Add margin-of-error ranges to poll estimates

Show the published margin of error directly on each poll estimate. For example, add a plus/minus range around a bar or value marker and state the margin on the chart instead of presenting one percentage as an exact result.

Uncertainty ranges keep sample-based poll estimates from being read like final election results. They make it visible when an apparent lead may still fall within the plausible range of error.

Mechanism: Showing a range changes the reading task from treating a poll as a settled outcome to judging whether differences are meaningful within sampling error.

Evidence: The post says election polls are extrapolated from samples of about 1,000 to 3,000 respondents and usually carry a margin of error of roughly plus or minus two to three percentage points, and that ignoring these margins is misleading, especially in tight races. (Jockers, 2021)

• User Goal: Report current standing from a poll without overstating certainty.
• Task: Compare close poll values.
• Data: Sample-based poll estimates with a published margin of error.
• Chart Setting: A single poll graphic at one timepoint with one reported percentage per candidate or party.
• Audience: Readers deciding how much confidence to place in a close race.
• Success Criterion: Readers can see both the reported estimate and the plausible range around it.

Break it when: The graphic shows actual election results rather than sample-based poll estimates. Why: The rule is for sampled estimates with statistical error, not realized vote counts.

Sacrifice: You give up some exact-looking simplicity. Risk: Readers must interpret a range instead of a single number. Mitigation: State the plus/minus value directly on the chart so the range is easy to read.

Mistake: Reporting the poll percentage alone and leaving the margin of error out of the chart. Why it fails: Readers still see the estimate as if it were a precise result.

Failure Sign: Close poll values appear as exact percentages with no visible range. Quick Check: Look at each plotted estimate and verify that the chart shows an upper and lower bound, not just one number. Stronger Test: Compare the displayed range against the published plus/minus margin and confirm they match.

• Add a visible range around each poll estimate using the published plus/minus margin.
• State the margin of error on the chart, not only in surrounding text.
• Redraw any exact-looking single value so the estimate and its uncertainty are shown together.