Critique By Design

Critique Process:

Identify a publicly available visualization
Critique the visualization
Sketch new designs of the visualization
Gather feedback via interviews and documentation about the sketches
Translate this feedback and sketches into a new design of the visualization on Tableau
Publish this Tableau visualization as the new and improved version of the original

Original Visualization

This visualization involves alcohol consumption for various countries. Each bar graph represents a year, as shown on the right. There is data for multiple years in this graph, which can be accessed by sliding the dot to the right or left to select a year.

I selected this visualization because it was very interesting; not only did it have a large amount of data from different countries, but also different time periods. I immediately knew I wanted to compile information from 2000 to 2024 on alcohol usage and see what the differences were between the USA and other major countries. Additionally, alcohol usage has been the topic of research papers and projects I have done in the past, so this visualization also connects to my past work and gives me a chance to pursue a topic I am interested in while also learning more about the topic. The visualization layout also spoke to me as it was so simple. I wanted to challenge myself by choosing an already simple visualization and trying to make it even better, rather than choosing a cluttered visualization and simplifying it. This was definitely a challenge for me on how to convey so much information with such a large dataset in the way I thought would be most effective.

Link to Visualization

Link to Data

Visualization Critique

Describe your overall observations about the data visualization here. What stood out to you? What did you find worked really well? What didn’t? What, if anything, would you do differently? The first thing that stood out to me when looking at the graph was the title, which is very short and states “Alcohol consumption”. Interpreting this title, I immediately thought that the data would show the amount of alcohol drank per country; immediately next, my eyes travelled to the bottom of the graph, where countries are listed in a row on the x-axis. The next thing I noticed were the many bars on the graph, which are all the same color. I noticed that the bars are organized from greatest on the left to least on the right. The main things that stood out to me on this graph were the title, the x-axis and the large number of same-colored, ordered bars on the graph.

One thing that worked very well was the sorting of the bars. The bars are sorted from highest/greatest on the left and lowest/least on the right. This works well because I can easily know which value is the highest and lowest in terms of location on the graph. Additionally, something else that works well is the simplicity of the graph. Because there is only one color on the bars, a very short title and a large amount of white space, my eyes are able to focus on the title and then the data more easily than if the graph were cluttered.

Unfortunately, there are many things about the visualization that do not work well. These are the elements which I would do differently. The first is the title. Although the title is simple, it does not fully represent the data. “Alcohol consumption” does not give information on the chart or how exactly the consumption is measured (amount drank, sales, trashed bottles, etc.). I would change the title to “Annual Sales of Alcohol in Different Countries (Litres Per Person Age 15+)” and remove the subtitle. The next part of the graph that doesn’t work very well is the coloring. Because the coloring of all bars is the same, the story is not entirely clear. I would color all bars the same except the countries with the highest and lowest values, which I would color red for highest and green for lowest, to indicate that more alcohol consumption is not good. Additionally, not all countries are labeled on the graph, so a reader does not know all countries involved unless they click on individual bars. I would change this so that all countries are listed in a more organized matter so that information is not left out of the graph. Another issue is the measurement across graphs. For each year, the measurements on the Y-graph are different. This makes it very difficult to compare measurements of each year. I would fix this by creating a new graph encompassing multiple years so the reader can see trends of specific countries in their alcohol consumption. The final issue is that the graphs do not tell a story. This data is objective, but the tags are “health”, “improving public health” and “harmful alcohol consumption.” This puts a negative correlation to alcohol, so a reasonable story would be to highlight that it is negative to have a large alcohol intake for countries. I would tell this story by highlighting the higher values in red to draw a negative correlation to high alcohol consumption.

Who is the primary audience for this tool? Do you think this visualization is effective for reaching that audience? Why or why not? The primary audience for this tool is likely health advisory committees which are global and international; the data source is OECD (Organization for Economic Co-Operation) which is known for developing policy and data analysis across a multitude of countries which can be found in the data. The tags on the graph are “Health”, “Improving Public Health”, “Harmful alcohol consumption” and “Directorate for Employment, Labour and Social Affairs.” These tags indicate that the visualization is meant for an audience which has an interest in alcohol negatively affecting public health internationally.

I think the visualization is not reaching this audience. The reason is because despite the negative connotation to alcohol consumption in the tags and its negative impact on public health, there is nothing in the graph that shows that alcohol consumption is negative or impacting the countries in a worsening way to their health. Because of this, it is difficult for someone interested in the public health of their country to look at the graph and know whether or not alcohol consumption has affected their country in a negative way. There should be some correlation of color or indication on the graph for which countries have good (lower) amounts of alcohol consumption and which countries have bad (higher) amounts of alcohol consumption.

Final thoughts: how successful what this method at evaluating the data visualization you selected? Are there measures you feel are missing or not being captured here? What would you change? Provide 1-2 recommendations (color, type of visualization, layout, etc.) This method was very successful in evaluating the data I selected. As I completed this form, I recognized more details of the visual that I hadn’t yet considered and it made me think outside the box in order to evaluate it. Some measures that would be helpful are ranking individual features of the graph, such as color, order, title, etc. so that these individual rankings can prompt us to think harder about how each of these elements could be improved on their own and all together. Additionally, I think that one element of the layout which could be improved would be to include the chart of the 4 rankings (similar to how the Google Sheet was formatted for the previous design assignment) so that we can see where our visualization places amongst the four “boxes.” This would give us a more visual representation of what the visualization is doing well vs. needs to be fixed.

Sketches

Completing the critique made me realize that while I could keep some elements of the original graph, the largest component which was lacking was a clear story. I wanted to tell a story of which countries were highest in alcohol usage and how this could be bad from a healthcare standpoint. I sketched out two designs to potentially use: the first was similar to the bar chart used to represent each year, just with different coloring and labeling to tell a story when showing the countries with the most and least alcohol consumption. The second graph was to provide an alternative to how the data was formatted; instead of individual bar graphs, it would be multiple lines telling the story overtime of specific countries and their alcohol consumptions. I presented both of these as alternatives to a redesign and asked for input on the sketches.

Interviews and Peer Feedback

In my peer feedback on my sketches, I learned that the use of two graphs was a bit confusing for some, and most preferred the graph with the continuation of data over multiple years rather than the individual bar charts for one year at a time. I also learned from my peers in class that although the colors of red, green and grey worked to tell the story, the grey was a bit jarring and provided a dull look to my graph. It was recommended that I keep the color scheme but add a bit more color rather than coloring the other lines/bars grey.

I interviewed two participants on my sketches. Their profiles and questions are below. Participant A: Full-Time Employee, 32 Participant B: Full-Time Student, 19

What is the first thing that stands out to you in the visualization? A: The green and red colors, the red seems like something negative and the green seems like something positive. B: The colors, assuming red and green is high and low, the year stood out in the graph and spikes and dips also stood out in the second graph.

Are there any parts of the visualization that seem unclear? A: Be careful with the bar widths, because it may seem unclear of bars with different widths, but this may just be because it was sketched. B: In the first graph, it is hard to tell what the exact Y-values are in the bar graph, maybe it would help if there were gridlines or written measurements. Also, some acronyms for countries may be unknown by the common reader.

Do you think that the graph conveys its message quickly and effectively? A: Yes, it is organized and draws attention to which countries are the highest and lowest in drinking. I think the line graph depicts this better than the bar graph and gives us the full picture. B: Yes effectively, and fairly quickly because it took a little bit of time to read the country acronyms for the lines on the second graph.

Is there anything about the visualization that seems surprising or confusing? A: The first bar graph seems like it is not needed, and I am more interested in the trends for each country in the second line graph. B: The data points are not as accurate between graphs, likely because this needed to be quickly sketched.

Who do you think is the intended audience of this visualization? A: Health workers. B: Healthcare boards and leadership at alcohol companies.

Do you think two graphs are necessary for the full picture or only one? A: It would depend on the messages being conveyed; one is for one year while the other is across years. Both are relevant if someone wants the big picture rather than just for a year, but I personally prefer the line graph for a larger picture of trends over time. B: The first graph may not be necessary if you want to view more than one year, and it is extremely helpful to view the trends of each country.

What is one thing you like about this visualization? A: It is very to-the-point. B: The colors stand out and indicate positive and negative connotations behind alcohol use easily.

What is one thing you dislike about this visualization: A: There should be slightly more clean-ness in the final design such as the connection of the lines and distance between bars. B: I wish the countries were spelled out because I don’t know all the acronyms.

What is one thing you would change about this visualization? A: The titles should be aligned better and the year should be included in the title rather than directly on the graph to reduce where the reader looks. B: I would make the title clearer rather than including text in multiple places and change the colors a bit to make it less gray.

Findings from Interviews

From the interviews, I realized that a common pattern was choosing the second line graph over the first bar graph alternative. This made it clear along with the peer feedback from class that the right visualization to go with was the line graph in order to properly encompass more data in a way that shows trends over time. I also learned from the interview feedback that a few things were unclear to the viewer, such as the acronym labelling for countries. Therefore it would be more logical to spell out the country names so that viewers are not confused. The title also needed changing, as I had used a title and somewhat of a subtitle and description in my graph, but this amount of text was confusing and led the interviewees to take more time to read it. I would be better off combining this information concisely in a clear title so that the graph can be read quickly. I will also play around with the changes of the color scheme to see if I can reduce the stark contrast in the red and green colors vs the grey colors in the chart.

Final Visualization

For my final visualization, I made sure to do the following:

I created a line graph, with multiple lines, each representing a country and the trends in how many litres of alcohol were consumed per year.
I created a color scheme, going from red to green, with red indicating a higher amount of alcohol consumption and green indicating a lower amount of alcohol consumption. The line which was the darkest red was the highest consumption and the line which was the darkest green was the lowest consumption.
I limited the amount of data used, as the data source has a large amount of data which is not necessarily all important for telling the story I wanted to create. Therefore, I chose the years 2000 to 2024 to show past to current trends, and I chose a few countries which have been at the top and bottom of the alcohol consumption range, and which are popular/known among people in the U.S., as this chart is in English and being presented to an audience of U.S. students.
I changed the title to be more concise and include additional information which I planned originally to put as subtitles, such as the measurement being liters per person 15+. I labeled the X and Y axis clearly so that the viewer could know what the measurements were as well.
I labeled each individual line on the graph to show which country it was; instead of using abbreviations, I made sure to spell out the whole country name so that the viewer knew which country belonged to which line. I placed the countries at the beginning of the lines so that viewers would not have to go all the way to the end of the line to find out which country it belonged to.

Annual Sales of Alcohol from 2000-2024 In Different Countries ( Litres Per Person Age 15+)

Anesha-Santhanam-Portfolio