Do you need statistics to understand your data?

Statistics and the accompanying procedures are necessary in helping you organize and analyse your data, but it will not help you interpret or understand your data.  Bangor Psychology undergraduate you will be aware of the statistics program that I will be referring to as ‘SPSS’.  For those of you who are un-aware, SPSS is a computer program that will take large quantities of data, organize it and then analyse it when prompted to by the user.  After analysing your data, SPSS will present you with a summary of the descriptive statistics derived from the data you provided in the form of graphs, tables and figures which can sometimes be difficult to decipher.  There is no statistical procedure to interpret the data for you.  Despite this, it is still a very helpful program and can save you a lot of time, especially when you are handling large quantities of data.

In order to be able interpret your data effectively you will need to have attained a sufficient understanding of the processes involved in analysing it.  In fact, I believe that SPSS can actually harness your ability to interpret and understand data by offering you a short-cut and carrying out the analysis for you.  It’s quite alike my mum when it comes to sending texts on her mobile phone.  She will decide who she wants to text, write her message and then call for me as soon as she has finished composing it.  I will then be asked to ‘do the rest’ i.e. adding the recipient and sending the text.  My mum seems to think that this process exceeds her knowledge of phones and she lacks the ability to ‘do the rest’ and will never attempt to carry out the finishing stages, thus, never actually learning or understanding the process.  After we plug our data into SPSS we just ask the program to ‘do the rest’ causing us to never actually understand the process, and it is understanding the process that gives us an even greater understanding of our data that is essential for us to interpret it.

By learning to carry out the appropriate statistical procedures manually you will develop the ability to understand why and how you obtain the figures that you do, you will be able to identify the story that your statistics will tell you.

Advertisements

7 thoughts on “Do you need statistics to understand your data?

  1. I like the way you gave a different perspective to the argument by looking at how us as students understand statistics. An interesting article I found on the internet http://www.userfocus.co.uk/articles/stats_not_maths.html suggests that SPSS is the “tool knowledge” for statistics…however I disagree.
    SPSS I find is the most aggravating program to use, statistics in its self I find difficult to understand, and SPSS goes ahead, and yes as you says “does the rest” but only when you input it correctly, which sometimes can be a mission by itself. (The amount of times last year when me and my friends turned up to the stats test with the “wrong” SPSS)
    I think it would be nice to leave the computer programs to the side one day, and actually have to sit back and learn which method we need to do ourselves, be it ANOVA, T-tests, standard deviations etc.

  2. I agree with you that while SPSS is a useful tool that can speed up our work, without a true understanding of statistics and the processes that SPSS uses to spit out the answers, one cannot comprehend one’s results fully or infer things confidently. It is for this reason that I think that it’s a terrible pity that we now do blogs (which, okay, do lead to discussion, but not much else) instead of truly learning statistics and all its associated methods. I feel that last year, everyone was actually beginning to understand why SPSS gave the answers that it did – and that the whole process became quite intuitive. (But maybe this is my love of maths and my nostalgia that is colouring my memory!)

    An important point to make is that your argument was good, (and I loved your mother texting analogy – very well judged and something I can relate to), but possibly a little one-sided as you didn’t discuss the fact that one can run a study and gather qualitative data and do away with the need for statistics in this particular case.

    Good blog on the whole and I am looking forward to next week’s one.
    🙂

  3. I like how you’ve put your point across and used the example of SPSS throughout. I agree that SPSS is a good short cut and produces the answers we need for tests and study results, however SPSS itself is difficult to understand (well i personally don’t understand it) but what about if we have qualitative data, such as that from interviews or case studies? We don’t need statistics in order to understand this kind of data. We can physically look at our findings (for example, listen to or view a recording of an interview to pinpoint the behaviours we are looking at). However by using qualitative data we would always be penalised as it isn’t considered as scientific as quantitative data, even though it is a well used method within psychology. Bandura, for example, conducted a very well known experiment in which he observed children’s modelled behaviour when given a bobo doll to play with. This is an established theory even though he did not use statistics to help to understand his findings, which shows us that we do not always need to use statistics and SPSS to analyse and understand findings from an experiment.

    But all in al I agree with what you have said, SPSS simply gives us the answers but that doesn’t mean we have understood our data.

  4. I agree with esh2 that SPSS is rather annoying and never seems to do what you want to do, and I would sometimes rather work it all out by hand, no matter how long winded the process, or how many stages are involved. The point about computer statistics packages taking away our true understanding is true – people can read off a p-value, but not tell you how it was calculated or what it actually means. They just know that if it’s smaller then the significance value then it means this.

    I have to disagree with you when you say that statistics will not help you to interpret or understand the data. I think that statistics – no matter how simple or difficult – is vital to understand it! Simple things such as knowing the means of 2 groups means we can easily distinguish the difference between them, and more complicated procedures can tell us more than this. Without the statistics, surely we would be trawling over the raw data trying to find the correlations or significant differences the statistics could tell us in an instant?

  5. Pingback: Comments I’ve made for week 2 blogs (for TA) « Statistics for Tea

  6. Pingback: Comments for Week 3 | psud6e

  7. Pingback: Comments I’ve made on the week 2 blogs (for TA) « psychmja1

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