Visualizing Data

By Alexander Tran, WSU Vancouver Neuroscience student

Pictures, graphs, charts and tables are all forms of visual aids but how do we know if our visual aids are conveying what we want the audience to perceive? For example, does your visual aid follow the saying, “a picture is worth a thousand words” or, does it require a caption in order to explain information that is not exactly obvious? In my opinion, it really depends on what you are writing for and who you are trying to appeal to. An article from Scientific America might aim to explain the side effects of opioids to opioid users, while an article from Pharmacology, Biochemistry and Behavior might attempt to present how pain behavior is changed based on chronic exposure to psychostimulants at different points in time. One article focuses specifically on a broad audience, one at this point might as well be the entire population as opioids are so common. The other article focuses on a specific group of scientists that could potentially use it for other scientific means. The audiences that these articles are completely different and as such, should have completely different data visualization styles. In this post, I will take you through visualizing data for a broad non-scientific audience and visualizing data for a scientific audience.

Non-Scientific Audiences:

In this section, I will be focusing on creating visuals based on data for a broader and non-scientific audience. For many new scientists, this is a very tricky and difficult process. Data is extremely important but as much importance it has, the general public may not understand or care for data that they deem unnecessary. Take for example these two different articles, one from a scientific journal (enlarged link here) and one from The Globe and Mail (enlarged link here):

If you were to take a guess, which one would be the one from the scientific article and which one would be from The Globe and Mail? It may be obvious but I’ll tell you.

The image showing the rising death toll of different opioids is from The Globe and Mail while the image showing 4 different graphs is from Neuropsychopharmacology (ACNP). We can see the differences between the two visuals immediately. One is extremely complicated, immediately isolating a non-scientific audience (hence why it is in a scientific journal). The other is much simpler and easier to understand. We will ignore the first visual and focus on the visual from The Globe and Mail. This visual is taken from an article reporting that fentanyl is now the leading cause of opioid death in Ontario. From the title of the article, automatically in my mind, I should expect an image showing that fentanyl-related death is above other opioid-related deaths. Carrie Cockburn did a phenomenal job creating this image, for the authors Karen Howlett and Andrea Woo, detailing the rise of fentanyl-related deaths as they relate to other opioid-related deaths in Ontario. The visual explicitly shows the rise of fentanyl-related deaths, the drop in oxycodone related deaths and where heroin-related deaths stand compared to the other opioids. This visual provides what the article needed, a simple visual that states the main point of the article and supports it. When we create visuals based on data for a broad audience we do not need to swarm them with information that does not relate explicitly to the point at hand. The visual should be straight to the point and should be able to stand on its own. To summarize, visual aids for a broad and non-scientific audience should:

• Be clear and easy to understand
• Be able to stand on its own
• Provides only the necessary data
• Address the main topic/answers the main question

Scientific Audiences:

In this section, I will be focusing on creating visuals for a scientific audience, specifically, creating a poster for a research conference. First, we must address what the basics are in creating a poster. Every research poster should have: a title, authors, department, some form of introduction (abstract, background information or rationale), methods, results, and conclusions. Additionally, a poster should have figures with captions. Optionally, posters may have contact information and a future directions section. Next, what should be in each section and what are some examples?

• Title: The major take home message of the poster
• Authors: The main author, principal investigator (if not the main author), others who have contributed greatly to the project
• Department: The field of the sciences that you/your principal investigator is a part of
• Abstract: The abstract should be a concise explanation of the major points of the whole project
• Introduction: The introduction should provide background information that is not explicitly obvious. An introduction should include a rationale for the project.
• Rationale: A rationale should provide the basis for the project. It should answer the question, why is this project important?
• Methods: The methods section should outline the general procedures on how the data was collected. This section may be combined with the results section. Refer to the first poster example in the next section.
• Results: The results section should present all the significant findings of the project and provide an analysis of the findings without drawing any conclusions.
• Conclusions: The conclusions section should draw all necessary conclusions from the findings and provide some discussion for the findings.
• Future Directions: What are the implications for the findings of the project?
• Contact Information: Any necessary contact information for future contact.

Our first example is a poster is from PhD Posters, a website for printing posters, is an example of a very clean poster (enlarged image here). The designer is Suzanne E. Wardell from Duke University. The first thing to note is that there are vivid colors that draw attention to the poster. It is appealing to the eye. Another thing to note is the amount of text. There is a lot of text present on the poster and it takes away from the appeal of the poster. Sometimes this is difficult to avoid such as in abstract/method sections, however, outside of these sections a poster should aim to condense long paragraphs into shorter bullet points that deliver the same message. This is especially true in the conclusion section. People reading the poster should be able to identify the main finding of the project(s) easily. Large blocks of text are not only boring but intimidating to a passerby. Furthermore, if the findings were to draw attention, it would be difficult for a passerby to find a take home message from the poster. It is important to note that on this poster, the methods and results are combined which was a smart choice by the author. If she had not done this, the poster would have two sections that would be filled with text. This would not be appealing to an audience. Let’s continue and look at another poster.

This poster was created by Kyrie Reyes, an undergraduate student performing research for Barbara A. Sorg at Washington State University Vancouver (enlarged image here). This poster demonstrates a visually appealing poster that can stand alone for an audience that is versed in this particular field of neuroscience. First, let’s examine the two sections that have a lot of text (rationale and methods). The rationale is straight to the point and it explains the importance of the project. This draws interest into project and the findings of the project. The methods are briefly explained to give information about what was done to obtain the data and does not include unnecessary information about the experiments performed. What I would like to draw attention to is the results section. This section draws from the methods section to explain the results of the project. The timeline makes the distinction between the different tests that were performed (e.g. initial preference, CPP test, memory test 1). A person can look at “Initial Preference” on the graph and locate when it was collected in comparison to when a “CPP Test” was collected. This makes the poster flow extremely well producing a poster that is easy to follow. One main problem with this poster is its lack of figure captions. In the first example each figure or graph is followed by a caption that explains the figure. This enables a reader to examine the poster and understand the figures without additional explanation (if a presenter is not present). Figure captions should briefly describe the experiment (if not already done so) and state the main findings of the figure. There should be one figure caption per figure. Overall, both examples are great posters that can be used as a reference. Don’t be afraid to put some creativity into your own poster and make your poster show the amount of effort that you have dedicated to a project. To recap, a poster should have the following aspects:

• Specific sections: title, authors, departments, introduction (abstract/background/rational), methods, results, conclusions. Optionally, contact information, references and future directions
• Figures with figure captions that are brief but informative of the figure.
• Flow: A reader should not be confused as to which sections to read first.
• Creativity: A poster should draw attention but not overwhelm the audience.

A poster should not have:

• Excessive text. Excess text is boring and intimidating. All sections should be concise. The audience should be able to identify the main take home message.
• Empty space: A poster should be completed and should not have excessive empty space that takes away from the poster.