Research poster layout: columns and reading order

A research poster is read standing up, by someone deciding in a few seconds whether to keep reading. The layout’s job is to make the order obvious without anyone having to think about it: where to start, where the answer is, and where to go next. Get that right and a dense board still reads easily. Get it wrong and a beautiful one gets a glance and a nod.

Columns are the reading unit

People read a poster in columns, not across the full width. A line of text that runs a metre across a landscape board is impossible to follow, so every poster layout divides the board into vertical columns that each hold a comfortable line length.

The column count follows the content. Colin Purrington’s poster templates show two common patterns: a horizontal board with columns and a big central column for results, and a vertical board, which he dislikes but provides, again with a wider centre column because charts and tables look poor when shrunk. The principle behind both is the same. The column that carries the evidence gets the most width.

The survey research poster example uses a landscape layout with columns above and below a ribbon of four numbers that runs across the middle. The ribbon is the one block allowed to break the columns, and it earns that by carrying the whole study: 1,204 responses, 61.2 percent allowing weekly remote work, 24.8 percent among the smallest businesses, and 78.5 percent among the largest.

Left to right, top to bottom, with one exception

Western readers start at the top left and work down the first column, then move to the top of the next. Ten Simple Rules for a Good Poster Presentation describes the job as guiding the passer by’s eyes from one succinct frame to another in a logical fashion. The standard order is question, methods, results, discussion, conclusion.

Eyetracking research on web text explains why a poster needs help. Nielsen Norman Group describes four scanning patterns, from the F-pattern, where readers skim the first lines and the left edge and miss most of the rest, to the layer-cake pattern, where they read the headings and skip the body. Poster readers behave the same way. Headings that say something, such as “Participant flow” or “Load against deflection”, let a layer-cake reader follow the board without reading a paragraph.

The exception is the band. A strip of key numbers or a timeline running across the full width is read before the columns, because it sits at eye level and is larger. On the survey board that is deliberate: the numbers are what a reader should leave with even if they read nothing else.

Where the hero figure goes

The main figure belongs where the eye rests after the title, usually in the upper middle of the board and never in a bottom corner. It should be the largest single block after the title.

Size it for distance. Purrington’s guide asks for chart details that can be read comfortably from six feet away and points out that most visitors want to read only the figures. On the survey board, Figure 1 shows four bars climbing from 24.8 to 78.5 percent across the size bands, and its caption prints the band samples, 512, 236, 242 and 214, which add to the 1,204 in the ribbon. A figure that reconciles with the numbers beside it is one a reader can trust without turning to the methods.

The economics research poster example makes a different choice. It is a portrait table led board, so the results table runs the full width under the title, with the grouped bar chart beneath it and a single key number block, 1.5 hours a week, beside the discussion. At a metre you read the estimate, the table and the interval before the method, which is the order an economist checks them in.

Where the conclusion goes

Convention puts the conclusion at the bottom right, the last block in the reading order. That works for a reader who works through the whole board and fails for the one who stops after ten seconds.

The answer is to say the conclusion twice. Put the finding where a passer by sees it first, in the title or a band of numbers, and keep a short conclusion block at the end for the reader who arrives there. The survey poster’s title carries both halves of its finding, six in ten allow remote work but only a quarter of the smallest businesses, and its conclusion points restate them as four bullets at the end.

Landscape and portrait change the order

Orientation changes how many columns fit and where bands can go, which is why the choice is covered separately in landscape versus portrait posters. Two points matter for reading order specifically.

A landscape board can give the study’s identity its own column. The epidemiology poster example stacks the title, authors, abstract and references in a left panel, so the rest of the width holds only evidence: the case definition, a records flow that reconciles from 9,412 notifications to 8,236 confirmed cases, and seven bars ordered by rate.

A portrait board reads more like a page. The clinical trial poster example runs a title band, then abstract and introduction, a labelled protocol, three key numbers with their intervals, a CONSORT spine that reconciles at every step, and three numbered figures. The order follows the trial as it ran.

The table below lists ten of the real A0 layouts behind the poster examples, where the eye starts on each and the kind of study each one suits.

Common mistakes

Columns of unequal importance given equal width. If the results need a chart and a table, give that column more room and take it from background.

A band that says nothing. A strip across the middle is read first. Filling it with a decorative image or a logo row wastes the most valuable position on the board.

Methods written as a paragraph. A methods block is the one most often skipped, and a paragraph makes skipping easy. Labelled entries, design, participants, intervention, measure, analysis, let a reader drop into the one line they care about and return to the reading order without losing their place.

Figures in reading order but captions elsewhere. A caption three blocks away breaks the frame. Keep it attached.

Headings that label rather than inform. “Results” tells a scanning reader nothing. “Anxiety fell 5.7 points” tells them everything.

Reordering by hand until nothing lines up. The builder has drag handles and arrow key nudges but no snapping or guides, so a board rearranged block by block drifts out of alignment. Choosing a layout that already matches your order is quicker. The block counts behind that choice are in what 40 poster templates have in common.

Build it

The scientific poster maker offers forty A0 layouts, twenty in portrait at 1656 by 2340 pixels and twenty in landscape at 2340 by 1656, each authored for its own orientation. Every layout has a logo, a title, an author block, sections and a figure; thirty eight add a methods protocol, thirty four a references block, twenty seven a key number block and ten a CONSORT flow. A chart figure offers ten chart kinds, and the charts and diagrams tutorial shows how to edit the data behind one. Choose the layout by reading order first and colour last.