Domain 6 of 6

Turning Analysis into Decisions

The part that changes the product. Deciding to ship, iterate or abandon on the evidence available, what happens to a measure once a team is judged against it, the questions behavioural data cannot answer at all, writing a finding somebody who was absent can act on, and building a practice that accumulates knowledge rather than dashboards.

5
Concepts
~17%
Of the exam
15
Practice questions
Concepts in this domain
01Reading an experiment resultHow a measured difference becomes a decision to ship, iterate or abandon, what the flat result actually says, the eight result shapes and the decision each one supports, and why shipping a change that did nothing has a cost.02Goodhart's law and metric gamingCharles Goodhart's 1975 observation, the three ways a product measure gets gamed once a team is judged on it, why most gaming is done by people acting in good faith, and the defences that keep a measure honest under pressure.03What analytics cannot answerThe limit behavioural data reaches, which is that it records what happened and never why, what session replay, surveys and interviews each recover from that limit and how each of the three misleads, the population no product measure contains, and the decisions no number settles.04Communicating analytics findingsWhat turns an analysis into a finding somebody who was absent can act on, the four parts of a written result, why a chart with no claim attached settles nothing and what belongs in the record after the decision is made.05Building an experimentation practiceWhat an organisation needs to run tests as routine work, how many tests can share the same traffic, the record that stops a question being asked twice, the arithmetic of experiment velocity and what a one in three success rate means for the way a programme is built.
Try a question from this domain

A crane hire booking product ran two tests. The first reports no significant difference with an interval running from a 0.4 per cent loss to a 0.5 per cent gain. The second reports no significant difference with an interval running from a 9 per cent loss to a 10 per cent gain. A summary slide lists both as flat. What does that slide hide?

  • ANothing that matters, since neither test cleared its threshold and both changes should therefore be removed.
  • BThat the second test is the stronger evidence of the two, since a wider interval has covered more of the possible outcomes.
  • CThat the two support opposite actions. The first has settled the question and the code can be deleted, and the second failed to answer it, so the honest report says the test could not detect an effect rather than saying the change did nothing.
  • DThat neither interval means anything without its p value, since an interval containing zero cannot be read on its own.
15 questions on this domain.

One per page, with a worked explanation.

Start the set