Tutorial Work

Tutorial 17/08/16 – Analysis of a Data Visualisation

The story this data visualisation tell is; what the heading already suggests,
How many of your health supplements are actually snake oil? The term snake oil means a product, policy, etc. of little real worth or value that is promoted as the solution to a problem, such as the products here that may be promoted for their health benefits. Recent

studies have shown that many vitamins and supplements do little for our health and are a

waste of money. This data visualisation attempts to convey how true this is with many everyday substances and their perceived health benefits. With this both well-known health remedies and the health benefit are compared to others and shown which has more scientific evidence to support their perceived value.

 

How does it tell it?

This image is a “balloon race” and is organized by the weight of evidence. The higher a bubble, the greater the evidence for its effectiveness but the supplements are only effective for the conditions listed inside the bubble. There are sub headings under evidence which ranges from strong to weak to indicate the amount of evidence that proves whether or not the supplement is a snake oil or not.

You might also see multiple bubbles for certain supplements. These is because some supplements affect a range of conditions, but the evidence quality varies from condition to condition. For example, there’s strong evidence that Green Tea is good for cholesterol levels. But evidence for its anti-cancer effects is conflicting. In these cases, we give a supplements another bubble. It uses size of the bubbles to represent either popular interest or scientific interest depending on which is selected. E.g the bigger the bubble the more interest it has to the general viewers. The bubble colour and the usage of gradient also adds to the balloon race, the lighter the colour the less evidence it has to not being a snake oil.

Does it allow for different levels of interrogation that can be seen or used on the part of the reader? eg can they drill down to discover more detail?

This data visualisation does in fact allow different levels of interrogation for viewers with its interactive features. When you hover over one of the bubbles it enlarges it and then you have the option to click on the bubble for a little blurb on how the substance is related to the health condition shown on the bubble. It then goes further and allows the viewer to click again and it links them to a more detailed source of information. This includes credited science websites which have all the details on the relationship of the substance and the health benefit hence explaining why the bubble is where it is on the data visualisation.

There are also filter options which makes searching for specific items easier. You can search for exact substances via health conditions and/or type of the item. The interactive options also allow the user to experience different perspectives in this data visualisation. There is the option to make the information appeal to the general public or those who specialise this this particular topic. This is done by clicking the option to change the weight of the bubbles to float in the order heaviest to lightest. You can click on a button to change the bubbles to float so that it represents ‘popular interest’ and have items that are more likely to appeal to the audience at top. There is also the option to change this to ‘scientific interest’ so that items that are more sought out by scientists and those in a similar field can investigate. There is also a link that leads you to a spreadsheet which has all the scientific notes and links to further studies that gives a clear indication of how the bubbles have been placed strategically in this data vis.

https://docs.google.com/spreadsheets/d/1CBVqSmGwgWzOmSEh35UluqjI3sshObzsB1mML 69wUSQ/edit?hl=en_GB#gid=0
Are you able to create multiple stories from it? If so what are they?

There are limited stories you can create from this data visualisation every aspect adds to the representation of whether each of items are snake oils. But if a filters are used you can gather many stories can be seen. If you filter the bubbles by health conditions for example cancer you can try and make connections between the items that are shown. You can also draw a connection between items that have scientifically interest and popular interest and see what items are considered more by scientists and the general public. Other than that there aren’t many stories that can be gained apart from the data visualisation’s main story.

What can you say about the visual design- layout, colour, typography, visualisation style?

We believe overall the design is a flattering design that appeals to many who set their eyes on it. The usage of the bubbles is a correct usage, because even though our eyes aren’t trained to read surface area well, the bubbles aren’t representing any sort of numbers. The size of the bubble simply reaches out to the audience to guide their eyes to something that is more likely to apply to them. The interactivity is what stands out more than anything. Its gives viewers more options and information to work with. The filter option and the key gives an easier way to navigate and is very user friendly.

There is a lot going on with the page though. When we first laid eyes on this data visualisation, in all honesty we were overwhelmed. There were so many questions, such as, what do the colours mean, what do the difference sizes of the bubbles mean, some of the bubbles only had one word. If you wanted to read this data visualisation I feel like you’d have to be searching for the topic to know what was going on. If a random came across this they wouldn’t know what was happening. The gradient is distracting. The heading needs to be larger.

  • The colours are not aesthetically pleasing. Although the gradient does allow the information to be more easily interpreted, there seems to be a colour clash with the blue- green gradient and the orange bubbles
  • There is too much clutter which makes it very overwhelming to understand. To improve it, we would perhaps space it out a bit more to make use of white space
  • The typography is in sans serif which is good for screen readability
  • The layout itself is in a linear fashion, so it’s not a one page reading. In order to see the rest of the layout, you can’t actually zoom out far enough, but would have to keep scrolling down to the rest of the data. This can especially create confusion when you want to navigate and make better comparisons with the various bubbles


What improvements would you suggest?

Overall, there aren’t too many improvements needed for this visualisation because it is already somewhat easy to interpret, however, there are a few things we believe that can improve the visualisation, for example:

  • Show numbers/ percentages of the results along with the circle to put it into better perspective and create accurate conclusions
  • In order to create accurate conclusions we would use a bar chart, rather than a balloon race, so you can make comparisons more easily
  • If someone needed exact numbers and results of each data set, this visualisation wouldn’t be a good source because it only shows an estimate size, and doesn’t provide the exact qualitative results
  • Although the data visualisation is a pleasing way of representing data, with the use of bubbles, the visualisation doesn’t however allow us to create easy conclusions in terms of the size. In Leon’s lecture on “Data Presentation Styles”, he mentioned that our brains can’t determine the surface area of a shape because we tend to look at the height and the length first. Hence, it’s hard to determine how much the difference in size is for each bubble.


Where does the data come from, and comment on its source.

There is a clear indicator of where the data came from. On the bottom of the page, it mentions that the research was conducted by Stephanie Smith and Miriam Quick and they’ve indicated that their data came from a number of research websites including pubmed.gov cochrane.org and examine.com.

They’ve also indicated that they conducted a meta study and large trials where possible which shows there has been a large breadth of information gained, hence making it more reliable.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s