Shrewd guesses

Cats, dogs, shrews and many other mammals have whiskers. Agreed? But what about boxes? Confused? Don’t be! Just have a look at Figure 6 on p. 28 and you will find an example of a ‘box and whisker graph’. Which are the ‘boxes’ and which are the ‘whiskers’? Even if you can’t make a shrewd guess, hopefully you will agree that there is a very obvious looking difference between the two data sets on the graph. We can only be confident that these apparent differences are real, and are unlikely to have been recorded by chance, if we know the outcome of a statistical test. The researchers who did the experiments did a statistical test — a paired t-test produced a value of t = 2.98. This is a highly significant value, but it isn’t necessary to know that, or indeed anything at all about t-tests, in order to guess from Figure 6 that there is a clear difference between the two sets of values. (If you want to know more about box and whisker graphs, t-tests, or are intrigued about interquartiles, 10th and 90th percentiles, get into the digital surround for this issue.)

One way that scientists show they have done statistical tests that reveal significant differences is by giving letters to data sets. Look at Figure 3 on p. 27. There you will find some values that immediately look very different from the others. A letter that is unique conveys that the value has been tested against the other values using a statistical test, and found to be significantly different. Values that share a letter with another value are not significantly different. But again, even if you didn’t know this, or understand it from the caption, you ought to be able to guess that certain values are really higher or lower than others — and be able to conclude things about the biology behind the values from the other information provided, or what you have learned.