Do you need statistics to understand data?

Once again a vague title. When i first read it i found myself struggling to lay down a definite opinion. After all, there are a variety of arguments for this topic. Luckily for you i will not even contemplate trying to cover all of them.

Firstly, lets take for example analysis of variance (ANOVA), an area which i hope i will not have to linger on for long. ANOVAs are a collection of statistical models and in its simplest form provides a statistical test, measuring whether or not the means of several groups are all equal. We all know that there is more to it than that but nevertheless, to understand the data an ANOVA produces you will need to understand the statistics for which it provides: p values, type 1 errors, degrees of freedom, the two go hand in hand.

However, lets consider something simpler such as graphical data. An example of this would be a bar chart. We need to collect statistics to make a bar chart, we need to organise statistics to help make a bar chart, but i would argue that you do not need statistics to help you understand the bar chart. A bar chart is a very simple form of displaying data, we are taught how to understand them in primary school. It is very easy to see a relationship between two variables when it is presented to you in this way. Otherwise why would it be taught to us at such a young age?

Nonetheless, even i will not dispute that analysis of variance data provides far more detail and is a value statistical procedure. But to say that we need statistics to understand data is far too ambiguous. I would have to argue that the need for statistics in understanding data is dependent on the type of data we are provided with.

In my case, i’d rather just be handed a bar chart. 🙂


7 responses to this post.

  1. Exacly I agree , whenever we need statistics to understand our data is dependent on type of data that we want to collect. Hovewer regarding your point on second paragraph I think that we still need statistics to understand graphical data. Yeah I agree with you on bar chart point because it is simple and easly understandable but there are differnt graphs that can be produced and require a bit of statistical knowledge to understant it. |For example steam graph or line or correlation it is not easly understable unless you know what relationship or differences it measures among variables). Further I agree with you on the example you showed in paragraph two ,(reelating quantitative type of data,) which is correct , that we do need statistcs to understand the outcome of the experiment. But how about the gualitative data?


  2. I like how this was quite informal and i enjoyed reading. I have to say i agree with everything you say, and i especially dislike ANOVA. However i felt that once you started getting into your explanations, it ended a bit abruptly, when more points could have been expanded. Also you made some very good points about quantitative data, but you mentioned very little on qualitative data. The conclusion does not state a definite opinion, but i have to say i agree, it does depend on the type of data that is used. Overall i enjoyed reading this blog post.


  3. Well done on this blog. I’ve enjoyed reading it due to it’s simplistic nature and professional form.One thing I liked in particular was how you acknowledged that statistics are important to create graphs, box plots, scatter graphs, etc but how it’s not important to simply understand what you see. I agree with your view point however I would have personally discussed something other than ANOVA, as an understanding of statistic is obviously important to work at this grade. Other than that, great blog


  4. Good blog, and you have made some good points. I think however you have underestimated statistics and i think that it is actually quite vital for understand data. Especially for psychologists, as you need to use statistics in order to try and prove hypothesis, if you didnt contain statistics within your work how are others supposed to know what you are saying is correct. There not just going to take your word for it, they will need evidence which is where statistics is used. However i do acknowledge that statistics cannot tell the whole story or give you the complete understanding. For people that havent studied statistics they are going to need someone to explain what all the statistics mean. This can be just as important as the statistics that you have gathered. Looking forward to reading your next blog.


  5. great blog, was a good read and you clearly understand the points you are making 🙂 statistics are indeed useful but not completely necessary, i think that the fact that not all experiments use statistics proves this point. many experiments, mostly qualitative but sometimes quantitative too, have data that is understandable without the use of statistics. for example Ainsworth’s experiment 1970 (strange situation) required no use of statistics whatsoever. 🙂


  6. Hey, congratulations on a great blog post.

    I agree with you that while statistics is essential is analysing some data, that there is a time when exploratory data analysis is more appropriate. I also agree with you that i would much rather handle a few graphs than a bunch of statistical tests. One thing i do think you left out was that you didn’t look into any ways of analysing qualitative data. When it comes to qualitative data statistics may not be necessary at all.


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 )

Google+ photo

You are commenting using your Google+ 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 )

Connecting to %s

%d bloggers like this: