Calculating t for a Test on the Slope

Several times in the last two chapters I have focused on the data set showing the relationship between alcohol and tobacco use in Britain, and the influence of extreme data points. In Figure 10.3 I presented SPSS printout showing the slope (unstandardized) and its standard error. Along with the sample size (N), that is all we for calculating t.

The applet CalcT is shown below for the statistics from a hypothetical data set. To perform this test on the alcohol and tobacco use data set, you would enter 0.302 as the slope, 0.0 as the null hypothesis to be tested (the hypothesis that the true slope is 0), and 0.439 as the standard error. You would also enter the sample size as 11. Each time that you enter a number, you must pressed the "Enter" key-don't just move your mouse to another box without pressing Enter.

• Enter the data from Table 10.5 for the data including Northern Ireland.
• Now enter the data from part b) of Table 10.5 with Northern Ireland excluded.
• Notice how the results compare with Table 10.5

Comments to: Gary.McClelland@Colorado.edu

 Return to index