"A good measure of safety is passenger miles traveled per reported passenger injury (defined here to include fatalities). A higher number is better: It means that a passenger can travel more miles before expecting to face an injury."+ + + +In the blog, the author attempts to show the bad condition of the US passenger train system by comparing passenger miles traveled per reported passenger injury with the worst rated rail systems of European countries. This is done using a time-series line graph that presents miles traveled for the US and the 6 European countries with the worst rated rail systems. + +With the graph, the author aims to demonstrate to the reader the frequency in which injuries occur in the rail system of the US as compared with the six worst rail systems in Europe for each of the years between 2004 and 2012. + +Although the chart seems to passably carry the intended message, it still required a close look and deeper observation to completely understand its intent. The choice of the line chart is one of the reasons for the confusion of the reader. When I first saw the chart, I immediately began to look for year-to-year changes in miles travelled per reported injury, for each of the countries instead of looking to compare the y-values of each country against the US. I believe the message can be better carried with a grouped/stacked bar chart or an area chart. + +Here is an example of a slightly better chart with the data for all the countries in one plot: + +
+
+But this still runs into the problem of too much information in one plot that may distract the reader.
+
+Also, while the author intends to compare the US against the lowest ranked countries in terms of their rail systems, he inadvertently leads us to compare each country with one another by putting everything in one chart. Instead, the author could make a better and stronger statement with six plots that show the numbers for the US vs. each of the other countries, as I will demonstrate below.
+
+Lastly, I must object to the choice of colors and dashed lines in the graph. Even though, the author tried to make the US data distinct by using a solid red line, the dashed lines for the other countries and the colors of the lines make the graph confusing. It's not very easy to identify the countries in the graph.
+
+Although they take up relatively more space, these six charts are better because we can better compare the US's rail system against each of the countries. The contrast between the US and the other countries is clearer and more dramatic.
+
+
+
+
+and alternative visualizations:
+
+
+
\ No newline at end of file
diff --git a/_posts/2015-03-24-shrutiblog.html b/_posts/2015-03-24-shrutiblog.html
new file mode 100644
index 0000000..2dbd68e
--- /dev/null
+++ b/_posts/2015-03-24-shrutiblog.html
@@ -0,0 +1,15 @@
+
+
+
+Thanks for visiting Shruti's Blog!
+ +Shruti Blog! + + + + diff --git a/_posts/2015-03-26-yuxin.md b/_posts/2015-03-26-yuxin.md new file mode 100644 index 0000000..d401e98 --- /dev/null +++ b/_posts/2015-03-26-yuxin.md @@ -0,0 +1,4 @@ +Please check my project page is [here](http://bell-wang.github.io/critique-on-graph). + + +Thank you! diff --git a/assets/alt_bar.png b/assets/alt_bar.png new file mode 100644 index 0000000..748c8b5 Binary files /dev/null and b/assets/alt_bar.png differ diff --git a/assets/alt_line.png b/assets/alt_line.png new file mode 100644 index 0000000..5e4120b Binary files /dev/null and b/assets/alt_line.png differ diff --git a/assets/grouped_all.png b/assets/grouped_all.png new file mode 100644 index 0000000..8cec17a Binary files /dev/null and b/assets/grouped_all.png differ diff --git a/assets/plots.png b/assets/plots.png new file mode 100644 index 0000000..6502a78 Binary files /dev/null and b/assets/plots.png differ diff --git a/assets/source_chart.jpg b/assets/source_chart.jpg new file mode 100644 index 0000000..ca7ec72 Binary files /dev/null and b/assets/source_chart.jpg differ diff --git a/carmem/bad_graph.png b/carmem/bad_graph.png new file mode 100644 index 0000000..31d4f19 Binary files /dev/null and b/carmem/bad_graph.png differ diff --git a/carmem/carmem_graph_critique_v01.Rmd b/carmem/carmem_graph_critique_v01.Rmd new file mode 100644 index 0000000..b67c54d --- /dev/null +++ b/carmem/carmem_graph_critique_v01.Rmd @@ -0,0 +1,43 @@ +--- +title: "Blogpost - Graph Critique" +author: "Carmem Domingues" +date: "March 24, 2015" +output: html_document +--- + + + +Here is a graph from NewsDay.com on employment in Long Island for june 2014 +(http://www.newsday.com/business/long-island-employment-for-june-2014-1.8501596): + + + +###The Good: +- One can read all the different categories of employment +- There is an attempt at color-coding different categories of types of employment +- They include amounts +- There is an attempt at comparing amounts based on the sizes of the rectangles + +###The Bad: +- One cannot easily distinguish between the different sub-categories of Service Jobs +- The quantities included are had to read and compare +- There is no indication of trends - whether the number of jobs in each sector has gone up or down, and no basis of comparison (last year, last month?) +- It is hard to compare the total number of jobs across categories +- The rectangle areas are hard to compare when they are different shapes and in different parts of the graph +- The breakdowns are not consistent among categories - some are in 2 levels, some 3, some more +- There is missing data - rectangles that are blank + +###The Ugly: +- The color coding fails for the sub-categories of Service Jobs, which are broken down on the Legend, but all one single color on the chart +- The fonts are just ugly as are the lines across the graph +- It is all just one heavy box - it's bulky + +