Data Science · Beginner
Data Science & Analytics
Answer questions with data: frame a decision, clean what arrives from the real world, describe a distribution honestly, and hand over a finding somebody can act on.
Data Science program
- Lessons
- 4
- Study time
- 40m
- Level
- Beginner
- Program
- Data Science
Sign in to open the lessons
4 lessons
Lessons, files and narration open after you sign in and buy the program. One payment opens every course in it.
Sign in to open a lesson. Buying the program opens every course and lesson in it, permanently.
About this course
Most data work is not modelling. It is deciding what question to ask, getting the data, and finding out why it is wrong. This course teaches that loop in the order it happens: frame a decision, pull and clean an export, describe the distribution honestly, then hand over a finding somebody can act on.
You will work with spreadsheets, SQL and a notebook, on data that resembles the data you meet at work — duplicates, missing values, inconsistent names, dates entered as text and numbers that cannot be true. Charts and summary statistics are treated as arguments: chosen to communicate a point, not to decorate a page.
It finishes with a one-page analytical note and the decision it changed, which is what an employer is buying when they hire an analyst.
Who this is for
- You are good with numbers and want the practical workflow that turns data into decisions.
- You already produce reports by hand in operations, sales, finance, health or education.
- You are moving toward a junior analyst role and want evidence on your CV before applying.
Before you start
- Comfortable with spreadsheets and with basic percentages and averages.
- No programming experience needed to start — the notebook lessons assume none.
- About five to six hours a week for two weeks.
Where it leads
- Databases & SQL goes deeper into the query language you used for aggregation.
- AI & Machine Learning is the next step if you want prediction rather than description.
- Analyst work in Ghanaian fintech, telecoms, retail and NGOs starts with exactly the deliverables this course produces.
Tools you will use
What you will build
- 1A framed question with a measurable decision attached
- 2A cleaned dataset with every cleaning step documented
- 3A summary of central tendency and spread that survives a second look
- 4Charts chosen for a specific audience and purpose
- 5A one-page analytical note and the decision it supports
Curriculum
2 modules · 4 lessons · 40m
Sign in to open each lesson
01 · Working with data
The analyst's workflow, and the cleaning work that decides everything after it.
1. The data analyst's workflow
Analysis is not a tool, it is a loop: ask a question precise enough to answer, get the data, clean it, explore it, model it if you must, then communicate a decision. This lesson names each stage and the failure modes inside it, distinguishes the roles people confuse — analyst, scientist, engineer — and covers the kinds of data you will meet, including the messy realities of a Ghanaian organisation: paper records, WhatsApp exports, mobile-money statements and spreadsheets that pass through three people.
2. Cleaning and preparing data
Real data is duplicated, missing, mis-typed, inconsistently named and full of values that are technically valid but impossible. This lesson is the craft of fixing that without introducing new errors: how to profile a dataset, the four kinds of missing data and what to do with each, how to handle duplicates and outliers, why tidy data is worth the reshaping, and how to make the whole process reproducible so that next month's numbers come out of the same steps rather than a fresh act of guesswork.
02 · Insight and decisions
Describing data honestly, then putting it in front of somebody who acts on it.
1. Describing and visualising data
Averages lie, charts lie more convincingly, and both are easy to fix. This lesson covers the summary statistics that describe a distribution honestly — mean, median, standard deviation, percentiles — the shapes that matter (normal, skewed, bimodal), the vocabulary of correlation and why it is not causation, and the chart choices that communicate rather than decorate, including the specific tricks that mislead: truncated axes, cherry-picked periods, dual-axis charts and 3D pies.
2. From analysis to decisions
Analysis only matters when it changes something. This lesson covers the last mile: aggregating with SQL, designing a dashboard people actually open, choosing the few KPIs that matter, and testing a change properly with an A/B test — sample sizes, significance, confidence intervals and the errors that make an experiment decorative. It closes the data science course by connecting the numbers to a decision somebody makes on Monday morning.
4 lessons, including the a one-page analytical note and a tested decision — from a messy export to a copy of it people trust project. Sign in and buy the program to open them. Create an account to track progress and earn the certificate.
Taught by a working engineer
Ama Boateng — Data Engineer. Every lesson is written to be used, not just watched.
A verifiable certificate
Complete every lesson and you get a certificate with a code and QR that an employer can check in seconds.
Practice in the Code Lab
Write and run code in the browser while you learn, with your work saved to your own account.
Still have a question?
Email hello@codemasterghana.com or send a message from the contact page — we answer within two working days, in English or Twi.