Probabilistic vs Deterministic Thinking
If the weather forecast says there is a 70% chance of rain and it rains: was the forecast “right” or “wrong” or neither?
Many people think and talk about the forecast as being “right” when it does rain, but if it doesn’t rain people think and talk about the forecast as being “wrong” (Tetlock & Gardner, 2015).
But the forecast cannot be right or wrong because it is a probability of a future event happening based on the data that we had at the moment it was made.
This example of how we think and talk about common daily experiences with data-based information I think is illuminating about some of the struggles students have working with data. Let’s explore.
There are two common ways of thinking about information:
-
Deterministic Thinking – For a situation, question, scenario, etc. there is a ”right” and a “wrong” answer. The forecast must be “right” if it rained and “wrong” if it didn’t rain.
-
Probabilistic Thinking – For a situation, question, scenario, etc. we make conclusions based on what is supported or not with existing evidence. The evidence indicates that there was a higher likelihood that it would rain then it not rain, but what actually happened was dependent upon the evidence that went into the forecast when it was made and many other components after the forecast was made.
I think this is relevant to data-based work in a few ways that can be helpful to think about for our students: 1) humans like to think deterministically (and are positively reinforced for that thinking both personally and in school a lot), and 2) we can never make deterministic claims from data. And here in lies the rub. Let’s use an example to explore.
Which plant is taller? —>
There is “an” answer to this question because 1) we can measure the height of each plant (here given in centimeters), and 2) we can use arithmetic to subtract 52 from 80 to get 28.
So we can use deterministic thinking to answer this number fact question.
<— Which group of plants are taller?
There is no one answer to this question. Even though we can measure each plant (like before) we then need to make a decision on how to think about what the “height of the group” (e.g., maximum? minimum? range? average?) for each is and then how we want to compare those “group heights”.
In the second example, we have data (multiple measured values of the variable) and thus we have to use probabilistic thinking to answer the question. Which also means that someone could report a different answer to the same question with the same data (aka they made a different decision on how to calculate the “group height”).
These two kinds of thinking also have ramifications on how we think about scientific conclusions... explain more and connect to science literacy
-
Deterministic Thinking – “more” research/work and/or a lack of “proof/proving” something suggests that there is something wrong with what is already known.
-
Probabilistic Thinking – “more” research/work is always needed to have more evidence to use and we of course we never prove anything because we cannot possibly have all of the information.
So, the question is how can we help our students think more probabilistically rather than deterministically about data? By embracing statistical thinking in all of our work with data!
What does that even mean? Great question! Join us at the next Data Literacy Series: Embrace Statistical Thinking session to find out :)
