1. Pearson's correlation or not?

Hi, I am a student at university of Birmingham, studying medical science. As part of one of my modules i have to do bit of stats...So I Hope someone can help me out.

Background to study
A study was carried out to see the effects of sleep deprivation on vigilance and reaction time. The study contained two groups sleep deprived and no sleep deprived groups, both conatined male and female participants. The following was recorded gender, numbers of hours kep awke, vigilance score and reaction time.

The question I am trying to analyse is " consider whether gender mediates relationship between hours of sleep deprived and reaction time"

I have considerd doing following-- do a Pearsons correlation to establish if there is a relation between hours kept awake and reaction time for male and female. Then see if my null hypothesis of r=0 for male and female can be rejected.

But I am confussed on how to show if gender mediates a relationship between the two variables.

So Please let me know what I can do?

2. Re: Pearson's correlation or not?

Originally Posted by Ruhofaston
Hi, I am a student at university of Birmingham, studying medical science. As part of one of my modules i have to do bit of stats...So I Hope someone can help me out.

Background to study
A study was carried out to see the effects of sleep deprivation on vigilance and reaction time. The study contained two groups sleep deprived and no sleep deprived groups, both conatined male and female participants. The following was recorded gender, numbers of hours kep awke, vigilance score and reaction time.

The question I am trying to analyse is " consider whether gender mediates relationship between hours of sleep deprived and reaction time"

I have considerd doing following-- do a Pearsons correlation to establish if there is a relation between hours kept awake and reaction time for male and female. Then see if my null hypothesis of r=0 for male and female can be rejected.

But I am confussed on how to show if gender mediates a relationship between the two variables.

So Please let me know what I can do?

First plot scatter diagrams of reaction time against hrs kept awake for both male and female subjects to eye-ball the data, then decide if a test is appropriate.

CB

3. Re: Pearson's correlation or not?

Sounds like you have 2 independent nominal variables (gender and sleep deprivation) and 1 dependent interval variable (reaction time).
I believe this is typically modeled with a 2-way ANOVA.
(Or alternatively a multiple regression with dummy variables.)

Of course it is always wise to eyeball the data in graph form.
See if anything weird is going on.

Thanks