Some Other Thoughts on Scientific Thinking

The type of thinking that we do in science is in many respects different from that used in other areas of academics and society. Scientific thinking follows a deceptively simple, intuitive, and yet somehow very difficult general line of reasoning often referred to as "Hypothetico Deducto." Simply put, this principle states that we make a hypothesis, and if that hypothesis is true, the data should follow in a certain way. That is, if it is the case that {hypothesis} is true, then our results should be {data}. Of course, this seems a rather obvious point, but it never ceases to amaze me how commonly people reverse this principle. That is, people will say "These {data} mean that {hypothesis} is true." At face value, these two statements seem to be interchangeable:

1. If {hypothesis} is true, then the results should be {data}

2. These {data} mean that {hypothesis} is true.

These two statements, though, are NOT equivocal. Take for example the statement that "Consumer spending increased this quarter, as did the Gross Domestic Product, therefore the growth of the U.S. economy is due to consumer spending." Certainly, this statement may be true, but the reasoning is not sound. The Gross Domestic Product and consumer spending could as easily have both increased because of a government stimulus that gave a bunch of free cash to everyone in the country, or the two things could have increased for no common reason at all. We could propose any number of alternate hypotheses to explain the data. If instead we say "If economic growth is due to consumer spending, then we should see consumer spending and GDP increase at the same time," we can better rule out alternatives in future studies. For example, we could then ask whether GDP increases with government spending, or we could ask whether consumer spending increases with global temperature, or any number of other alternatives.

What's Next?

• Complete the Quiz 1: Critiquing an Article.