Correlations & Causation: A Norwegian winter study

[sue_frantz]

There must be a cognitive bias that explains why it is so hard for us to learn that correlation does not mean causation. Okay, we can learn it fine. Applying it consistently is the hard part.

Last week here in the Pacific Northwest we had a few days of our typical late fall weather: highs in the low 60s (upper teens Celsius) and rainy. And that’s when I discovered that I’m not ready for late fall. Fortunately for me, it’s early fall, and we’re now looking at a 10-day stretch of mid-70s (low 20s Celsius) and sunny.

This morning, I opened my news feed to see this article: “Thinking like a Norwegian may help you cope with a winter lockdown.” Oslo's average temperature in January is 32 degrees (0 Celsius), and they get about 6 hours of daylight, so I thought, “Perfect! Let’s see what the experimental evidence says. I could use some tips!”

Researchers gave a sample of Norwegians a questionnaire asking about their attitudes about winter, their mental health, and their life satisfaction. Those who had better mental health and higher ratings of life satisfaction had better attitudes about winter. Those who had poorer mental health and higher ratings of life satisfaction had poorer attitudes about winter.

This correlational—not experimental—research doesn’t help me at all.

After covering correlations and experiments in Intro Psych, in a synchronous or asynchronous discussion, provide students with this discussion prompt:

Read “Thinking like a Norwegian may help you cope with a winter lockdown.”

Is this article describing a correlational study or an experimental study? How do you know?

The article suggests that if you change how you think about winter, you will feel better. In other words, they’re saying that how you think about winter will cause you to feel better. Based on this research study alone, is that conclusion warranted? Why or why not?

Monitor the group discussions. If they’re falling into the trap of thinking “it makes sense that how one thinks about winter would affect one’s attitude toward winter,” prompt students with this:

This was a correlational study. They measured attitudes toward winter and measured levels of mental health and levels of life satisfaction. They found positive correlations. When attitudes toward winter were high, mental health was better (and vice versa). When attitudes toward winter were high, life satisfaction was high (and vice versa).

It is possible that those who have better mental health and higher life satisfaction already would see winter more positively. It may not be that attitudes toward winter cause life satisfaction/mental health. It may be that life satisfaction/mental health cause attitudes toward winter.

Or maybe there are third factors. Can you think of any third factors that could cause people to have positive attitudes toward winter and have better mental health and higher life satisfaction?

Given that this research was correlational, what would have been a better article title?

If you’d like to extend this discussion or spin it off into an assignment, give students this prompt.

While correlations cannot tell us which variable is the cause and which is the effect, experimental research can. Here are two hypotheses. Design an experiment that would test each of these hypotheses.

Hypothesis 1: Seeing winter in a more positive way can cause people to feel more satisfied about their lives.

Design an experiment that would test this hypothesis. Be sure to identify the independent variable and dependent variable.

Hypothesis 2: Feeling more positive about our lives can cause us to see winter in a more positive light.

Design an experiment that would test this hypothesis. Be sure to identify the independent variable and dependent variable.