Science: Cause and Correlation

I was planning on holding off on discussing research and science on this blog.  I felt like it was important to make sure people were interested in what I had to say before I risked overwhelming people with the science of stuff. But after today I just couldn’t keep it in any longer!

Given, my background it’s hard for me to not discuss science and research and given that I do scientific research about child development everyday, it definitely plays a role in what I blog about here.  What’s interesting is that many people don’t really think about what scientific research is or how it is conducted when they read findings in their local newspaper or on their favorite websites.  Most people read a finding like, “TV Causes Autism” and think oh great, I better turn my TV off immediately or else my kid is going to get Autism.  Its unusual for people (myself included) to stop and think about exactly what these findings mean as you read them.

It’s easy to fall into just believing findings and not looking into the details of the studies- especially when they are “finding” scary things like cancer or autism or even language delays. But today, I just want to make a push for looking into the details of studies when you see words like “findings” and “causes” and today I “found” the perfect example for why this is important.

If someone was interested in the research I am doing today and wanted to publish it, I can almost guarantee the headline would read “Books Make Babies Taller”.  Yes, today I was running analyses on some data- I was interested in understanding if there were any relationships between media and health outcomes.  I happened to include the variable “baby’s length” in my analyses and to my surprise, I found a shockingly strong correlation (statistically significant, in fact) that the more books the family had the longer (taller) the baby was.  So like I said, this correlation would likely be reported as “Books Make Babies Taller”, implying a cause and effect relationship.  The more books you buy the taller you child will be. Now, this would be great news for me!  I’m 5’3″ so genetically I’m not doing a whole lot to help my future children become NBA basketball starts, but if owning more books causes babies to be taller, well, then my children will be dunking on Shaq! 🙂

Unfortunately for me, my finding was just a correlation and an incredibly odd one at that.  So what is the difference between a correlation and causation and what types of scientific research can determine causation?  How can we really understand the findings that we read about?  Here is a very quick tutorial:

Correlation:  two or more things that vary together.  If one thing increases so does the other.  Examples of things that are frequently correlated: height & weight (when your child grows 3 inches, usually she gains a few pounds too); grades in school and class attendance (as attendance decreases, grades decrease); crime and outside temperatures (crime increases in the summer when temperatures increase).

Causation: when something causes something to happen- a direct impact.

Correlations exist when there is causation but not necessarily the other way around.  Yes, growing in height can cause you to gain weight, and these things are correlated, but gaining weight doesn’t cause you to grow taller (we all know this one from personal experience, especially after the holidays!).

So how does science determine these differences?  Well, you have to scientifically test  to determine causation whereas with correlations you can just look at how things change over time.  If I wanted to determine if TV caused Autism, I would have to take two identical groups of young kids and require that the only difference between the two groups be whether they watch television or not.  Then, if significantly more children in one group were diagnosed with Autism compared to the other group we could say that TV caused Autism.  However, it is extremely difficult to control for  all the factors that may also play a role in Autism.  In the study that was discussed in the article “TV Causes Autism”, the researchers didn’t even attempt to conduct a scientific study like I described. Rather, they looked at correlations.  They asked parents of children  how much Television their children watched and if their child was diagnosed with Autism .  And yes, they found a correlation between Autism and TV- so that means that kids with Autism were more likely to watch TV than kids without Autism.  But that is not causation! Kids with Autism may be soothed by television, or maybe parents of kids with Autism are more stressed and rely on television to calm their child more than parents of children that do not have Autism- this doesn’t mean that watching TV is what caused Autism.

As with my data from today, it could be that tall children like books and therefore their parents by more books, or it could just be a funny statistical anomaly. There are many explanations for correlations and it’s important when you see a correlation to look at the ways in which it can be interpreted and how it was measured!

Advertisements

2 Comments

Filed under Uncategorized

2 responses to “Science: Cause and Correlation

  1. Amanda

    Taller people tend to be more successful in the workplace, right? And wealthier parents tend to have more books for their kids to read!

  2. Jacob

    even as a scientist I enjoyed reading your explanation!

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