Inferential Statistics help!

holla

New member
Joined
May 18, 2013
Messages
1
So my statistics group is doing our final project. Our research question is: Does caffeine affect short term memory?

We had 10 subjects do our experiment. We gave each of them a list of words and gave them 3 minutes to memorize the words. We then gave them 1 minute to write all the words they remember.

After that, we gave 5 of them the caffeinated coffee and the other 5 decaf. We then gave them a different list of words and gave them 3 minutes to memorize the words and another minute to write down the words they remembered.

The results look like this:

CaffeinatedDecaffeinated
Subject #Amount of wordsSubject #Amount of words
1Before: 13
After: 15
6Before: 12
After: 11
2Before: 8
After: 9
7Before:10
After:11
3Before: 9
After: 8
8Before:7
After: 7
4Before: 8
After: 11
9Before:9
After: 8
5Before: 12
After: 12
10Before:10
After: 11
















What statistics testing should I use?

It's kind of like difference of mean of difference because there is the before/after factor and also the caffeine/decaf. I really don't know what to do.

Thanks in advance.
 
So my statistics group is doing our final project. Our research question is: Does caffeine affect short term memory?

We had 10 subjects do our experiment. We gave each of them a list of words and gave them 3 minutes to memorize the words. We then gave them 1 minute to write all the words they remember.

After that, we gave 5 of them the caffeinated coffee and the other 5 decaf. We then gave them a different list of words and gave them 3 minutes to memorize the words and another minute to write down the words they remembered.

The results look like this:

CaffeinatedDecaffeinated
Subject #Amount of wordsSubject #Amount of words
1Before: 13
After: 15
6Before: 12
After: 11
2Before: 8
After: 9
7Before:10
After:11
3Before: 9
After: 8
8Before:7
After: 7
4Before: 8
After: 11
9Before:9
After: 8
5Before: 12
After: 12
10Before:10
After: 11
















What statistics testing should I use?

It's kind of like difference of mean of difference because there is the before/after factor and also the caffeine/decaf. I really don't know what to do.

Thanks in advance.
I don't know what test you know! The first thing I would do is calculate means and standard deviations for some aspects of the data. For instance, you could find the average for the full group of 10 for the "before" case. That would represent your estimators of the "population distribution."

Then I would look at the DIFFERENCES for each group:
Caffeinated: +2, +1, -1, +3, 0
Decaff: -1, +1, 0, -1, +1
What are the means and std.dev. for those two sets?
Are they different? -- that is where you have to choose a test.
I think I would just find the confidence interval assuming a normal distribution.
"What is the probability that the observed difference may be due to random fluctuations?"

An interesting experiment, but two small of a sample size to get really formal about it.
 
Top