The plural of “anecdote” is not “data”

Students from Flowing Wells Jr High attending UA Science Sky School don't just make anecdotal observations: they take measurements in a variety of environments (leaf litter in burned and unburned sites).
Students from Flowing Wells Jr High attending UA Science Sky School don’t just make anecdotal observations: they take measurements in a variety of environments (leaf litter in burned and unburned sites).

I first heard this phrase from my friend, Brad Boyle, last weekend (though he was cheekily commenting that the plural is data). An anecdote, in case you’re not sure, is a story about something that happened, usually to you or someone you know. It might be something you noticed once, or even a number of times. And that provides you with some information. But that does not necessarily make it data.

Data is information that should be collected and analyzed in a standardized way to answer a question or estimate some value (like the growth rate of a population or the strength of a bridge). One problem with anecdotal information is that it is not collected with the intention of providing a large or unbiased sample. Basing our understanding of the world around us on anecdotal information could lead us to the wrong conclusions, resulting in poor decisions.

For example, many medical procedures have variable outcomes. Imagine you have some inflammation of the elbow (lateral epicondylitis) from playing tennis, rock climbing, or just everyday life. Corticosteroid injections seem to be the most effective way to get a short term reduction in pain and inflammation so that you could use that elbow in the next few weeks. However, the study I linked to reported a 92% success rate at the six week mark. What if you had a big competition coming up and wanted to decide whether to go for the treatment? If you talk to only your aunt, who had had the procedure, and was part of the 8% it did not help, you might decide not to get it. Especially if you heard a second anecdote from your best friend, who was also not helped! There you have multiple anecdotes, but a larger sample might reveal there is a pretty good chance it would help you be able to compete.

Understanding the scientific method is important for everyone, not just scientists. If your vacuum cleaner breaks, to borrow an example from another friend, Marielle Smith, scientific thinking can help you to fix it more efficiently. Instead of replacing parts at random, you might form a hypothesis about what piece failed. That hypothesis might be based on anecdotal information, like what part of your friend’s vacuum cleaner broke. Or it might be based on observations, such as where smoke is emerging from. Either way, you can test your hypothesis by isolating that part of the vacuum to test it, or by replacing it and trying the vacuum, ultimately spending less time and money than replacing parts at random.

The same students from Flowing Wells presenting graphs of their data at the Sky School Symposium. (They even have error bars on their graphs!)
The same students from Flowing Wells presenting graphs of their data at the Sky School Symposium. (They even have error bars on their graphs!)

Helping each other to better understand science and the scientific method is important for all of us, as well. With outbreaks of preventable diseases, like measles, currently happening in the United States, we can see the negative effects that not vaccinating children has on those who are too young or sick to be vaccinated. The data shows that vaccines are safe and measles are not, but some parents are scared anyway. With 2014 the warmest year on record (though one year does not a trend make), we more urgently need political action on climate change than ever, but a large proportion of the US electorate first needs to better understand the risks of unchecked climate change.

So what can YOU do to increase scientific literacy?

1. Practice asking questions.

Question everything. Question sources, and learn who has the expertise to be right about a subject. Just as a theoretical ecologist might not have the skills to remodel a kitchen professionally, a lobbyist with no scientific training may not have the skills to validly contest a research finding. Make up questions while you’re driving (or biking) to work, or waiting in the grocery store check out line. Just be curious about the world.

2. Practice answering questions.

Design a way to collect data to answer them, even if you do not intend to hang out at the grocery store all day collecting that data, or use Fermi estimation to get close. Why? It’s more entertaining than watching the driver singing (or worse) in the car behind you, and as my junior high math teacher always said, it’s like going to the gym for your brain. (If you enjoy this kind of problem,  you will really enjoy the What If? page of xkcd.com.)

3. Learn the difference between causation and correlation (and more about trends, variability, and probability in general).

To use the famous example, both ice cream consumption and murder rates go up during the summer, but that is probably not because eating ice cream makes people want to murder each other. I spend a lot of time these days analyzing data I collected for my dissertation, and discussing probability and analyses with other graduate students. We have to formally analyze data because human brains are not really wired to understand probability. The more you can learn about this, and how to bolster your own ability to understand and communicate about variable results, the better you can interpret the world around you.

4. Interact with scientists and science educators at your local university.

If you live in Tucson, check out the Cosmic Origins lecture series, which is free and open to the public (but crowded – get there early!), or Science Cafe talks at local breweries and restaurants. Visit the Laboratory of Tree Ring Research, see a laser show at Flandrau Science Center and Planetarium, use the telescopes on Mount Lemmon during SkyNights programs or with Sky School, tour Biosphere 2 or the Mirror Lab, or hike up Tumamoc Hill. The Gem and Mineral Show is going on now – find the UA Geosciences students teaching there! And Tucson Festival of Books has a whole Science City where you can see rockets, volcanoes, gila monsters, and more – for free March 14-15. Even better, volunteer for Science City this year (they still need many, many volunteers)!

5. Support science fairs.

Whether you’re a parent who can encourage your kids to do an experiment, or a scientist who can volunteer to judge, get involved! In Tucson, the regional Southern Arizona Research, Science and Engineering Foundation Fair is coming up!

6. Use science to shape how you communicate with people around you.

There are very rational reasons that people may not have all the data or scientific interpretations on a topic, like vaccinations. Going out and doing research takes time and effort, and people have lives to live. Researching every little thing is impossible. So instead, we lump issues together into our identities to decide how we feel about them. Cognitive science has demonstrated that attacking a person’s beliefs, or implying that she or he is an idiot for not knowing something, is the best way to solidify a previously held belief. So when engaging with someone less scientifically literate than you, use what evidence suggests is the best way to get them to change their beliefs: acknowledge their fear or doubt are real and reasonable, acknowledge that data and scientific papers are often not open access, and… ask questions.

One thought on “The plural of “anecdote” is not “data”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s