robyndavies
Badges: 2
Rep:
?
#1
Report Thread starter 3 years ago
#1
I am running an experiment where people have to turn over cards that have a hazard word on. There are two conditions, one using uppercase words and one with lower case words. I am measuring the speed in which they say that they have turned over all of the cards with a hazard word. Is it right to do a paired-sample t-test to compare the mean speeds from the two conditions?

I am also wondering what statistics I should use for the errors made.I am measuring if a card with a hazard word is missed (not turned over), or if a card with a non-hazard word is turned over. How do I analyse this data?
0
reply
Interrobang
Badges: 20
Rep:
?
#2
Report 3 years ago
#2
Moving to Study Help, where you're more likely to get replies
0
reply
AnnaBanana2000
Badges: 11
Rep:
?
#3
Report 3 years ago
#3
(Original post by robyndavies)
I am running an experiment where people have to turn over cards that have a hazard word on. There are two conditions, one using uppercase words and one with lower case words. I am measuring the speed in which they say that they have turned over all of the cards with a hazard word. Is it right to do a paired-sample t-test to compare the mean speeds from the two conditions?

I am also wondering what statistics I should use for the errors made.I am measuring if a card with a hazard word is missed (not turned over), or if a card with a non-hazard word is turned over. How do I analyse this data?
Well you have to answer 3 questions to decide which inferential statistic test to use:

Which level of measurement? If you are counting how many people turn over hazard words, how many people do not turn over hazard words, etc... It is NOMINAL data.
Is the data related? Which experimental design are you using? independent groups=unrelated, matched pairs, repeated groups=related data
Is it a measure of difference or correlation? I'm guessing your data is a test of difference between hazard cards and not etc

I think that for your data it would be Chi Squared?
0
reply
iammichealjackson
Badges: 18
Rep:
?
#4
Report 3 years ago
#4
For the reaction time data its probably fine to use a t-test. Because reaction times don't follow a normal distribution (they are usually modelled with the gamma distribution), the distribution of means from a non-normal distribution will approx follow a normal distribution because of central limit theorem if you have enough N.

I think chi-square is what you use for the other too. not sure if I've understood correctly though.
0
reply
TryingIsGood
Badges: 9
Rep:
?
#5
Report 3 years ago
#5
Can anyone help? I'm looking at sample questions for a job test, but my stats are a bit rusty and this one's got me stuck. Thanks!

A study into the number of cigarettes smoked by male and female smokers found that the number was normally distributed for both groups with a mean of 7.2 per day. However, the standard deviation was 4.0 for men and 2.5 for women. Based on this information, which ONE of the following MUST be true?
A. The median is higher for men than for women.
B. A man at the 70th percentile smokes more than a woman at the 70th percentile.
C. A man at the 10th percentile smokes more than a woman the 10th percentile.
D. Variance is less for men than for women.
E. The overall mean is 14.4.
0
reply
X

Quick Reply

Attached files
Write a reply...
Reply
new posts
Back
to top
Latest
My Feed

See more of what you like on
The Student Room

You can personalise what you see on TSR. Tell us a little about yourself to get started.

Personalise

Have you made your mind up on your five uni choices?

Yes, and I've sent off my application! (150)
54.74%
I've made my choices but havent sent my application yet (38)
13.87%
I've got a good idea about the choices I want to make (31)
11.31%
I'm researching but still not sure which universities I want to apply to (26)
9.49%
I haven't started researching yet (16)
5.84%
Something else (let us know in the thread!) (13)
4.74%

Watched Threads

View All