Variables do not correlate

[uminka]

I used paired samples t-test and got 0.003 signifikance, but my variables did not correlate. I read, that if they don`t I should use another test. My question is: is it a mistake to use t-test anyway? If it`s so, which test should I use instead? I`m a student and not very experienced in statistics yet...

(Additional info: I did within-subject experiment and need to compare two variables, which represent decision making score of the same group in different conditions. Variables are normally distributed and so are their differences.)

[Bruce Weaver]

You say the two scores "did not correlate".  I take it you mean the correlation was not statistically significant.  Please report the sample size, the value of Pearson r and its corresponding p-value.

uminka wrote

I used paired samples t-test and got 0.003 signifikance, but my variables did not correlate. I read, that if they don`t I should use another test. My question is: is it a mistake to use t-test anyway? If it`s so, which test should I use instead? I`m a student and not very experienced in statistics yet...

(Additional info: I did within-subject experiment and need to compare two variables, which represent decision making score of the same group in different conditions. Variables are normally distributed and so are their differences.)

[CG]

I am out of the office until February 15.  For immediate assistance please contact Jessica Jaramillo ([hidden email]). Thank you.

[uminka]

Hi Bruce and thank you for your answer,

to tell you the truth, sample size si quite small. N=24; r=0,1; p=0,642; and in the second comparision: N=21; r=0,236; p=0,303. But small correlation isn`t unexpected in this case, because I`m comparing score from situation, where participants had to guess , which symbol is on the card,  with scores from situations, where they also had to guess the symbol, but they were unconsciously influenced by various stimuli. So the possibility of statistically significant correlation was pretty small in this case. I just don`t know, if the small correlation will not interfere with the test, or something like that.

Uminka

[uminka]

I`m not sure, if you can understand me clearly, because I`m not used to using English, when talking about statistics, so it`s hard for me to find the right expressions.

[Art Kendall]

Okay lets start over. 
make sure that you have completed the variables name, variable labels, value labels, and missing values and their labels in the variables view.
In a syntax window type then run.
DISPLAY dictionary.

Please cut and paste the output into a post.

Do you have 1 group  of respondents with scores under 2 conditions?
or
Do you have 2 groups of respondents with scores under 2 conditions?

if you are using menus etc, exit them via <paste>,  then post the syntax.

How did you create the scores? were there a number of stimuli shown to the respondents?



On 4/13/2012 2:08 PM, uminka wrote:

I`m not sure, if you can understand me clearly, because I`m not used to using
English, when talking about statistics, so it`s hard for me to find the
right expressions.

--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Variables-do-not-correlate-tp5638650p5639093.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD