Online discussion questions – Day 17 (Feb 20) – Avoiding Data Distortion
Please share misleading data visualization you’ve found today in-class with URL. Why is it misleading, and how can we correct it?
15 Responses to Online discussion questions – Day 17 (Feb 20) – Avoiding Data Distortion
Leave a Reply
You must be logged in to post a comment.
http://image.mathcaptain.com/cms/images/41/example-of-misleading-graph.PNG
This graph is misleading because it exaggerates the difference between the two years since the graph’s x-axis numbers don’t start at zero. For the graph to look for accurate the number line on the x-axis would have to start at zero.
http://www.statisticshowto.com/wp-content/uploads/2014/01/Bush_cuts2.png
This graph is misleading because the y-axis does not start at zero; and even though there is only 4.6 percent change, it looks like there 5 times that. To correct it, the news should instead display the bar graphs proportionately to each other and start at zero on the y-axis.
http://image.wistatutor.com/content/feed/u830/misleading%20bar%20graph%205.PNG
This graph is misleading because some years are missing in the x-axis making seems as if hurricanes only increased during the 90s whereas we don’t really know since some years are missing and maybe hurricanes decreased during those missing years. To correct it, all the years should be written down.
http://mediamatters.org/static/images/countyfair/fnc-an-20120220-gasprices.jpg
This graph is misleading because the y-axis does not start at 0, which makes it seem like the gas prices dramatically changed from last year, to last week, to today. In order to correct this, we could start the y-axis at 0 to see the clear relationship between the recent gas prices.
http://www.yale.edu/ynhti/curriculum/images/2008/6/08.06.06.11.jpg
This data is misleading because the graphic sizes are different. It looks like Michael got about 4 times as much candy instead of only 2. To correct this, we can change the graphics to something of the same width or just make them bars.
http://gator.gatewayk12.org/~smcgrail/myweb/powerpoint/misleading_graphs/ma03039.gif
This data visualization is misleading because first of all the axes does not start at zero. This not only results in an uneven y axis but creates a skewed visual of the average house price in 1998 compared to 1999. It makes it appear that the price in 1999 was over twice as much as the price in 1998 when in reality the prices only went up by $2,000. To correct this false representation, we can make a graph with even intervals and start at zero on the y axis. This will create a visual that represents the true prices in 1989 and 1999.
http://wiki.stat.ucla.edu/socr/uploads/thumb/6/6a/EBook_IntroUses_Misleading_Graphs_F2.png/300px-EBook_IntroUses_Misleading_Graphs_F2.png
This graph is misleading because each medal shown does not represent the same amount of medals won. For example, shown visually France has one more medal than Germany but that one extra medal is used to represent 24 medals. While there is only one medal shown visually to represent the difference between USA and Russia but that one extra medal represents 976 medals. The scale is inconsistent making the graph misleading.
http://www.montereyinstitute.org/courses/DevelopmentalMath/COURSE_TEXT2_RESOURCE/U08_L3_T1_text_final_2_files/image009.gif
This bar graph is very misleading. It states that the Central College has the highest average salary. However, we do not know which colleges were used for this graph or how many colleges were surveyed. Also, there is nothing on the y axis. Therefore, we don’t know how much the average salaries range from.
http://blog.visual.ly/wp-content/uploads/2013/09/05-what-not-to-do-618×308.png
This data visualization is misleading because it doesn’t add up to 100%, and the thickness of color that is supposed to represent a certain percentage is not porportionate to the corresponding percentage or compared to the other percentages. For example, the Direct mail (87%) is represented by a color block that is almost twice as big as Special Events (88%) even though it is less. This is misleading and does not show an accurate breakdown of the percentages within this data visualization. To fix this, one should make sure the percentages are represented by an appropriate color block thickness so one can analyze the data correctly. Also the percentage should add up to 100% to make sense.
Image URL: https://technaverbascripta.files.wordpress.com/2014/04/y-axis3.png
This data visualization is misleading because the pie chart is cut into thirds, but the percentages are not equal. In addition, the percentages equal 193% rather than 100%. Lastly, the source for this graph is “opinions”. This graph does not seem factual at all.
http://www.statisticshowto.com/misleading-graphs/
The graph I found misleading was the one titled, “Those d%$n Liberals!” This is misleading because the graph does not start at 0 but rather at 50 which makes the gap between the Democrats and Republicans a lot larger than it actually is.
http://www.narragansett.k12.ri.us/resources/necap%20support/gle_support/Math/resources_data/ident_rep_elem.htm
This graph width is misleading because it causes the bars be unproportional. the best way to fix it is to make them bars and not pictures up cans. these bars should have the same width to help with the proportions.
https://s3.amazonaws.com/heapdatablog/misleading1_yaxis.png
This graph one left side is misleading because the Y axis does not state at zero. The other graph is better which has the right Y axis.
http://www.quora.com/What-are-examples-of-bad-data-visualization-thats-misleading-and-confusing
In this data visualiztion, there is a pie chart that is divded into 100 sections therefore making the data confusing to measure and understand because there are so many variables.
This website shows different company and graph which were posted and lead to misleading information. It shows a graph that was misplotted which caused errors in calculating the incline of unemployment.
http://www.statisticshowto.com/misleading-graphs/