2.2 Science is not like sausage-making

As you have seen, there are difficulties and, perhaps surprisingly, subjective judgements in deciding:

• whether a weather event is unexpected
• whether past and present climate distributions are different
• whether measured and simulated climate distributions are different.

You might be tempted to think of the scientific method and statistical analysis as objective recipes or methods to follow: put the data in the top, crank the handle and out fall neat ‘sausages’ of results (Figure 5). But it’s not as simple as that.

Figure 5 Science is not like sausage-making, with sausage meat (data) going in and neat sausages (results) automatically coming out. Instead, there are many subjective judgements to be made about data and how it is interpreted that influence the outcomes.

Show description|Hide description

This figure is the drawing of a manual sausage machine, with data forced in through a funnel like mincemeat and a handle to turn to force the meat out as sausages (finished results). Instead, there are many subjective judgements to be made about data and how it is interpreted that influence the outcomes.

Figure 5 Science is not like sausage-making, with sausage meat (data) going in and neat sausages (results) automatically coming out. ...

As you have seen, there may be more than one type of machine.

Statistical analysis is permeated with subjectivity because it involves making choices – assumptions, decisions and judgements – for which more than one option may be justifiable, even though the results may differ.

As such, these choices can be challenged – for positive or negative purposes.