Archive for September, 2011

Are there benefits to gaining a strong statistical background?

September 28, 2011 6 comments

Hello!  As you may or may not know, this is my first blog in the statistics series.  Today, I shall be discussing whether there are benefits to gaining a strong statistical background.
Of course, as Psychologists we need to have a close and personal relationship with statistics.  Statistics help us to analyse our data.  Statistics help us to understand the results we have collected.  Find out whether our data is even valid, if we found what we thought we’d find and if not, what our data actually shows.  We need to know what kind of data we will be collecting and how to figure out what all the numbers mean.  We need to know about statistics to make any sense of what we find.

Lets think about Asch’s conformity experiment for a minute. For those of you who are unaware of Asch’s experiment, I shall explain a little.

(If you already know about it – feel free to skip over this bit, make some food, stare at facebook waiting for someone to post a status…the usual).

Asch’s experiment, as you may have gathered, is on conformity.  He wanted to know if people would go against what they knew was right if everyone around them agreed on something that was wrong.  He set up an experiment, which participants believed to be a vision test. Participants were placed in a room with a group of confederates (who were basically in on it with Asch and had been told what to do).  The group of people, including the real participant, were shown a series of lines – a different length line every time. The group were asked to match this line to either A, B or C, one of which was the same length.  If I’m not making much sense, here’s a nice picture for all you visual learners out there:

In the picture above, the answer is clearly C. But, the confederates had all been told by Asch to give a wrong answer. So, they may all have answered A.  What Asch wanted to see is if the participant would answer C, because they knew that this was the right answer or whether they would conform to what the rest of the group said, and answer A.

Nearly 75% of participants conformed at least once. In other words, 75% of them gave the wrong answer to ‘fit in’, even though they knew what the correct answer was.

People who stopped paying attention above, now is when I try to make my point.

Even though this is a fascinating piece of research, think about what Asch actually had when he finished collecting his data.  He would have ended up with a whole mess of numbers and  to even make any sense of these numbers, there is a need to know about statistics.

If you want to have a look in a bit more depth about Asch’s study, there’s some information about it here:

and I’m sure a simple google search would provide a wealth of links if you’re really interested.

Having said all this, is just using statistics enough to tell you the whole story? For example, we don’t know what it was that made the participants conform. Also, 75% did conform…but what about the 25% that didn’t? What makes them different?

Can statistics be used and manipulated to look one way, when really they’re not that way at all?

(Any men out there may not be all too interested in this one, but just try to keep focused)

Have a look at this video for Avon’s extra lasting make up:

Make up that stays on all day, looks and feels great??  AND if you look about 14 seconds into the video, it says that 75% of women agree! Great, it must be true then.  But wait…does that say 75% of 52 women agreed?  52?!  So, that means that to have 50% of women agree only 26 women would have to agree?!  Which means that for 75%, only 39 women out of those 52 would have to say they agree with what Avon are claiming.

52 is a TINY sample size to be able to generalize the results to all women!

This is an example of how statistics can be used to ‘trick’ someone.

So, are there benefits to gaining a strong statistical background?  Of course there are,  if we want to conduct research and find things out, we can’t get anywhere without having a knowledge of statistics and how to do things to make sense of our results.  I just think that it’s important to not let it end there.  Find out the full story and make sure the amazing results you’re being shown actually are amazing.

Thanks for reading the ramblings of my inner mind.  Come back next week when I shall be answering that long standing question on everyone’s mind:  Do you need statistics to understand your data?


Categories: Uncategorized