32 Research Posters
This chapter is part of a book in active development and has not yet been through the author’s review. Content may change as the review advances.
The research decision. If your artifact carries a quantitative visual, decide what that visual must let the audience see and what uncertainty must stay visible. If your venue is a poster, also decide how the figure participates in the page’s scan path. You own the headline verb, and you own the call that the picture never says more than the data.
32.1 Why this decision matters
The decision on the table: whether your headline figure’s picture says exactly what its numbers say, and nothing more.
“From four feet away I read one thing: does your tallest bar mean what it looks like it means? If your axis starts halfway up the scale, I have caught you before I read a word of your methods, and now I distrust the whole board.” — a skeptical peer walking a poster session
A journal reader can study a figure for minutes. A seminar audience may see it for thirty seconds. A poster visitor may first meet it from four feet away. The venue changes the viewing conditions; it never excuses an unlabeled axis, a hidden denominator, an inaccessible color choice, or uncertainty detached from the claim. At a poster session a stranger gives you ninety seconds and one glance. The figure does most of the talking, and a figure can lie with true numbers. An axis that starts just below your data turns a rounding-error gap into a landslide. A reader who cannot tell your red bar from your green one receives no message at all. This chapter is the decision that keeps your strongest visual honest before it ever reaches the wall.
32.2 The concept
The decision is one judgment made four ways, and each way has a name.
- Headline claim. The one sentence you want a reader to leave with (Erren and Bourne 2007). Its verb carries the whole meaning. Example: “Version B raised our sales” promises far more than “Version B sold slightly more.”
- Claim boundary. The line between what your evidence licenses and what it does not. Example: an observational comparison may say “is associated with,” never “causes.”
- Compass position. The kind and reach of the question your project answered, which fixes what the headline is allowed to say. Example: a descriptive question about the weeks you measured, not a causal claim about why.
- Figure honesty. The rule that a figure’s picture cannot say more than its data. Example: if the real gap is three points, the bars should look about three points apart, not twice as tall.
- Truncated axis. An axis that starts above the bottom of its scale instead of at the floor. Example: a completion-rate axis running 70 to 76 instead of 0 to 100 magnifies a small difference into a cliff.
- Uncertainty on the page. The interval or caveat printed in the same glance as the claim, not buried in a footnote (Alley 2013). Example: a “±4 points” error bar drawn right on the headline figure.
- Accessibility. Whether your message reaches readers who perceive differently. For the web version of a poster, WCAG 2.2 is the published standard: color is never the only visual means of carrying information, and every informative image needs a text alternative that serves the same purpose (World Wide Web Consortium 2024). Its normative scope is web content, so treat the same two rules as good practice for a printed board rather than a formal requirement. Example: a red-green bar pair carries nothing for a red-green color-blind reader, roughly 1 in 12 men (Crameri et al. 2020).
- Redundant encoding. Carrying the same distinction in more than one channel, so no single channel is load-bearing. Example: label each bar with its value and give each a pattern, so color becomes a bonus, not the only signal.
A poster that fails any one of these overclaims, even when its numbers are exactly right.
32.3 A worked example
You are presenting a business-analytics project comparing two versions of an online store’s checkout page. The outcome is the completion rate, the share of shoppers who start checkout and finish the purchase. Across weekly samples, Version A gives a mean completion rate of 71 percent and Version B a mean of 74. The 95% interval on each mean is about ±4 percentage points, so the two intervals overlap. Your first draft draws a bar chart with the y-axis running 70 to 76, titles it “Version B BOOSTS completion,” and colors the bars red and green.
Now run the decision. Figure honesty: the real gap is 3 points, but on an axis floored at 70 the taller bar is drawn four times the height of the shorter one, so the picture claims a landslide the data never delivered. Uncertainty on the page: the intervals overlap, so a difference the figure sells as decisive might be noise, and nothing on the board shows it. Claim boundary: “BOOSTS” is a decisive, causal verb for a three-point, overlapping gap. Accessibility: told apart by red and green alone, the bars carry no information for a color-blind reader, and a screen reader reads no value off the image.
The fixed figure keeps every number and changes only the honesty. You run the axis 0 to 100, print each mean as a label on its bar, draw the ±4-point interval on top, separate the bars with a color-blind-safe pair plus a hatch pattern, and soften the headline to “Version B completed about 3 points higher, within overlapping intervals.” Same data, a picture that now tells the truth.
Designing so the message survives a color-blind reader and a grayscale print is a documented requirement, not a courtesy (Crameri et al. 2020).
The block below computes both means with their intervals and draws the honest version of the figure: a zero baseline, the intervals visible, and a pattern so the bars survive a color-blind reader and a grayscale printer.
import numpy as np, pandas as pd
import matplotlib.pyplot as plt
SEED = 464
rng = np.random.default_rng(SEED)
weeks = 12
a = rng.normal(0.68, 0.070, size=weeks)
b = np.random.default_rng(SEED + 1).normal(0.705, 0.070, size=weeks)
means = np.array([a.mean(), b.mean()]) * 100
ci = np.array([1.96 * v.std(ddof=1) / np.sqrt(weeks) for v in (a, b)]) * 100
print(f"Version A: {means[0]:.0f}% ± {ci[0]:.0f} "
f"Version B: {means[1]:.0f}% ± {ci[1]:.0f} gap {means[1]-means[0]:.0f} pts")
print(f"intervals overlap: {means[0] + ci[0] > means[1] - ci[1]}")
# the honest figure: zero baseline, intervals drawn, pattern not colour alone
fig, ax = plt.subplots(figsize=(4.2, 3.2))
ax.bar(["Version A", "Version B"], means, yerr=ci, capsize=6,
color=["#4a4a4a", "#a8a8a8"], hatch=["", "//"], edgecolor="black")
ax.set_ylim(0, 100)
ax.set_ylabel("completion rate (%)")
ax.set_title("Completion rate by checkout version")
for i, (m, e) in enumerate(zip(means, ci)):
ax.text(i, m + e + 3, f"{m:.0f}% ± {e:.0f}", ha="center", fontsize=9)
plt.tight_layout()
plt.show()One note on how far this chapter travels. If your venue wants a slide deck, a brief, or any format that carries a quantitative figure, the figure rules you just worked (honest axes, values on the marks, uncertainty shown where the design licenses it — an interval or an honest caveat — and distinctions in two channels) travel with the figure. The poster’s scan path and its print lock do not; they belong to this format. Borrow by naming the technique, never by claiming the whole poster transfers.
32.4 An AI failure case
You paste the described figure and ask your AI for the hardest reviewer questions. It returns a confident, well-organized list of six: sample size, how many weeks were sampled, generalization to other product pages, seasonal traffic, and two more. Every question is reasonable, and not one mentions the truncated axis, the single worst flaw on the board. The list looks complete, so it is tempting to trust it and move on.
You catch it by mapping each returned question to an audit. Claim boundary, sampling, and methods all draw questions; figure honesty draws none. That empty row is the finding. You check the number your own cell printed, the exaggeration the floored axis creates, and you add the figure-honesty hit the tidy list skipped. Fluency is not completeness. The check lives in your own audit list, the one you wrote before you asked.
32.5 It is your turn
You are working inside Studio 10: Prepare to publish or present. This lesson adapts your work to the poster format. Work the format your venue actually asks for; the studio page routes you.
Your claims now have rows and traces; this step turns them into the one page, or the one-figure summary, that a stranger meets first.
The hands-on half of this section lives in the chapter’s companion notebook: open it in Colab with the badge at the top, and work the steps there.
Commit your own audit first, then delegate. Each prompt below is a checkable job, not a request for a verdict.
Run these as a loop. First answer, then your objection, then the sharper answer. The second and third rounds are usually where the useful criticism shows up, because a first pass tends to be generic. If your tool runs its own rounds before replying, you get the polish without seeing the drafts, so your own audit list matters more, not less.
The words of your headline and the verb that carries its boundary are yours. So is the judgment that the figure’s impression matches its number, the call that the uncertainty is visible in the same glance, and the decision that the page is honest enough to lock. AI can generate the skeptic’s questions and scan your code for color-only channels. It cannot see your board, and it cannot decide the claim was earned. Those stay with you.
Choose the single figure that carries your lead claim, and draw it with the axis running the full scale of the measure.
Print each value as a label on the mark it belongs to, and show the uncertainty the design licenses right on the figure so the uncertainty arrives in the same glance as the number.
Carry every meaningful distinction in at least two channels, so color is never the only signal a reader can lose.
Locate the color-only channels.
Here is the code that draws my poster's headline figure: [paste the cell]. List every place this figure encodes a meaningful distinction using color. For each, tell me exactly what a red-green color-blind reader and a screen-reader user would fail to receive. Do not rewrite the figure.After running, verify: render your figure and check each named channel against the picture you can see. Counters plausible-but-wrong-method (generic accessibility advice that does not match how your figure is actually drawn).
Write the headline above the figure using the verb your claim boundary allows, and lay out the rest of the page so a stranger finds that headline in under a minute.
List the skeptic’s questions, then map them to audits.
Act as a hostile poster reviewer. My headline figure is a bar chart of mean checkout completion rate for two page versions, y-axis 70 to 76, red versus green bars, no interval shown, titled "Version B BOOSTS completion." Generate the six hardest questions you would ask, then name the one this figure cannot answer at all.After running, verify: map each question to one audit (claim boundary, figure honesty, uncertainty, accessibility); an audit with no question is the gap, and the truncated axis had better appear. Counters illusion of completeness (a long, tidy list that quietly skips the flaw that matters most).
A second angle, optional:
Red-team the headline verb.
Here is my headline: "[paste it]" and my compass position: "[descriptive / associational / causal]". Act as a skeptical reviewer and find every word that claims more than that position licenses. List the exact words. Do not soften them.After running, verify: if it praises the headline or finds nothing, push back and demand the single worst word, then check each flagged word against your interval. Counters sycophantic agreement (praise that reviews your nerve, not your evidence).
Log it in your AI Research Ledger, and verify at least one number on the page with a named method from the Verification Guide. An AI reviewer may run the check with you; the decision to accept or reject stays yours.