Skip to content

MODULE 9 · LESSON 53 · Dashboards and storytelling

Writing evidence-based summaries

Write a concise finding that names the evidence, calculation, scope and limitation.

By Laurence Svekis · Self-paced · All worked data is synthetic unless you supply your own source.

Try the practice ↓

PredictInspectCalculateCheckExplainTransfer

Your starting point

Prerequisites: Use the ideas and vocabulary from lesson 52; revisit its worked examples if needed.

Definition: An evidence-based summary connects a claim to specific observations or calculations and keeps interpretation within their scope. A strong summary distinguishes what was measured, what was derived and what remains an explanation to investigate.

Lead with the supported finding

Name the population or subset, period, measure and comparison. Include a useful number and unit rather than a vague adjective such as dramatic. A reader should be able to locate the corresponding calculation or chart. Avoid adding precision that the measurement does not support, and use percentage points or relative percentages accurately.

Separate explanation from evidence

A higher rate is an observation under defined rules; why it is higher may be uncertain. Put potential explanations in a separate sentence and label them accordingly. Do not use causal verbs such as caused or improved when the design supports only association or before-and-after description. A correct calculation does not authorize a broader claim about unmeasured outcomes.

Include the boundary and next step

State the limitation most relevant to the reader’s decision, then identify a proportionate check or action. A concise summary can still be rigorous: one finding sentence, one interpretation or limitation sentence and one next-step sentence often suffice. Keep the detailed method and value table available for readers who want to verify the result.

Two worked examples

A scoped campaign finding

“In the synthetic period, Email recorded 10 converting sessions from 50 sessions, or 20%, compared with Search’s 20 from 200, or 10%. The channels have different audiences and volumes, so this does not isolate a causal effect. A follow-up comparison should align the audience and outcome definition.” Each sentence has a distinct role.

A misleading rewrite

“Email doubled customer satisfaction” is not a valid rewrite of the conversion-rate comparison. Satisfaction was not measured, and the relative rate difference does not establish causation. Keep the measured outcome and the denominator instead of replacing them with a more impressive but unsupported business claim.

Where this becomes useful

A project portfolio should let a reviewer understand the finding quickly while still tracing it to source data. Clear summaries are especially important when AI has helped draft prose that may overstate the evidence.

Make an attempt

Complete a small workbench analysis and write the report’s finding, limitation and recommendation fields. Check every numerical phrase against the plotted-value table or grouped summary. Rewrite any sentence that introduces an unmeasured variable.

Open the accompanying interactive lab →

Misconception check

Claim: A concise summary must omit uncertainty.

Inspect the correction

A short, specific boundary often improves clarity. Brevity should remove clutter, not the conditions that make the claim valid.

Independent application

Write a three-sentence summary for a synthetic project: a scoped finding with numbers, an interpretation boundary and a next check. Have another reader identify the source calculation and flag any word that exceeds the evidence.

Completion records practice and self-review. It is not certification of statistical competence.