Making a Statistical Poster - Guidelines

(see ISLP guidelines adapted for Irish competition)

What is a statistical poster?

It is a one-page presentation that tells a story about a set of data.

It should:

• be simple and have a logical progression (contain a goal, an approach, main findings and key conclusions)
• include graphs and descriptive summaries of data
• contain commentary on the meaning of the data
• be self-contained (viewers should not need any extra material or information to understand the poster)
• be visually attractive and creative
• be able to be read from a distance of about 2 meters (7 feet) when printed on a single one-sided A1 sheet.

Examples

Here are some links to get a general picture about posters:

http://iase-web.org/islp/Poster_Competition_2014-2015.php?p=Prizewinners

https://iase-web.org/islp/documents/Poster_Competition_2014-2015/ISLP%20poster%20session%20presentation%202015.pdf

Steps in Poster Making

Find a question

First, find an issue or question to study. It should be:

• clearly defined to make it easier to collect relevant data
• be interesting so people want to read the poster
• not too difficult but also not have an obvious answer

Collect Data

Before collecting data plan carefully what measurements etc. are needed.

More about different types of data collection:

http://www.statcan.gc.ca/edu/power-pouvoir/ch2/types/5214777-eng.htm

When using data from other sources, they should be reliable and cited.

Do not forget to consider randomness and numbers of measurements or observations.

Examples:

If the topic is to find out students’ opinions and the study is only made by interviewing two best friends, the results may be severely biased.

If a problem is to measure how far a grasshopper can hop, measuring just one hop from one grasshopper will not give good results. You need a bigger sample. For example, it is better to take 10 grasshoppers and measure 10 hops from each.

More about selection of a sample:

http://www.statcan.gc.ca/edu/power-pouvoir/ch13/sample-echantillon/5214900-eng.htm 

Data quality

Do the data represent a general situation?

Data are variable (which is why we need statistics!) but also may have errors. What are the possible sources of errors in the data?

No matter how well the measurements or observations are made, there may still be some sources of variability or errors. Think about sources of variability or errors and if possible mention them in the poster.

Suggestions on how the study could have been improved can also be given.

Analyze data

Analyze the data collected either by hand or by using a computer program.

Remember that you are investigating the original issue(s) or question(s).

Use numbers and graphs to describe the data: for example, histograms, bar charts, line charts, pie charts and box plots. Statistical quantities like mean, median or standard deviation are also useful.

Practice making a graph:

http://www.statcan.gc.ca/edu/power-pouvoir/ch9/create-creer/5214819-eng.htm

Making the Poster

Electronic Format Only

Posters can be:

• An online one page poster presentation (oriented either vertically or horizontally) i.e. multiple page presentations are not acceptable and posters must be presented in a single page poster format. No paper poster entries are accepted, this is an online poster competition.
• Acceptable file formats include PNG, JPEG, GIF and PDF format only
• Maximum poster file size is 2MB
• The poster should be readable from  a distance of two metres (seven feet) when printed on an A1 sized sheet
• The poster should have no identifying information on it i.e. student names or school names.

Content

Posters should:

• contain at least what was studied and how;
• the main results, discussion about results and the principal conclusions;
• be presented using pictures and key graphs;
• have simple text telling the story of the data;
• include summaries but not all the raw data.

Remember that there is only limited space so do not say things twice.

Pick only the graphs that best present the results. All graphs should be titled and commented on.

Try to avoid using a dark background colour in your poster as it can make some of the text very difficult to read. Sometimes colours and fonts can look different on screen and on paper, so it might be a good idea to print a copy of your poster to see how it looks before you submit it. Also, give consideration to the balance of blank space versus text/graphs; while a certain amount of blank space is needed in the margins etc. so that the poster is not too cluttered, small font sizes and small graphs are difficult to read.

Think about the appropriateness of the graphs that you use. Bar charts are good for comparisons, while line charts work better for trends. Scatter plot charts are good for relationships and distributions, but pie charts should be used only for simple compositions — never for comparisons or distributions. Graphs should help the reader to understand your results so be sure to label your axes and use a legend if required.

Finally, make sure to proof-read your poster before you submit it as typos and spelling mistakes can take from the overall impression of a poster.

Try not to over-complicate your poster by attempting to test or analyse too many things. 

If survey data comes from a particular school, some background on the school should be provided so that the context of the findings may be more transparent.

The possible effects of non-response could be explored.

Exercise caution when reporting your findings. Avoid overstating the inferences/conclusions that can be made from the results – usually the inferences are limited to the sample and probably can’t be extended to the overall population.

It is very important to choose a statistical analysis that is appropriate for the type of data collected.

Try to keep the commentary on the conclusions objective rather than subjective.

Make sure to label the axes of all graphs, label them correctly and try to select a style of graph that best conveys your message. Avoid the use of 3D graphs as they can often detract from the readability of a graph. Choose the graph type that is appropriate for your data, for example, do not use line graphs to summarise categorical data.

In presenting your results, limit the number of decimal places displayed to 2 or less – unless the data specifically needs to be displayed with more decimal places.

Take care with spelling and the overall alignment and formatting of the poster. Try to avoid having too much text and over-filling the poster area. Use a mixture of text, graphs and images, but keep in mind graphs and images have a greater impact than text.

Do not over-complicate the flow of the narrative.  

Make sure all text, graphs and images are legible in the final poster.  

Try to create an eye-catching poster but be careful in choosing the overall colour scheme, 2 – 3 different colours should suffice.  Use an overall colour scheme that is not too garish and hard on the eye.  Avoid backgrounds that are too busy.

Originality and creativity in the research question are key components of the competition.

Avail of the many online guides that can help to improve the overall impact of your poster, e.g. http://hsp.berkeley.edu/sites/default/files/ScientificPosters.pdf