Correlation = Causation?

Okay. After trying to put off blogging for as long as I could I’ve finally gathered the courage to at least ATTEMPT to write something understandable; statistics related and (hopefully) short. – I cannot guarantee that you will not die of boredom, sorry guys.

So, let’s start.

Correlation in psychological terms or scientific terms means that there is a relationship between two variables.

There are 2 types of relationships: positive (in which when one variable increases the other variable also increase) and negative (in which when one variable increases the other variable decreases).

An example of a positive correlation may be.. high sales of ice-cream is positively correlated with deaths by drowning.

And an example of negative correlation may be.. The older the person is negatively linked to donation making.

What is important to point out is that correlations are looking to see if there is a relationship. It’s an observation. A collection of data they haven’t manipulated. These variables are called co-variables.


Causation on the other hand is saying that one variable directly changed the other variable without the interference of extraneous variables. It’s saying that the Independent variable directly manipulated the Dependent variable. This isn’t just an observation anymore, it’s a manipulation that resulted in an effect.

So therefore, if a study published the findings that high sales of ice-cream positively correlated with death by drowning.. it does not mean high sales of ice-cream LED TO an increase in death by drowning or vice versa. All the correlation can tell us is that it COULD BE that one variable led to a change in the other variable or vice versa OR that the relationship is actually due to a third variable that effects the co-variables that were not accounted for (such as visits to the seaside during summer).


Okay so now that the difference is established. What does Correlation mean for Causation then?

Well, correlation can be quite useful. As a correlation is merely an observation of data to see if there is a relationship between two variables.. If a correlation is found this would give you a reason to see if the correlation found is in fact a causation!


So to conclude, correlation does not mean causation. It is a relationship. But it could be that the relationship COULD be due to causation. But in order to test that you would have to control third party variables.

I know this post has been boring. And I’ve overused the term variable way too much. But yeah, thanks for reading. And don’t shy away from buying and eating ice-cream during summer.

I promise to try and put more effort into my next blog post. =/


9 thoughts on “Correlation = Causation?

  1. Wonderful explanation of the two term that i am always getting confuzzled.
    what i would like to contribute to your blog is confounding variables. As you mentioned causation is how the Iv changes the Dv regardless of extraneous variables, however how can we be sure that there is no confounding variables? they are by there vary nature sneak and difficult to spot, yes we can do our best to remove them but we can never be sure there are none there, does this make the entire theory of causation redundant?
    luckily not, but to truly be sure of causation, a heavy experimental design is needed and on top of this a lot of human judgment and control before, during and after the research! Particularly during data analysis as statistical tests can often be too accurate and in a way misguided us to weather we have identified a causation or correlation.
    The key component to my argument here is yes its vitally important that we don’t say causation when it could be a simple correlation, but even if we can establish this difference psychologists tend to jump on board without considering extra confounding variables they didnt account for.

    • nim2152 says:

      Yep. I just knew I was missing the argument component of this question. Thanks for pointing it out! (^______^)
      I totally agree with your point.
      It is very difficult to be 100% sure that causation occurred and that no confounding variables were present making it a highly subjective matter.
      I guess at the end of the day it will always be subjective as Psychology is not a ‘hard’ science; in that there will always be at least one flaw in every psychological study we come across.
      However I think causation can be implied if the study was replicated many times by other researchers and still led to the same conclusion; as despite the possible finding being due to a confounding variable in one study, if another experimental method was used instead that takes care of the previous confounding variable; and other replications which also used different experimental methods and confirmed the original studies findings.. then we could state that there is a causation until it has been refuted?
      So I guess a studies findings on its own should be questioned and never 100% trusted until many many other researchers replicate the original findings. – Although exactly how many times is also subjective.

      Thanks for reading and commenting!

  2. poeywycheung says:

    Hi 😀
    It is good that you have explained what correlation (positive/negative) is and definition of causation and it is easy to read. (Even outsiders can understand)
    The example (I heard from someone) is that, there is a positive correlation between the sales of TV and the sales of cars in the 1980s. There is a relationship but does not necessary means this is due to the cause and effect.
    And there is a “famous” phrase in stats: Correlation does not imply causation.
    I just found the definitions in oxford dictionary, correlation:”mutual relation between two or more things” and causation:”causing or producing an effect”. Pretty much the same as yours so they are just backups!
    Relationship does not equal to causation, to determine causation, it seems to require not just a correlation, but a counter-factual dependence. And well-designed methodology is also needed. AT least, we should aware of the bias which might have caused in the research/study if we want to see if there is causation.
    (I am not sure what I am talking about but it still counts as a comment:D)

  3. prpj says:

    I feel like this blog has communicated to me clearly the issue of correlation and causation. We can see that correlation does indeed NOT explain causation. I found this site with some good examples of this It is important to have a knowledge of these correlational experiments showing up in the media and how to spot them because they can be used to claim findings are causal and can be used for all the wrong things!

  4. the WHO says:

    Nice blog, easy to understand and the terms are well explained.

    David Hume argued that causality is based on experience, and experience similarly based on the assumption that the future models the past, which in turn can only be based on experience – leading to circular logic

    I would like to add something new which is related to the field of media, also I will provide an article example.

    I found this press article with the headline: “Diet of fish ‘can prevent’ teen violence. This press article regarding the research study and the result that they have found. Here are some facts from the article –

    •Participants were a group of 3-year-olds given an “enriched diet, exercise, and cognitive stimulation.” They were compared to a control group who did not go through this same program.
    •By age 23 they were 64% less likely than a control group of children not on the program to have criminal records.

    Assume, of course, that the enriched diet included fish. However, note that the media article does not mention what the other kids ate or did. The data do not support the headline, it simply suggests a correlation between two variables. Unfortunately, in most of the cases, the popular media often misrepresent the research on which they are based. This article suggest causal relationships when, upon closer reading of the article itself, one finds that the research was correlational in nature, and the headline is not justified.

    To this end, correlations are looking to see if there is a relationship, but very often people assume relationship means causation. Refer back to the article that I have found, There are many other underlying factors which may decrease violence in teen, factors that could have contributed to this relationship, such as the participants had good educational background or good parental caring (if any) which minimise the exposure of criminal acts, alcohol etc and less likely to act like a one. So to conclude, media often blindly mis-justified research studies results as correlation cannot tell you much about cause and effect.

    Thank you.

  5. psuc97 says:

    Hey you’ve really clearly defined each term and made some good points 🙂
    You’re right saying that correlation definitely does not mean causation. Correlation studies are highly useful though as they used as preliminary research and show researchers whether there is any point further researching the impact one variable has on another.
    One major problem with correlation studies though is that they can get into the wrong hands and be misinterpreted. This is a great book (Sorry about the massive link, didn’t realise it was that long!) but yeah this is a great book and explains how the media loves to blow things out of proportion and tell the world that if you do one thing, something crazy will happen. For example it was stated that cannibis cured a young boys cancer his father slyly fed him cannibis oil through his feeding tube. However the boy was receiving intense chemotherapy and it was more likely that this was the reason for the boys recovery. However that isn’t exciting enough for the media and they have to go with the most outrageous or interesting option in order to sell papers.

Leave a Reply

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

You are commenting using your 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

%d bloggers like this: