Thread Closed 
Week 9: Most Fuel-Efficient Cars
10-21-2014, 02:23 PM
Post: #1
Week 9: Most Fuel-Efficient Cars
http://www.scientificamerican.com/articl...ient-cars/

[Image: sa1114Gsci31_web.jpg]

This visualization shows the most fuel efficient car companies (either domestic or foreign). It also shows the average MPG (miles per gallon) improvement, or lack thereof, for each company and also for both types of cars: domestic and imports. The first graphic shows the CAFE (Corporate Average Fuel Economy) MPG trend as a line graph, beginning from the year 1978 when the CAFE MPG was 18. For roughly 25 years, the CAFE MPG stayed the same, but due to political pressures, most car companies are complying with federal mandates that enforce new MPG standards. The gray area is the projected average fuel efficiency and the federal mandate. It's interesting to see that by 2015, the average MPG must be 54.5.

The second graphic shows each car company's improvement over time. They small multiples are sorted by the amount of sales per company. Within each view, the red line corresponds to the company's average MPG, while the black line corresponds to the standard that must be met over time. The company name associated with each view is colored according to location: blue is for domestic manufacturers and green is for imported cars.

After studying this visualization for a while I decided that I really liked it. I think the first visualization accurately shows how the trend has started to change in the last 10 years, which I believe to be the author's goal. I can't think of a better way to display that type of data besides using a line chart. It also provides a categorical aspect to the graphic, showing both the CAFE score for domestic and imported vehicles. The only criticism I would have is the extra "chart junk." The bubbles and car graphics may not be necessary, but because it's not printed, I think it might be okay to have them there.

The small multiples view of each companies average MPG was very detailed, but perhaps was a little small to decipher anything meaningful. Yes, it is easy to see what the trend is, but it is not easy to determine what the actual values are. The visualization creator must have only been interested in the change. It did take me a while to figure out that each company has a different CAFE line, which could be slightly misleading. Toyota had the highest increase in average MPG, but how do we know they didn't have a lower CAFE starting point? I think if the actual values were present, instead of the overall change, the views would be much better.
Find all posts by this user
10-24-2014, 01:29 PM
Post: #2
RE: Week 9: Most Fuel-Efficient Cars
I think the critique is through and accurate.

The dataset contains temporal data and are classified into different categories. There is a hierarchy in the category division. At the higher level, the two categories are the CAFE standard and the actual car fleet efficiency. The later is further divided into different categories according to car manufacturers.

The author use line charts to represent the time series and use multiview to organize the different categories.

Since the CAFE standard is in a higher level, it is presented in a single larger view.
In this view, the author use hue to represent different types of standards: light blue for a uniform standard for both domestic and imported cars between 2010 and 2014, dark blue for standards for domestic cars between 2010 and 2014, green for imported cars after 2010, and grey for cars after 2016.
The author also used shape, that is, dashed line, to double encoding the information of "future".
This view contains annotations as well as chart junk. The chart junk is not too much and is quiet appealing in this case.

The categories for different car manufacturers are organized in small multiples. The encoding for the small multiples are the same.
Within each view, two line charts are presented: one for CAFE standard and one for the actual performance for that manufacturer. This is very useful for comparison.
Each view also contains a legend describing the manufacturer and the rank in sale. The legend for manufacturer is color encoded: green for imported and blue for domestic.

I think this visualization does a good job in illustrating the data. One may argue that the usage of small multiples makes it hard to make comparisons between different manufacturer directly. It might be good to put all the manufacturers into a single view. There are 12 different manufacturers and 12 is within the limit of number of hues discernible in a single view. Thus we may put all the manufacturers into one single view and use hue to encode the manufacturers.
Find all posts by this user
10-24-2014, 04:16 PM
Post: #3
RE: Week 9: Most Fuel-Efficient Cars
This is an interesting and informative visualization. While I agree with most of the critique I would probably keep the bubbles in one form or another since it informs the viewer what CAFE is and the context in which it appears.

The use of small multiples broken down by manufacturer is also very informative. I agree that the CAFE per manufacturer is hard to read. I think some of it has to do with the line thickness relative to the height of the chart. Looking closely it appears that they all flatten out to the same value for the majority of the chart. What really bothers me is that they include the top 11 (not 10, not 12, but 11) manufacturers and then suddenly Porsche at 19. Who's at 12? Why Porsche? Very strange.
Find all posts by this user
10-25-2014, 04:00 PM (This post was last modified: 10-25-2014 04:01 PM by Abhishek.)
Post: #4
RE: Week 9: Most Fuel-Efficient Cars
The visualization presents some interesting data about the car fuel economy over the years in a very simple way. The visualization effectively conveys the stall period of 20 years where there was no improvement in MPG and it was stuck around 28 miles per gallon. I like the bubbles where they give the reason for the stall and about the uphill drive where the MPG is expected to improve drastically as most companies will meet new fuel economy standard. One might argue that the bubbles and two cars icon in the visualization is redundant, but i find the bubbles to be very informative and the car adds some context to the line curve and makes it attractive. There should have been mouse hover property on line curve so that we can easily check the exact MPG at any instance.
Its also good to see different car companies fuel efficiency ordered by the sales in the company. Toyota tops the chart while GM is second despite having lower MPG compared with companies lower in sale. This company level comparison is well presented though there is no absolute values.
Find all posts by this user
10-25-2014, 11:12 PM
Post: #5
RE: Week 9: Most Fuel-Efficient Cars
I think the visualization does a good job expressing the data in charts that are clear and detailed to read. For the first visualization on CAFE standards for imports/domestics the pictures of small cars catches the viewers eye to the data being visualized in the form of a line graph, The bubbles up top help detail this quantified data and let the viewer know what is going on with the trend. In the visualization below of it provides detail of some notable import/domestic vehicle mpg efficiencies through a span of 30 years. The CAFE standards and the actual standards can easily be seen and compared through an effective color encoding and legend.
Find all posts by this user
10-26-2014, 11:32 AM
Post: #6
RE: Week 9: Most Fuel-Efficient Cars
I liked the critique especially the part where you argued that small multiples are not useful in comparing data which almost similar.
Instead of small multiples, I felt they could have come up with two separate clusters of car brands which have almost similar data.
By doing this, users can compare between similar data in these cars' mpg in just two graphs which can fit the space which earlier had 16 graphs.
And the chart junk in first main graph ,such as the car driving uphill, is misleading.
I would say the first graph can be omitted. They could have gone with just the two graphs of clusters of data with the CAFE line with all information from the first graph.
Find all posts by this user
10-26-2014, 01:21 PM
Post: #7
RE: Week 9: Most Fuel-Efficient Cars
braden.edmunds presents a good overview to the Car Fleet Efficiency visualization. One thing that is a bit confusing is why there appears a single, solid line (Domestic? Imported?) from 1978 to 2010 at which point the line splits between Domestic & Imports for MPG. Granted the graph does state "Stalled Progress" for that time period, but it is strange to only have the legend come into effect after 2010.

The small multiples for car manufacture, sales rank, & how the standards were met if also a good approach, as braden.edmunds stated. Also along the lines of what braden.edmunds said, I agree that the graphs might be a little small to best approximate the exact values. I am not sure if each company did have a different CAFE standard to be met, but some of the companies might not have existed when the 1978 cafe standards came into creation, which might be a little confusing. Also for the small multiples, it is not clear why the the top 11 companies in sales are present then the jump to Porsche at #19 in sales.
Find all posts by this user
10-26-2014, 02:34 PM
Post: #8
RE: Week 9: Most Fuel-Efficient Cars
Quote:The only criticism I would have is the extra "chart junk." The bubbles and car graphics may not be necessary, but because it's not printed, I think it might be okay to have them there.
I believe in this case the "chart junk" actually improves the readability of the chart enough to warrant it. This is because it naturally draws the eye towards the right hand side of the chart, and gives important explanations of key data points.

Quote:The small multiples view of each companies average MPG was very detailed, but perhaps was a little small to decipher anything meaningful. Yes, it is easy to see what the trend is, but it is not easy to determine what the actual values are.
I believe showing the trend was probably the whole point of the small multiple views. As such, it was only effective in the sense that you could tell how much an individual company has improved over time. (Not which company actually has the best mpg or best mpg increase over time.)

Quote:I think if the actual values were present, instead of the overall change, the views would be much better.
I think in this case a separate interactive line chart would be a better solution. That way you could use a baseline average for the cafe standards, and overlay each company on top of that to show which company actually followed the cafe standard the best.(And/or did better than the standard.)
Find all posts by this user
10-26-2014, 03:29 PM
Post: #9
RE: Week 9: Most Fuel-Efficient Cars
This visualization is very interesting and informative, and I agree with most points in the critique, especially the point that it is misleading that different car firms have different CAFE standard lines. At first glance I did not realize the different CAFE lines of all the firms, but after reading the critique did I notice the difference, and it is truly a little bit misleading.
The critique states that the bubbles and car graphics are extra chart junk and not necessary. I disagree with this point. I think these bubbles and car graphics provide a more attractive and readable visualization to the readers and make the whole visualization more interesting and straightforward.
Find all posts by this user
10-26-2014, 04:56 PM
Post: #10
RE: Week 9: Most Fuel-Efficient Cars
I think the visualization is quite interesting.
It effectively conveys how the fuel economy standards have trended over the years starting from 1973 and also the projected fuel economy standards of the future. In order to convey the fuel economy standard evolution in a more interesting manner, the chart junk has been used. The two decades of stalled CAFE progress is figuratively compared to a car drive along a flat surface(a period of no effort) and the projected CAFE standards are compared with that of an uphill drive. This makes sense since it relates the effort required in meeting new standards with that of the effort required for an uphill car drive.

The small multiple views of the performance of each car company is also nicelyrepresented. I agree with braden.edmunds that the different CAFE standards for each company can be misleading at first and difficult to figure out. Also it takes time to figure out what the positive/negative value at the end of the plot of each company represents.
Find all posts by this user
Thread Closed 


Forum Jump:


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