Exploratory Data Analysis Step by Step
Explore a new dataset with a clear routine and turn what you notice into testable questions
A taste of a lesson
While exploring I found customers in one small city spend three times the average. Should I report that?
Note it, but do not report it as a finding yet. First check the group size: if it is 12 customers, one big spender can produce that by chance. Look at the raw rows for those customers, and check whether a business account or a data error is in there. Also remember you probably compared many cities; with enough comparisons, one will look extreme. Write it as a hypothesis to confirm with next month's data. How many customers are in that city?
Written by the teacher as an example. In your lesson the tutor answers your own questions, and like any AI it can be wrong.
What you will be able to do
- Follow a repeatable routine to explore any new dataset
- Summarise each variable and spot suspicious values
- Explore relationships with the right plot for each pair of types
- Compare segments and time trends without overreading noise
- Write observations as hypotheses to test later
Lesson plan
- 1 Questions and structure Start with purpose and understand what one row means. Start
- 2 One variable at a time Summarise and plot each variable and flag oddities. Start
- 3 Pairs of variables Explore relationships with suitable plots. Start
- 4 Segments Compare groups fairly. Start
- 5 Trends over time Explore trends, seasonality and breaks over time. Start
- 6 From noticing to hypotheses Record findings as questions to test, not conclusions. Start
Try asking
About this tutor
A beginner tutor that gives you a repeatable routine for getting to know any dataset. You will start from questions, check size and structure, look at every variable on its own, then pairs of variables, segments and time trends, and always keep an eye on the raw rows. Lessons stress writing down what you notice as hypotheses rather than conclusions, because patterns found by browsing need checking. Datasets include sales records, app usage logs and public data. The routine works with any tool, from spreadsheets to notebooks.
Reviews
Students can review a tutor after a paid lesson. Nobody has yet.
About the teacher
Data cleaning, SQL, exploratory analysis and honest charts
9 tutors 427 lessons taught Sample
I teach the part of data science that takes most of the time: getting data into a shape you can trust, querying it, exploring it and showing it honestly. I came to data from operations work, where reports drove real decisions and a wrong join could cost a week. I teach by handing you small, deliberately messy tables and asking...
See Lin's profile and tutorsMore like this
Other tutors on the same or nearby topics.