Chapter 10 Good and Bad Graphs PDF

Preview

Summary

Stewart Hamilton...

Description

Chapter 10: Good and Bad Graphs Data Tables: What Makes a Clear Table? ● For large sets of data summarize the information ● Use Easy to understand labels ○ Main heading should describe the general subject ○ Identify variables and units of measurement ● Include date ● The source of the data should be obvious Data Tables ● Distribution of a variable ○ What values a variable takes ○ How often it takes these values ● Rounding ○ Common to see ○ May appear to imply an error ○ For percentages or proportions, the total may sum to a value slightly different than 100% ○ E.g. ■ 1/3 reported as 33% • 3(33%) results in 99% Pie Charts and Bar Graphs ● Pie charts show how a whole is divided into parts ● To make a pie chart: ○ First draw a circle. The circle represents the whole. ○ Wedges within the circle represent the parts, with the angle spanned by each wedge in proportion to the size of that part ● Pie charts force us to see that the parts do make a whole ● Angles are hard to compare: a pie chart may not be a good way to compare the sizes of the various parts of the whole ● A bar graph can make it easier to compare categories Categorical and Quantitative Variables ● Categorical Variable: Data separated into Bins (categories) ○ E.g., Eye color ○ Distribution best displayed with a pie chart or a bar graph ● Quantitative Variable: has numerical values ○ Can perform arithmetic operations (e.g., adding, averaging) ○ Pie charts and bar graphs can be useful for quantitative variables ○ Can also use histograms (Chapter 11) ● Distinguish between variables whose values have a meaningful numerical scale and variables that just place individuals into categories. ● REMEMBER: ○ A pie chart can only compare parts of a whole ○ Bar graphs can compare any quantities Beware the Pictogram ● Bar graphs ○ Compare several quantities ○ Heights of bars represent the quantities

● Eyes react to ○ Height of the bars and ○ Area of the bars ● All bars should have ○ Same width ○ Area varies in proportion to the height ○ Our eyes receive the correct impression ● Tempting to replace the bars with pictures and create a pictogram…however, this can be misleading! Change Over Time: Line Graphs ● Many quantitative variables are measured at intervals over time. To display change over time, use a line graph. ○ A line graph of a variable plots each observation against time ○ Time is on the horizontal scale ○ Variable you are measuring on the vertical scale ○ Connect the data points by lines to display the change over time ● When studying a line graph, what should you look for? ○ Overall pattern ○ Trend: upward or downward movement ○ Deviations from the pattern ○ Seasonal variation: more than 2 years of data Watch Those Scales! ● Because graphs speak so strongly, they can mislead the unwary ● The careful reader of a line graph looks closely at the scales ● Perceptions! ● When examining the change in the price or value of an item over time, plotting the actual increase can be difficult to understand ● Often better to plot the percentage increase from the previous period Making Good Graphs ● Graphs are the most effective way to communicate using data ● A good graph frequently reveals facts about the data that would be difficult or impossible to detect from a table ● Principles for making good graphs: ○ Make sure labels and legends tell what variables are plotted, their units, and the source of the data ○ Make the data stand out. Be sure the actual data, not labels, grids, or background art, catch the viewer’s attention ○ Pay attention to what the eye sees ■ Be careful choosing scales ■ Avoid pictograms ■ Avoid fancy “three-dimensional” effects that confuse the eye ● Only 5 data points make a simple line graph ● Background sketch and gridlines clutter the graph ○ Provide a table if exact numbers needed ○ Use no more ink than needed ○ Decoration is often a distraction and a distortion of the data...