questions into stakeholder decision requirements
• Model interpretation, uncertainty statements and limitation reporting
• Variable importance and partial-dependence style interpretation approaches
• Reusable functions and script organisation for maintainable analysis
• Quarto documents, code chunks, parameters and rendered outputs
• Visual storytelling for technical and executive audiences
• Analytical governance, version control practices and peer-review checklists

Workshop: Participants complete and present a Quarto report containing their data pipeline, visualisations, validated model and evidence-based business recommendation.

Tools & standards covered

R, RStudio, tidyverse, tidymodels

A typical training day

Live online deliveries follow the same structure in the East Africa Time zone, with shorter screen blocks and longer breaks.

What the fee includes

How you can take this course

Certification

Participants who complete the full five days receive the Skillset Development Certificate of Completion, stating the course title, course code, dates and delivery format — suitable for professional-development records and employer reimbursement.

Frequently asked questions