Thread Closed 
Week 11 : An Average Consumer's Spending
11-06-2014, 10:55 PM (This post was last modified: 11-07-2014 08:03 PM by Yuedong Zhang.)
Post: #1
Week 11 : An Average Consumer's Spending
I found this visualization from New York Times and here is the link:
http://www.nytimes.com/interactive/2008/05/03/business/20080403_SPENDING_GRAPHIC.html

Description:
This visualization shows how much the average American spends in different categories. It uses area to encode the share of each category and color to encode the change in prices for each category.
There are 8 big categories and separated with thick lines and each small category thin lines. When the mouse hover on a specific area, it will show the share of spending and the price change. And you can use the zoom in/out button on the top-left corner to zoom in/out to see the details of some small categories. When zooming in, you can drag the mouse to move you view to see different area.

Things I like:
1. I like the idea of using area to encode the share. It looks like a pie chart, but we cannot use angle to encode each category with so many categories. If so, there will be a large amount of category getting small angles that cannot be tell. With area encoding, it uses the space efficiently to make every category can be seen which I think is a smart idea.
2. It uses a diverging color map to show the price change with blue below 0 and red above 0. It is easy for people to tell which categories' prices go up or down.
3. I also like the geometric zoom in/out and interactions like mouse drag in this vis. This makes people easy to get a general view of all categories and also for specific area.

Things I don't like:
It is also about the area encoding. As we all know, area is a less effective encoding method comparing with position or angle. With different shapes, we can hardly compare shares of two categories with similar area. For example, we cannot tell which one is bigger for Food and beverages and Transportation if we don't look at the actual data. But since the topic is about inflation, I think it is reasonable to use more effective encoding like color to show the price change. After all, that is what the topic cares about.
Find all posts by this user
11-07-2014, 10:09 AM (This post was last modified: 11-08-2014 10:20 AM by hehe.)
Post: #2
RE: An Average Consumer's Spending
I agree with YueDong's opinions for the most part. I think this visualization is cool. It uses area encoding strategy to present average American spends from 2007 to 2008. Users could easily tell larger areas make up a larger part of spending. Red and blue color help user to know the percentage of changes. Interactivity in this visualization also facilities user to obtain more details about spendings. For example, by zooming in subcategory of a certain specific area would be presented. Here, I would like to add a scrolling view to move view area by mouse otherwise it is inconvenient if user wants to jump to another zoomed view. Just like Yuedong said area encoding is less effective at some points. For example, Frozen juice and other beef, veal are 0.0% in share of spending but the area of other beef and veal is apparently more than Frozen juice. This may make viewers think this visualization lacks of accuracy. Overall, I like the design of this visualization.
Find all posts by this user
11-08-2014, 04:53 PM
Post: #3
RE: Week 11 : An Average Consumer's Spending
I believe, this is an effective visualization showing the average American spending in different categories. I agree with Yuedong's opinion that the area encoding makes it difficult to compare shares of two categories and hence we might feel that the data representation is not accurate. But, the mouse hover feature provided helps us find out the actual data in each category and hence clear our doubts if any.

Overall, I think this visualization represents the average American spending effectively.
Find all posts by this user
11-08-2014, 07:36 PM
Post: #4
RE: Week 11 : An Average Consumer's Spending
This visualization is effective in conveying the average american spendings and I agree with YueDong that this is with a positive and negative aspect of the visualization. Though the area encoding is a good idea to represent larger spendings for larger shapes, the area of shapes would be misleading without the actual data. Interactivity is simple and provides more information with the mouse over view, but I agree with Hehe that a scroller view would be helpful as we had to zoom out every time we need to focus on other areas in the chart.
Find all posts by this user
11-09-2014, 09:45 PM
Post: #5
RE: Week 11 : An Average Consumer's Spending
There is an amazing amount information shown here, but displayed in a very strange way. The mouse over feature is a great feature, allowing finer resolution of a lot of categories. The colors are easy to distinguish and see how categories changed during the last period. But why these shapes used to separate the categories? I would think that a square style would be much more effective and easy to use for comparisons. Overall, I agree with Yuedong Zhang, some things are implemented well, others not so much.
Find all posts by this user
11-09-2014, 10:17 PM
Post: #6
RE: Week 11 : An Average Consumer's Spending
The visualization has been neatly critiqued by Yuedong and like all the others who have commented on this thread before, I agree that area encoding is both the good and bad aspects of this visualization. But when we consider a tradeoff between the information about the different categories and the details represented, I feel the area encoding works well to bring out a lot of information in this case.

I like the way the designer has shown the summarized percentages of the different categories on which an average American consumer spends, making it slightly easier for us to interpret the values which wouldn't have been obvious just with the area encoding particularly for categories which have same percentages. I also feel that if they had used squares for encoding the information, it wouldn't have been appealing like how it currently is and could not have captured the user's attention.

I also like the description provided below few of the categories, which clearly shows what contributed to the percentages or how an addition or removal of item from a market changed the prices in that particular category. This is quite useful when we are actually trying to understand about inflation.

The color coding for the change in prices is also done very well.

Overall I feel this visualization does an excellent job of describing a lot of information without any cluttering and misleading.
Find all posts by this user
11-09-2014, 11:30 PM
Post: #7
RE: Week 11 : An Average Consumer's Spending
I really loved the designers idea in representing different individual spending of an average household using area encoding.

Pros:
1. This type of encoding enabled the designer to depict almost 200 different categories in one circle.
2. Pop overs used to show the change across different categories between March 2007 to March 2008 provides us great way to monitor the change which took place over the time period .

Cons:
1.As correctly mentioned by Zhang and others. Different shapes may indicate similar values across different categories but would have different shapes and volume . It could get confusing as they would have same color encoding .
2. Also I thought instead of just having percentage on pop over, it would be nice also show if he could provide exact value by which the prices increased or dropped.

However I agree with others , and I feel it is an great visualization design.
Find all posts by this user
11-09-2014, 11:39 PM
Post: #8
RE: Week 11 : An Average Consumer's Spending
The visualiation is well explained by Yuedong Zhang. I agree with almost all the points mentioned. The visualization is clean and it conveys what it needs to. It is quite evident that people spend most of their money in housing. Some of the improvements that I would like to see in this visualization is the option for the user to choose the category he selects and that will form a new bubble. Also a better color coding could had been used and 0 could had been considered as base.
Find all posts by this user
11-09-2014, 11:48 PM
Post: #9
RE: Week 11 : An Average Consumer's Spending
I find this piece of visualization very interesting. Its been well explained by Yuedong Zhang. The use of area to represent the amount spent and color coding used to represent the percentage of increase in price from previous year is good. The graph is clutter free and interactive which is an added advantage. It would had been great if one more feature was added to this. The option for the user to arrange these items in increasing or decreasing manner of the percentage increase or decrease. This will help the user to visualize over the year what prices have come down or have shot up.
Find all posts by this user
11-10-2014, 12:35 AM (This post was last modified: 11-10-2014 12:59 AM by holtvg.)
Post: #10
RE: Week 11 : An Average Consumer's Spending
I think the visualization does a good job of portraying the spending costs of people, magnitude of different categories can easily be compared in the different chunks of the chart which are seperated by clear thick lines. The subcategories are a little confusing and I think they could use a legend to guide the user as some of the small areas are unlabeled. In addition the color could use a more effective coding scheme as some places have color that seems to not be used in any particular effective way.
Find all posts by this user
Thread Closed 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)