Category: Uncategorized

As the exercise is designed to illustrate both of these charts are the same data set. At first glance though, they appear to be polar opposites. In an era of flash media, and “shocking” statistics, simply changing the order the data is presented in could easily sway a reader to look at it differently.

I played with the ‘reverse’ tool on the x axis. So, although the data is the same as chart 1 it appears as if percent is going down by flipping the years. I am unsure if this a statistically correct way to present years on a line graph – but, something tells me that those who ‘lie’ with stats. are not always following all the “rules”. In addition by allowing the y-axis to illustrate a larger span of percentages in the second chart the line does not appear to be drastic by starting at the bottom of a chart and raising. instead it appears to illustrate less drastic change by remaining in the middle of the chart.

Jack, asked in class if we recognized anything about his racial change chart that may have been done on purpose to present a viewer with “guided” interpretation of the results. What I noticed after further examining the chart is:

1. the percentage values associated with a color in the middle of the data from 90% white residence to 10% white residence are in 15% increments. So, for a color to change with in those figures it would have to be a substantial difference of a 15% racial population change.

2. In contrast – the highest and lowest percentages have color changes on the map at just 2% racial population change, 100% to 98% white reflects a color change, as well as 0% white population to 2% white population results in color change.

3. Data sets between 90% to 98% white population and 2% to 10% white population are alone in a reflecting a color change at an 8% racial population shift.

I can only speculate why this was done, but I would not have noticed it had I not been told to look closely at the map.

Here is my practice post!

In chart 1, we show the full range from 0-100%, which makes the change in percentage of the data seem smaller. In chart 2, the scale ranges only from 40-80%, which makes the full change in percentage seem larger. Even though the percent change is the same in the second chart, the change appears to be much more dramatic (larger) because of the scale of the chart. One can be mislead if they are not checking the scale values carefully.

In this exercise, I showed how the same set of data could be depicted in two different ways. While both graphs provide the exact same data, the first graph looks like it has a more dramatic increase than the second one. By changing the scale on the Y-axis I was able to make the slope look very different on both graphs. It is extremely important to look at the scales on a graph because it is very easy to be fooled by data. Creators of graphs can use these manipulation tactics to help convey a message they want to send out about their data, even if it is not truly representative.

Extreme One – Minimizing Differences

Extreme Two – Maximizing Differences

Explanation
It is clear from this exercise that it is rather simple to manipulate chart settings in order to make the data appear to fit whatever argument is being made. If I were writing a paper in which I wanted to show that the percent did not change over time, I would use the settings in the first graph, in which I chose an axis minimum far below and a maximum far above the actual data in order to “scrunch” the data into a small space. If I were making the opposite argument, that in fact huge changes occurred over time, I would use the settings from the second graph. Setting an axis minimum and maximum close to the data minimum and maximum spreads out the data and emphasizes differences. The reason this manipulation is so powerful is that, generally, readers are lazy. We expect graphs and charts to demonstrate a point in the quickest way possible, so it is easy to overlook details like how the axes are spaced. The human brain also looks for patterns, order, and continuity, so it is even easier to get away with a misleading graph if the title or an accompanying explanation have already told readers what they should expect to see. What is interesting to me is whether we classify this type of manipulation as lying or simply as making an effective argument.