Start Learning Data Analytics with a Better Question

A manager asks, “Why did completion fall?” You can begin useful analytical work before opening a spreadsheet. First, make the question specific enough that an answer would help someone decide what to do.

In this fictional practice example, a learning program reports an 80% completion rate last month and 60% this month. The catalog changed, more learners used phones, and a new group joined. These facts suggest questions. They do not establish the cause of the decline.

Define the number

Write down the numerator and denominator. Does completion mean finishing a lesson, finishing a course, or passing an assessment? Are you dividing by people who registered, people who started, or people whose learning period has ended?

For illustration, 80 completions among 100 starters is 80%; 120 among 200 is 60%. The rate fell even though the count of completions increased. Neither number alone explains whether the learning experience improved or deteriorated. Your first task is to establish which measure answers the manager's question.

Check whether the comparison is fair

Ask whether both groups had the same opportunity to finish. A group that started yesterday cannot fairly be compared with a group that had a month. Also ask whether the course mix changed. A longer activity may have a different completion pattern from a short introductory lesson.

Record the catalog, learner-group and device changes as possible influences to investigate. Do not promote one into the explanation simply because it sounds plausible.

Write a question brief

Try this structure in your own document:

• Decision: what action will the analysis inform?
• Measure: what exactly are we counting, and over which period?
• Comparison: which groups can reasonably be compared?
• Evidence: what information is available, and what is missing?
• Limit: what could this analysis show without proving a cause?

A useful opening question might be: “Among learners with the same time since starting, did completion change within comparable courses?” That is narrower and more testable than “Why did completion fall?”

Explain the next check

Finish with one concrete request. For example: “I need course, start date and completion status under a consistent definition to compare groups with equal follow-up time.” This is a fictional data request, not an instruction to collect real learner records.

Your practice output is a one-page question brief. Review it for vague definitions and unsupported explanations. A reader should understand what you would compare, why it matters and what would remain uncertain.

Explore READY's independent Google Data Analytics-themed training for more practice defining questions and checking evidence. Check the current course overview and lesson list; access to the free learning materials requires a free READY account. READY does not award the Google Data Analytics Professional Certificate and is not affiliated with Google or Coursera.