scs
07-09-2007, 12:10 PM
Hi,
Firstly I feel the need to apologise for the long (and probably confusing post) but I am in real need of some help with my research.
I am currently trying to complete a piece of research for my MSc. I would really appreciate some help with the stats as they are not my strong point and have been driving me crazy.:( :confused: :eek:
My research question is looking at weather a treatment programme helps people who are psychopaths recognise emotions in others better(emotion representation).
I have three groups. Group 1 who are not psychopaths, Group 2 who are high scoring psychopaths and have been in treatment for a long time and Group 3who are high scoring psychopaths and have not been on treatment very long.
The way the emotional representation is tested is by timing participants reading a sentence after a story that either matches the emotion infered in the story (congruent sentences) or mismatches the emotion infered in the story (incongruent sentences). The reasoning behind this is that it takes people longer to recognise (and therefore read) things they are not expecting. If they have infered the emotion right they will not be expecting the mismatching word to come up.
I am looking at 6 emotions, so for each participant I have a congruent and incongruent time for each emotion and a total congruent and incongruent time. I have also worked out the differences between the congruent and incongruent scores for each emotion and the total differences.
My question is which stats test do I use. I was using a one way independant ANOVA which semed to be working okay but my tutor told me I need to use a MANOVA. My difficulty is that my numbers are very small (group 1=10 group2=5 and group3=7) I thought that a MANOVA would not work with such small numbers
Also should I compare the congruent and incongruent scores for each emotion and the total or do I compare the diffences that I have calculated for each emotion.
Hope all that made sense.
Thanks in advance for your help.
scs
Firstly I feel the need to apologise for the long (and probably confusing post) but I am in real need of some help with my research.
I am currently trying to complete a piece of research for my MSc. I would really appreciate some help with the stats as they are not my strong point and have been driving me crazy.:( :confused: :eek:
My research question is looking at weather a treatment programme helps people who are psychopaths recognise emotions in others better(emotion representation).
I have three groups. Group 1 who are not psychopaths, Group 2 who are high scoring psychopaths and have been in treatment for a long time and Group 3who are high scoring psychopaths and have not been on treatment very long.
The way the emotional representation is tested is by timing participants reading a sentence after a story that either matches the emotion infered in the story (congruent sentences) or mismatches the emotion infered in the story (incongruent sentences). The reasoning behind this is that it takes people longer to recognise (and therefore read) things they are not expecting. If they have infered the emotion right they will not be expecting the mismatching word to come up.
I am looking at 6 emotions, so for each participant I have a congruent and incongruent time for each emotion and a total congruent and incongruent time. I have also worked out the differences between the congruent and incongruent scores for each emotion and the total differences.
My question is which stats test do I use. I was using a one way independant ANOVA which semed to be working okay but my tutor told me I need to use a MANOVA. My difficulty is that my numbers are very small (group 1=10 group2=5 and group3=7) I thought that a MANOVA would not work with such small numbers
Also should I compare the congruent and incongruent scores for each emotion and the total or do I compare the diffences that I have calculated for each emotion.
Hope all that made sense.
Thanks in advance for your help.
scs