Zach’s facts have been extracted from the book to remind you of the key concepts you and Zach have learned in each chapter.

Zach's Facts 5.1  Presenting data

• The primary objective of a graph is to enhance the reader’s understating of the data. To achieve this:

• Don’t create false impressions of what the data show or hide effects by scaling the y-axis weirdly.
   
• Don’t distract the reader with unnecessary chartjunk: don’t use patterns, 3-D effects, shadows, drawings of space lizards, photos of Milton the cat or anything else.
   
• Avoid excess ink.
   

Zach's Facts 5.2 Graphs

• Bar graphs can be used to present different types of information. They are used to show the frequency of scores (histograms), but the bars can also represent summary statistics such as the mean score for different groups of cases, or the same cases in different situations.
   
• Line graphs are an alternative to bar graphs. They too are often used to display means of groups, or means in different conditions or over time. The means are usually displayed with symbols that are connected by a straight line.
   
• Error bars can be applied to both bar and line graphs when they show means. They are vertical lines that protrude from each mean to show the precision of the mean. One way to indicate precision is to have the error bar show the standard deviation; in other words, the error bar extends 1 standard deviation above and below the mean – for example, if the mean is 5 and the standard deviation is 3, the bar would start at 2 (1 standard deviation below the mean) and stop at  8 (1 standard deviation above the mean). Errors bars can represent things other than the standard deviation, so always note what they show in your figure, and when interpreting graphs check what the error bars show.
   
• Scatterplots show scores on one variable plotted against scores on another variable. They are good for summarizing the relationship between two variables. If the data cloud seems to point upwards, this suggests a positive relationship (as scores increase on one variable they also increase on the other); if the data cloud points downwards, it suggests a negative relationship (as scores increase on one variable they decrease on the other). If the cloud doesn’t seem to slope up or down, it indicates a very small or non-existent relationship between the variables.
   
• Pie charts show the relative frequency of cases falling into different categories. It is usually clearer to present this information as a bar chart. If you must use a pie chart, never ever apply a 3-D effect, otherwise Dr Tuff will collapse in a heap and sob, which is just awkward for everyone.