Repository mining study, 41,000 pull requests, 14 repositories
Software engineering research poster example
A software engineering research poster has to make a mined dataset readable from two metres away. This A0 landscape board does it for code review: 41,000 pull requests in 14 open source repositories, with the median wait for a first review falling from 31 hours to 22 over three years, shown as a four point timeline across the top.
The whole board
The poster at full size, exactly as it prints. Every number, citation and caption on it was written for this example, so the layout is being judged on real content.
Block by block
What each block on the board is for, in the order a reader walks it.
- Title band and two authors
- The title asks the question as a reader would, how long a pull request waits for a first review, and names 41,000 pull requests and 14 repositories so the scale is set at once.
- Findings timeline
- Four points: 31 hours in 2023 on 12,400 pull requests, 27 hours in 2024 on 13,700, 22 hours in 2025 on 14,900, and the 29% fall from first year to last.
- Abstract and methods
- A short abstract, then five labelled entries: how the 14 repositories were chosen, the yearly pull request counts, what counts as a first review, what was excluded and why medians were used.
- Figure 1, median hours by quarter
- Twelve bars, one per quarter from January 2023 to December 2025, falling from 33 hours to 20. The design draws this figure as bars, so quarters are numbered 1 to 12 and the caption gives the dates.
- Results, with the size table in prose
- This design has no table block, so the three size bands go into the results text: small repositories 14 hours, medium 20 and large 27 in 2025, with their pull request counts.
- Key number and key finding
- The stat block holds 22 hours, the 2025 median. The callout says reviews got faster every year but large repositories still wait almost twice as long as small ones.
- Discussion, key points and references
- Why the fall is gradual, why large repositories lag, and two limits: a first review is not a merge, and survivors are not a random sample. Four checked points and three references close the board.
What makes this board work
The timeline carries the headline
Four large numbers across the top, 31 h, 27 h, 22 h and a 29% fall, give the whole result before a reader reaches a paragraph. Each carries its year and pull request count underneath, so the denominator travels with the number.
Medians, and the reason for them
Review waits have a long tail: a few pull requests sit for months. The methods block says why medians were used, which answers the first question a software audience asks about this kind of data.
The pooled trend is checked against size
A falling pooled median could just mean busy repositories got quieter. The results split 41,000 pull requests into small, medium and large repositories, 7,800, 14,600 and 18,600, and show every band got faster.
Questions people ask
What should a software engineering research poster show first?
The one number your study changes and its denominator. Here that is the median first review time by year, 31, 27 and 22 hours, each with its pull request count. Mining studies often lead with the pipeline, but a reader at a poster session decides in seconds whether to stop, and a pipeline diagram rarely makes them.
How do I describe a mined dataset on a poster?
Give the selection rule, the window, the count and the exclusions in one block. This board says the 14 repositories each merged at least 50 pull requests in every quarter, the window is 2023 to 2025, the total is 41,000, and bot pull requests and quick self closes were removed before counting.
Should I report means or medians for review times?
Medians, and say why on the board. Review times are skewed: most pull requests get a response within a day or two, and a handful wait for months, which drags a mean far above what a typical contributor experiences. This board states the reason in its analysis entry so nobody has to ask.
Can this layout hold a table of repositories?
Not as a table. Only three of the forty poster designs carry a table block, and this timeline layout is not one of them, so the three size bands are written into the results section. If your repositories need a full table, pick one of the table led designs instead and keep the timeline numbers in the abstract.
Why is the quarterly figure drawn as bars?
Each figure block has its chart kind set by the design, and this one is a vertical bar chart. Twelve bars still read as a trend when they fall steadily, as these do from 33 to 20 hours. Short tick labels, 1 to 12, keep the axis clear of the legend, and the caption carries the dates.
How big does a landscape A0 board export?
The working canvas is 2,340 by 1,656 pixels and image export captures it at twice that, so about 4,680 by 3,312 pixels, close to 100 dots per inch at full A0 width. Text stays sharp at reading distance. There is no bleed, crop mark or CMYK option, so a print shop that needs them will add them.
Build your own in about a minute
The button below opens the generator with this use case already described. Change the wording to match your own, generate, then edit anything you like.
Make one like thisOther poster examples
Biomedical engineering research poster example
A biomedical engineering research poster for a device test is an agreement study: how close is the new sensor to the reference, and where does it fail. This A0 landscape board tests a wearable heart rate patch in 40 volunteers, 72,000 paired readings, with a hero line chart of error by activity and the patch within 5 bpm on 96% of exercise readings.
Renewable energy research poster example
A renewable energy research poster earns its place when it tests a number people rely on. This Australian A0 landscape board checks the design model behind rooftop solar quotes against a year of meter data from 220 systems, and finds the model overstates yield by 3.1% overall and by 7% on south facing roofs.
Robotics research poster example
A robotics poster has to show the method as a sequence before it shows a result, because the result only means something once the reader knows what the robot does at each step. This board runs map, plan, detect, replan and execute across the top as a five stage chain, then gives run times at three obstacle densities underneath.
Want the steps in the builder? Read Add charts and diagrams, then choose the template, theme and size. For everything this generator can do, see the poster maker.
Written and checked by the OneCraft team. Last checked .