LabWriteUp is a report writing application for students and researchers in science
|
|
Graphs Summary data can be presented graphically or as a table, as well as textually. If you want to plot the means on a graph, select 'Plot This'. You can display the data as a histogram or as a line plot, with or without error bars:
Figure 1. Plot with error bars showing data from two levels of one IV.
Figure 2. Histogram with error bars showing data from three levels of one IV The text in the graph title, the bar labels, and the axis labels can changed and the graph can be saved as either a bitmap or a jpeg (generally bitmaps have greater clarity but jpeg files are much smaller). When you save the graph it will open in a picture application (usually Microsoft Office Picture Manager if it is installed). If the graph is opened in a picture editing application, such as Microsoft Paint, then aspects of the graph can be modified (such as the fill colour of the bars, and so on). The graph function is particularly useful for plotting interactions from ANOVAs.
Figure 3. Interaction plot as a clustered bar graph. This graph has errors bars and shows the data from a 2 x 2 ANOVA.
Figure 4. Interaction plot as a line graph, showing the data from a 2 x 2 ANOVA. What are error bars? Error bars display the confidence intervals. LabWriteUp displays 95% confidence intervals. So for each mean we can be 95% confident that the population mean is within 1.96 standard deviations from the sample mean, that is, between the upper and lower erros bars.
|
When to choose a line plot and when to choose a histogram: Histograms should be used when the levels of the variable are categorical (nominal), such as sex (men and women), and line graphs can be used when the levels of a variable are on an interval or ratio scale, such as time.