Data Science Foundations and Exploratory Analysis Training Course

Course overview

Teams increasingly hold customer, operational, financial and digital-service data, yet many analysts and subject-matter professionals struggle to turn raw extracts into reliable evidence. Common problems include unclear business questions, inconsistent fields, missing values, misleading averages, untested assumptions and charts that obscure rather than explain variation. This course gives participants a disciplined foundation for examining data before modelling or reporting, so they can identify what the data can support, what it cannot support, and what action decision-makers should consider.

Participants learn a practical end-to-end exploratory analysis workflow using Python, Jupyter Notebook, pandas and Tableau. They frame analytical questions, inspect data structures, profile quality issues, clean and transform tabular data, calculate descriptive statistics, investigate distributions and relationships, segment records, identify outliers, and create purposeful visualisations. The course also addresses sampling, correlation versus causation, bias, reproducibility and the responsible communication of uncertainty. By the end of the week, participants can produce an evidence-based exploratory analysis rather than a collection of disconnected charts.

Instructor demonstrations are followed by guided notebook work, short analytical challenges and a multi-day business case using a realistic dataset. Participants maintain an analysis log, document assumptions and receive feedback on their interpretation and visual design choices. They leave with a completed exploratory data analysis notebook, a Tableau dashboard, a data-quality findings register and a concise stakeholder briefing that can serve as a reusable template for workplace projects.

The course is best suited to professionals who already work with spreadsheets, reports or operational data and now need a structured route into data science practice. It is particularly valuable for analysts, technical specialists and managers who need to commission, review or explain data-led investigations with greater confidence.

Course objectives

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

• Frame business questions as testable analytical questions with defined measures, populations and decisions
• Profile tabular datasets using pandas to identify data types, completeness, uniqueness and validity issues
• Clean and transform data using documented rules for missing values, duplicates, dates, categories and derived fields
• Calculate descriptive statistics and distribution measures to characterise central tendency, spread and skew
• Investigate relationships between variables using grouped analysis, correlation checks and visual comparison
• Detect and evaluate outliers using IQR rules, z-scores and business-context validation
• Build a Tableau dashboard that communicates trends, segments, exceptions and supporting evidence
• Produce a reproducible exploratory analysis notebook with findings, assumptions, limitations and recommended next steps

Benefits of attending

Target competencies

Who should attend

• Data Analysts — who need a repeatable method for turning raw extracts into defensible findings
• Business Intelligence Analysts — who must validate source data before publishing dashboards and performance reports
• Business Analysts — who translate operational questions into measures, data requirements and evidence
• Operations Managers — who need to interpret process data and challenge conclusions drawn from reports
• Digital Product Managers — who use customer and product-usage data to prioritise improvements
• Subject-Matter Experts — who contribute domain judgement to data-quality checks, segmentation and outlier review

Requirements and prerequisites

Participants should be comfortable working with rows, columns, filters and formulas in a spreadsheet, and should understand basic percentages, averages and charts. Some prior exposure to Python is helpful: participants should be able to recognise variables, lists, functions and simple conditional logic, although the course refreshes the Python and pandas commands used. Bring a laptop able to run a current web browser and, where organisational policy permits, install Anaconda or access a managed Python environment. Prior data science, statistics beyond basic descriptive measures, machine learning, SQL or Tableau experience is not required.

Training methodology

The instructor uses short demonstrations to introduce each method, then participants apply it immediately in Jupyter Notebooks against structured business datasets. Exercises progress from data profiling and cleaning to statistical exploration, chart selection and dashboard design. Small-group reviews require participants to compare interpretations, challenge unsupported claims and document assumptions. A running case study provides continuity across the week, while instructor feedback focuses on code readability, analytical judgement and communication. The final session includes application planning, enabling participants to identify a suitable workplace dataset and define their first exploratory analysis deliverable.

Course outline

Day 1: Data science questions and data understanding

• Data science lifecycle and the role of exploratory analysis
• Translating business decisions into analytical questions
• Units of analysis, populations, measures and data dictionaries
• Jupyter Notebook interface, cells, markdown and reproducible narratives
• Python variables, data structures and basic control flow refresher
• Loading CSV and Excel files with pandas
• Initial dataset inspection using head, info, describe and shape

Workshop: Participants inspect a customer-service dataset and produce a question brief, data dictionary review and initial profiling notebook.

Day 2: Data quality, cleaning and transformation

• Data quality dimensions: completeness, validity, consistency, accuracy and timeliness
• Missing-value patterns and documented treatment options
• Duplicate detection and record-level reconciliation
• Standardising text categories, dates, identifiers and units
• Filtering, sorting and boolean conditions in pandas
• Creating derived variables with assign, apply and vectorised operations
• Joining reference tables and validating merge results

Workshop: Participants clean a flawed operational extract and produce a data-quality findings register with documented transformation rules.

Day 3: Descriptive statistics and pattern discovery

• Measures of central tendency: mean, median and mode
• Dispersion measures: range, variance, standard deviation and interquartile range
• Distribution shape, skewness and the impact of extreme values
• Frequency tables, cross-tabulations and percentage calculations
• Grouping and aggregation with pandas groupby and pivot tables
• Sampling, representativeness and denominator checks
• Correlation analysis and correlation-versus-causation limits

Workshop: Participants analyse service-resolution performance by customer segment and produce a statistical findings table with interpretation notes.

Day 4: Visual exploration and dashboard communication

• Selecting charts for comparison, composition, distribution and relationship questions
• Histograms, box plots and scatter plots for exploratory analysis
• Encoding categories, scales, labels and reference lines responsibly
• Detecting outliers with IQR rules and z-scores
• Tableau data connections, dimensions, measures and calculated fields
• Building interactive filters, tooltips and highlight actions in Tableau
• Designing dashboards around a decision question and evidence hierarchy

Workshop: Participants build a Tableau dashboard that highlights segment performance, unusual records and the questions requiring follow-up.

Day 5: Findings, limitations and workplace application

• Structuring an exploratory analysis narrative for non-technical stakeholders
• Distinguishing observations, interpretations, hypotheses and recommendations
• Documenting assumptions, transformations and analytical limitations
• Bias, privacy and responsible use of organisational data
• Notebook organisation, code comments and reproducibility checks
• Peer review of evidence, visualisations and claims
• Scoping a workplace exploratory analysis using CRISP-DM

Workshop: Participants complete and present an exploratory analysis pack comprising a notebook, dashboard, limitations statement and 30-day workplace application plan.

Tools & standards covered

Python, Jupyter Notebook, pandas, Tableau

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