For the duration of this blog I would like you to assume that when I say statistics I mean the level of statistics that is above what is learnt at GCSE level. The level I mean is the level above the basic mean, median and mode that my maths teacher desperately tried to yell across the room; over the cries of “I don’t GET IT”. And with that disclaimer over I will try to answer a question that I am sure has passed through many a mind of a 1st year psychology undergraduate during a particular distressing lecture about analysis of variance; Do you NEED statistics to understand your data?
With fairly basic levels of numeracy it is easy enough to work out the average or the median of a set of data. Then, if you have collected this data from more than one group/condition, you can draw conclusions about the different groups and voilà… you have interpreted data on a basic level. You can suggest that Group A could be happier/smarter/faster than Group B because the average was much higher. But that is where you will begin to get stuck if you have no idea (of even just the basic workings) of more advanced statistics. You will not be able to say that the difference was significant, and therefore, even relevant or if the difference could have just arisen by chance. Here is a link http://www.skorks.com/2010/03/you-dont-need-math-skills-to-be-a-good-developer-but-you-do-need-them-to-be-a-great-one/ to a lovely article that explains how statistics/maths is crucial if you want to have an interesting career (granted the article is talking about software developing but I think the point is the same for almost all professions). If you want to do research then, ok, maybe you don’t need statistics, but if you want to find your research interesting and become a great researcher then you will definitely have to do much more than comparing two means with each other time after time!
So maybe we do need statistics. But then there is the beautiful invention of the computer! The brain that was born from numbers and science and so has no problem crunching numbers all day to give us that glorious printed page of words that tells us quite (almost) plainly whether the data we have given it is in anyway worth bothering to write about. So if the computer is going to tell us the answer anyway, why should we bother trying to shove all of this information into our brains about calculations workings out when the computer can spit the information at us and we can just write it down? I believe that the answer is that as potential scientists, surely we want to really understand the work and research. When I am looking at data, especially if it is my own, I want to know WHY personal differences between the people in the study can make such a difference (and why the computer thinks it can work it out without knowing them). I want to know WHY anomalies can occur and WHY we can sometimes exclude them (and HOW we do this). When creating experiments I want to know WHY three people normally isn’t sufficient where as thirty is quite good. But maybe that is just me…
Even if you aren’t really that bothered about about finding your own research interesting or understanding why your computer knows that your data isn’t significant, I bet that you want to be able to tell when you are being lied to. If you read a research report and something about it just doesn’t feel right, you want to be able to go along and find out why you feel like that and understand their data for yourself. Here is an article that tells us about statistics in the world today and how, without statistics we won’t know when we are being lied to http://www.wired.com/magazine/2010/04/st_thompson_statistics/. Now, this obviously is not going to happen very often but how often has some news programme or newspaper reported some statistic that you just can’t believe? With statistics you have the tool to go along and check their facts and see if what they are reporting is being reported in a full descent manner or if they have left out some specific detail that changes everything. Even the BBC’s Science & Environment news page has been thought to contain flaws that mean the full story is never really told to it’s audience (http://www.guardian.co.uk/science/the-lay-scientist/2010/sep/28/science-journalism-spoof). Although in the case of the BBC it seems that it is their desire to remain COMPLETELY impartial that is leading to this problem.
So there we have it. Technically you could go along and try to complete research without statistics but this research could possibly be quite dull or you might not be able to answer questions about the variance between participants because it was your computer that worked it out. You might not be able to tell if there is a mistake in your own work or the work of others. Is that the kind of scientist you would want to be?
Pictures from:
psuc3d 
Hii 🙂
First of all, you managed to retain my attention for longer than a few minutes with a funny but straight to the point opening to your blog. I like it as it adopts the “upside down triangle” structure that lecturers often tell us to adopt in our essays, where the bold statement leads into the arguments which become much more specific as the writing carries on.
It seems you have adopted a different approach to some of the other blogs I have read, so it’s great to see a breath of fresh air. Whereas most people have gone straight for the examples of research you have gone for the application of statistical methods in real life and how they are useful in all walks of life, and that’s before you even mention a computer! There are a number of different points you have made as well, such as the importance of statistics to enable understanding and misuse of statistics which I think shows a broad understanding of the issue. Well done!
The only thing I would suggest, (even though you have justified not using one) is to mention an opposing point to your arguments to show that you have considered both sides of the issue, and also possibly use more examples of actual psychological research as I myself deem these as slightly more trustworthy than newspaper articles etc (which kind of relates back to your point about the flawed BBC articles!) Other than that, good job and good luck for next week.
Hello,
Great pictures, the effort of looking for these pictures online showed you actually spent some time on it.. or have you? LoL. Anyways, lets get to the business, I find it amazingly interesting reading this blog, you have sort of creating a debate here? Like a newspaper, not tabloid ones, you know the broadsheet types of newspapers, where they create a debate about an issue, standing neutral towards this issue / event, and let readers to think about it themselves? So good job here, I really like your style.
One problem though, is that, you’ve mentioned different ideas, views or points about whether or not statistics = is a must, but since there are quite a few points included, I kind of lost of way in the middle (not till the end at least), If i understand your blog correctly, You thinking stats = a must right? Or at least know it = a qualify researcher?
But then, from my personal view, I think stats is really important for any science based subjects, since without it, you cannot draw your conclusion confidently nor you can say out loud that your results are representable for a whole population, so I totally agree with you here.
Lastly, I actually took a look at your Links, (http://www.guardian.co.uk/science/the-lay-scientist/2010/sep/28/science-journalism-spoof), I think just have to be really careful in general, These kind of tricks occur globally, and its a tactic used often by the huge evil empires (Companies) or information released by the officials.
Thanks! Really liked it, Keep this up! Looking forward to read your next one.
Hello, really liked your blog, was really interesting to read unlike other blogs ive read 🙂
I would have to add that sometimes statistics arent needed in order to understand the data, sometimes it is best if the raw data is looked at because sometimes this shows a clearer picture of the data rather than inserting it into SPSS and producing a load of statistics in which you can interpret. This would be the case for case studies, such as freud.
Hi there, I have just read your blog and loved the way you introduced it. I do believe that a writer’s introduction is the key, to whether or not the rest of what you have written will actually be read by the reader. You managed to get me interested in what you had written by starting with a funny comment in which we can all relate to. By restating the question at the end of the introduction, you have made it clear to me what the blog is about and you have kept it professional.
It’s great that you have supported the reasons for why statistics are needed (so “we are able to find out when we are being lied to”). What would be nice to see here is maybe a link within scientific research, where statistics have helped us get a more meaningful view rather than just having a load of data e.g. with medical research into chemotherapy. Here statistics have given patients and the medical profession a clear view on whether chemotherapy if effective. The paper published in the Australian journal Clinical Oncology states that statistics found that “ in lung cancer, the median survival has increased by only 2 months [during the past 20 years] and an overall survival benefit of less than 5 percent has been achieved in the adjuvant treatment of breast, colon and head and neck cancers.” Link: http://www.icnr.com/articles/ischemotherapyeffective.html.
Although you have highlighted that statistics is a way of being able to check, whether a certain statement is true or whether it’s a lie. What you could of also highlighted, is that sometimes statistics are published in the media, or stated by politicians which are false, but the public trust them (they don’t check them with other statistics) For example when the conservative party stated that knife crime was on the rise in 2008 a lot of the public believed this and it lead to a moral panic. Here is a link: http://www.guardian.co.uk/uk/2008/may/13/ukcrime.boris . In reality knife crime was actually on the decrease.
Also adding on to this, you could say an advantage however of psychological published research, is that if it is published in the British psychological Journal, you know that the statistics are truthful, as they have had to pass the criteria of the peer reviewing body. It’s research statistics within journals such as the British psychological journal, which can help us understand data.
Great blog, I look forward to reading your next one !
Pingback: Homework for my TA – week 3 (14/10) « rgjblog