---
title: "The Data Science Workflow"
description: "Useful analysis runs in a loop: state the question, find and check the data, build something, look at what it says, then report it in a form someone can act on. Most of the work is in the first two st"
canonical: https://lightmysky.com/learn/computing/the-data-science-workflow-mt_99QNfiyQ55
source: https://lightmysky.com/learn/computing/the-data-science-workflow-mt_99QNfiyQ55.md
retrieved: 2026-09-12
---

> **Agent view.** This is the Markdown twin of the page, for tools and assistants.
> When to use this site, and the call that answers each job: https://lightmysky.com/agent-instructions.md
> API description (OpenAPI 3.1): https://lightmysky.com/openapi.json · Authentication: https://lightmysky.com/auth.md
> Pricing: https://lightmysky.com/pricing.md · Catalog: https://lightmysky.com/llms.txt · Full catalog: https://lightmysky.com/llms-full.txt
> Every machine-readable file on this domain: https://lightmysky.com/.well-known/ai-catalog.json
> Ask for Markdown with `Accept: text/markdown`, a `.md` address, or `?mode=agent`.

# The Data Science Workflow

Useful analysis runs in a loop: state the question, find and check the data, build something, look at what it says, then report it in a form someone can act on. Most of the work is in the first two stages, and skipping them is what produces confident nonsense.

Subject: Computing · Area: Data & Databases · Ages 21 to 22
Page: https://lightmysky.com/learn/computing/the-data-science-workflow-mt_99QNfiyQ55

## Ready when they can

- Turn a business question into one that the available data could answer
- Say what could make a dataset unable to answer a question at all
- Keep an analysis reproducible so a second person gets the same numbers
- Pick a summary the data can honestly support, such as the median when a single far-off value pulls the mean.

## Lesson: From vague question to honest answer

Pipelines move data to where it can be analysed. Useful analysis then runs in a loop: pose the question, find and check the data, build something, study what it says, then write it up so someone can act. Most of the work sits in the first two stages, and the life cycle names them acquisition and exploration.

**Example.** A vague ask is why customers leave. You turn it into how many new customers never placed a second order, using the order records your nightly pipeline loads with dates and repeat counts. Now the available data can answer it.

Some datasets cannot answer the question at all, like daily money totals cannot say which pastry sold best. Skipping the check is what produces confident nonsense. And you keep every step written so a second person gets the same numbers.

Before you build, write the question and the data check side by side. If the data never captured what the question asks about, no calculation can bring it back.

**Recap.** Ask a question the data can answer, check the data first, and record each step so the numbers reproduce.

## Practice

15 questions on this page, each with its working shown.

## Needs first

- [Sampling Methods and Their Trade-offs](https://lightmysky.com/learn/mathematics/sampling-methods-and-their-trade-offs-mt_aSsRQvnKa6)
- [Errors, Power and the Design of a Test](https://lightmysky.com/learn/mathematics/errors-power-and-the-design-of-a-test-mt_D7ZXbvSD0l)
- [Data Pipelines: Batch, Streaming and Doing It Twice](https://lightmysky.com/learn/computing/data-pipelines-batch-streaming-and-doing-it-twice-mt_o0zBEPLbmF)

## Opens up

- [Cleaning Real Data: Missing, Duplicated and Wrong](https://lightmysky.com/learn/computing/cleaning-real-data-missing-duplicated-and-wrong-mt_qIsLXA-7EC)