Factor Analysis

#1
I'm trying to run a Factor Analysis on a large environmental data set with a combination of data types, and i'm having some difficulties. I get a Warning message listing a number of possibilities of what could have gone wrong. One option is that I have variables with zero variance. i've run a descriptive statistics listing variance and eliminated from my analysis those variables with zero variance listed. I've also run a scree plot and found some variables giving null data, so I've eliminated those from my factor analysis, still no luck. I'm not sure what to try next?! Any idea on what might be causing this? I'm using SPSS 19 with over 300 variables. The data was imported from Excel. Any suggestions would be wonderful!
 

bugman

Super Moderator
#2
Two suggestions:

1) in spss check how your data are coded in the data view tab. There might be some inconsistencies with this

2) you have alot of variables. If the number of observations is less than the number of variables you might run into difficulties.

:confused:
 
#3
Thanks for the suggestions. I'm working w/329 variables and 973 observations at this point, so well more observations.

Some data transfered into SPSS as 0E-9 or 0E-10 and shows in the toolbar writing space as .0 - could this be a problem?
I'm not including string data in the analysis. The warning I get follows:

Warnings
There are fewer than two cases, at least one of the variables has zero variance, there is only one variable in the analysis, or correlation coefficients could not be computed for all pairs of variables. No further statistics will be computed.

Again, I've used descriptive statistics to identify variables w/zero variance and eliminate them from my analysis.
Suggestions on getting around this problem are appreciated.
Thanks
 

bugman

Super Moderator
#5
Some data transfered into SPSS as 0E-9 or 0E-10 and shows in the toolbar writing space as .0 - could this be a problem?

Thanks
I don't use spss, but it sounds like this is the problem.

try adding in a constant and see if it will run the model.
 
#7
Hello, could you kindly help me look into this please?

here is an urgent appeal for help:

In ANCOVA,

Can this model be correct : One dependent variable; One categorical independent variable(two levels)+ two independent continuous variables + one or two continuous co variates?

Many thanks!
 
#8
I'm trying to run a Factor Analysis on a large environmental data set with a combination of data types, and i'm having some difficulties. I get a Warning message listing a number of possibilities of what could have gone wrong. One option is that I have variables with zero variance. i've run a descriptive statistics listing variance and eliminated from my analysis those variables with zero variance listed. I've also run a scree plot and found some variables giving null data, so I've eliminated those from my factor analysis, still no luck. I'm not sure what to try next?! Any idea on what might be causing this? I'm using SPSS 19 with over 300 variables. The data was imported from Excel. Any suggestions would be wonderful!
Could you share info on how you overcame this issue ?