Statistics Help @ Talk Stats Forum - Statistics
http://www.talkstats.com/
Statistics course and homework discussion. Elementary statistics.enFri, 25 Apr 2014 05:13:40 GMTvBulletin60http://www.talkstats.com/images/misc/rss.pngStatistics Help @ Talk Stats Forum - Statistics
http://www.talkstats.com/
<![CDATA[Testing Model Assumptions - Randomized Complete Block - Normality & Homogeneity Var]]>
http://www.talkstats.com/showthread.php/55242-Testing-Model-Assumptions-Randomized-Complete-Block-Normality-amp-Homogeneity-Var?goto=newpost
Fri, 25 Apr 2014 04:13:23 GMTF1 - 3 Levels; F2 - 2 Levels; Block - 3 Levels.

I have a data set with one rep per Block and Treatment combinations of F1 & F2.
An example is
F1(1), F2(1), B(1) = 34.4
F1(1), F2(1), B(2) = 58.9
...
F1(3), F2(2), B(3) = 58.9
In total, I have 18 observations (3*2*3).

I can build a model with interactions, blocking, etc. in SAS, but how do I test my model assumptions ? I am unable to test the variance within any given Treatment*Block combination since I have only one rep. I was thinking maybe residuals could be used for testing Normality & Variance but am not sure.

Any ideas ?
]]>Statisticssimplemtshttp://www.talkstats.com/showthread.php/55242-Testing-Model-Assumptions-Randomized-Complete-Block-Normality-amp-Homogeneity-VarWhat type of analysis is best ?
http://www.talkstats.com/showthread.php/55239-What-type-of-analysis-is-best?goto=newpost
Thu, 24 Apr 2014 23:54:07 GMTTo sum up I have a population of 36 respondents split into two groups High perfomers (18) and low performers (19). I would like to analyse the mean...To sum up I have a population of 36 respondents split into two groups High perfomers (18) and low performers (19). I would like to analyse the mean variables for statistical differences between the two sets of data.

I have little to no stats experience and any help is greatly apreciated.
]]>StatisticsMorgan1991http://www.talkstats.com/showthread.php/55239-What-type-of-analysis-is-bestHelp with mixed measures design
http://www.talkstats.com/showthread.php/55237-Help-with-mixed-measures-design?goto=newpost
Thu, 24 Apr 2014 21:46:49 GMTHi,

I'm carrying out analysis on a mixed measures design. The within subjects factors are 2 (male/female) x 2 (high dependence/low dependence) x 3 (control, long term depression, short term depression). The dependent variable is probability of treatment choice.

I used 12 videos to expose people to the within subjects factors, which all differed according to the within subjects factors. However, they also differed by intensity of emotional expression (varying from 5 to 9, most of them were 7). The way my SPSS sheet is set up won't allow me to control for the intensity of emotional expression. I created a new SPSS sheet and pasted the data so that I could correlate intensity of emotional expression with the dependent variable. There were no significant correlations. However, pasting the data this way meant that cases appeared on more than two rows, which doesn't seem right. Did I do the right thing? Sorry if this question is confusing, I'm really confused myself. Thanks.
]]>Statisticscoppersulphatehttp://www.talkstats.com/showthread.php/55237-Help-with-mixed-measures-design