Skip to content
NTPU AI Learn

DS 101 Apply

Turn what you learned into your own ESG Data Analysis Use Case

What the Use Case is for

The ESG Data Analysis Use Case is the final artifact of DS 101. Pick a question you actually want to answer with data, then write down the context, the data and sources, the grain and units, the data-quality risks, the metric formula, the analysis steps, what AI may help with, your verification plan, the limitations and the expected output. It becomes the starting point for further implementation discussions.

Info

Human, AI, Human: a person defines the question, AI helps write, explain and debug the code, and a person verifies the data, formulas and conclusions and stays responsible for the result.

What you fill in

  • The problem context, the analysis question and where the result will be used
  • The data and sources needed, and the data grain and units
  • Data-quality risks
  • The metrics to calculate (with numerator, denominator and unit)
  • The analysis steps
  • What AI can help with, and how you will verify (at least three checks)
  • Limitations and unknowns, and the expected output
  • What could happen if the analysis is wrong

A structurally complete Use Case

A structurally complete Use Case has an analysis question that data can test; the data, grain and units stated; at least one metric; the data-quality risks; the analysis steps; a bounded role for AI; a verification plan with at least three checks; and at least one limitation.

Info

The platform does not judge whether your organization should adopt the analysis. The Use Case is for learning feedback and follow-up discussion, and people review it.

Exit Ticket

Finally, the Exit Ticket records the most important idea of the day, the question you most want to answer with data, the step you are least confident about, how you will verify an analysis AI helped with, and the topic you want to explore next.

Before you submit

Warning

Do not enter company secrets, client or personal data, or sensitive information you are not authorized to share. Do not upload confidential institutional or company data to an AI tool without authorization, and check your organization’s policy before using real ESG or internal operational data.

Start

What you need to complete

  • Complete both DS 101 modules and score 80% or higher on the course post-test
  • Complete the Guided Lab, the Data Quality Challenge, and the Independent Lab
  • Submit the Independent Analysis Artifact and have it reviewed by a teacher against the Independent Analysis rubric

What you are assessed on

A teacher scores the Independent Analysis Artifact across six weighted domains: question framing and evidence boundary, data understanding, analysis correctness, verification, communication and limitations, and reproducibility and AI-use record. A total of 80 or above, with Verification at least 2 out of 3, is a PASS; below 65, or triggering any critical-fail rule, is a FAIL.

Info

Publishing this scoring logic is meant to show you what is being assessed: it is not the same as an answer key.

Practical evidence

  • A recorded rubric result (PASS, REVISION_REQUIRED, or FAIL) with no critical-fail rule triggered
  • This result is recorded as competency evidence shared by this course and later DS courses (for example DS 102): this wave does not issue a second certificate for it

What you can earn

The DS 101 Course Completion Certificate: completing both modules, the practice work, and the post-test, with the Independent Analysis Artifact reviewed and passed. It is never issued from the post-test score alone.

What this credential does not mean

This credential confirms the holder completed DS 101's required modules and post-test, and demonstrated question framing, data inspection, analysis correctness, and verification skills in the Independent Analysis Artifact.

  • It does not mean the holder holds a formal data-analysis professional certification or is qualified for employment
  • It does not mean any specific dataset the holder analyzes later will necessarily be correct
  • It does not mean human review before high-impact decisions (Human, AI, Human) can be skipped

What next?