I don't have a dog to eat my homework. Q about Experimental Design

Grunfeld

Long shot but no harm asking cos I'm stuck. 
Back story:  I have to write a case for support for funding a bit of research based on a  previous bit of research [this is an assignment for the MSc I'm doing].  The problem is:  the "bit of previous research" is quantitative and... I have no background in designing quantitative research.  (I've not been able to think of a qualitative follow up but this might be my best option still.)
So....
As it's a case for support I don't need a heroic amount of stats -- just enough to show I'm not as dumb about this as I really am.
Follow up experiment has 2 variables.
I'm hoping that would be a 2x2 factorial design? 
But...
what if you wanted to run those 2 variables on two different groups?  Would that be a shit-load extra statistical analysis which I'd clearly show my ignorance of?

To flesh it out:
Imagine you want to cause volunteers some experimental pain.
You've got two variables.  (1) metaphors about the pain; (2) suggestions for being able to tolerate  the pain.
So you run 4 groups:  1+2; just 1; just 2; neither
This experiment has been done already. 

To progress the research you could run it again on two types of people:
People trained in practice A
People great at talent B.

My dumb question forum people is:  is the A, B variation too difficult for a statistical ignoramus like me to attempt to describe?
I know "too difficult" is vague.  I think I may have answered my own question here.  I don't know enough about stats to even know what question I'm asking!!
I'm gonna go qualitative!  It's the only thing I know (and I'm okay at that). 
Err, as you were!

Danny1969

Wow Oli I was so bored reading that my advice is to get a dog

Kilgore

I would use yoghurt pots and sticky back plastic.

Sporky

If you post it to me I can present it to Sprocket for digestion.

LuttiS

What field is it in out of interest? 

In my experience.. quantitative is always the way

Grunfeld

LuttiS said:

What field is it in out of interest? 

In my experience.. quantitative is always the way

Experimental psychology. 
For this specific thing then yes I really know that quantitative is the way. 
E.g. Stick your hand in cold water, time how long you can keep it there.
Then retest +/- the variables of (1) metaphor and (2) being able to tolerate pain. 
So basically you say (1) imagine you're in cold, disgusting swamp (2) there's something on an island you really want five minutes swim away and you'll do anything to get it.
It yields interesting results.
Swamp and a reason to tolerate pain = 200% longer times
Just swamp, or only "you can tolerate pain", or neither  = about the same.
Metaphor + reason seem to make a difference to tolerating pain.

The paper I'm supposed to build on is actually in a different area:  mindfulness and hypnosis.  They might seem similar but they're kind of opposites. 
Imagine a line, a continuum of awareness:
|------------------------------------------------------------------------------|
really engrossed                                                                   totally disassociated
in a thought                                                                           from a thought

That's hypnosis, either end of the line.
You're either so totally in the thought that it's as if it's really happening, or you're so disassociated from a thought that it's as if it is not happening, or it's happening to someone else. This is why some people feel it's like mind control. E.g. Lift your arm. You create an intention and lift your arm. Imagine if you were not aware of your intention. You would actually form the intention to lift your arm, but not be aware of it, and when your arm moved it would be as if it were moving all by itself. And that's what happens in hypnosis. (Kind of, a bit more complicated.). So you can for example dissociate from pain or really believe that your arm is being lifted by lots of petty coloured balloons.

Mindfulness is like having your awareness in the middle of that line. You are aware of thoughts happening but you don't engage with them. It's like you're a shepherd watching a flock of sheep. You want to keep awareness of the sheep (the thoughts) but you don't want to get so engrossed in them that you run into the flock and start playing with them and going "baa", nor do you want to lose all awareness of them and you go wandering off into the next field ignoring your sheep.

So.... I was thinking, run the pain/swamp/tolerate test but with experienced meditators and highly hypnotisable subjects.  But I have no fucking idea why!  Other than it sounds fun.  But results analysis?

NelsonP

Grunfeld said:

To flesh it out:
Imagine you want to cause volunteers some experimental pain.
You've got two variables.  (1) metaphors about the pain; (2) suggestions for being able to tolerate  the pain.
So you run 4 groups:  1+2; just 1; just 2; neither
This experiment has been done already. 

To progress the research you could run it again on two types of people:
People trained in practice A
People great at talent B.

Wait, are we the volunteers that you are causing some experimental pain to? Because my head hurts.

From what you have said the interesting comparison is 1+2 vs neither (i.e. Metaphor + reason seem to make a difference to tolerating pain).

Do you really need A1, B1, A2, C2 (as these seem to make no difference vs the neither option)?

Then you would have four groups:
A1+2
B1+2
A neither
B neither

But don't just take my word for it as I don't have an MSc.Is your tutor allowed to advise on this? And I would not want to be responsible for screwing up your hard work!

Grunfeld

NelsonP said:

But don't just take my word for it as I don't have an MSc.Is your tutor allowed to advise on this? And I would not want to be responsible for screwing up your hard work!

All screw ups are totally mine!  I'm sure my tutor is totally allowed to advise on this.  But you know that thread on here about Do you have any bad habits?  And there were a couple of people saying, "oh I procrastinate..."  Amateurs!  I'm in a wee time-crunch.

And in all seriousness I think you have seen something useful.  Only the (metaphor + reason) group had a difference.  That might simplify things.  Once I've got the ideas clear in my head then I should be fine but I can tell that I've not seen what I want to see clearly yet. 
EDIT -- sorted!  (Beer is bloody wonderful sometimes.)
Flash of insight and totally different approach. 
All dogs waiting to assist may now stand down. 

NelsonP

Maybe this is one of those times when procrastination (looking at a guitar forum when you should have your nose buried in a stats book) leads to inspiration?

Interested to know more about the inspirational approach....

Grunfeld

NelsonP said:

Maybe this is one of those times when procrastination (looking at a guitar forum when you should have your nose buried in a stats book) leads to inspiration?

Interested to know more about the inspirational approach....

  It occurred to me I was trying to get too much out of one experiment.  What's needed are three experiments.  I won't waste electrons listing them, but really my thing is clinical and pain so three experiments that link from working out which component of mindfulness impairs hypnosis (probably "thought awareness"), to experimental pain, to clinical pain.  I can't fudge the lack of statistical insight so will just have to take that on the chin.

NelsonP

Well I'm glad you've sorted it. Good luck with the thesis!