Error massage in Factor analysis using FACTORAN

Hi, I am doing factor analysis in Matlab using factoran. I have 5 variables and considering initial eigen value criterion (eig>1) to extract significant factors. My data shows, I need to extract 3 factors. I am using following matlab code:

[Lambda,Psi] = factoran(data,3,'normalize',0) % data is already normalized

However, its giving following error massage,

"The number of factors requested, M, is too large for the number of the observed variables."

i went to the factoran code in matlab and says, to analyze data with factoran code, following conditions need to be satisfied:

(m < d) and (df > 0)

where m= number of factors= 3 (my case) d = number of variables = 5 df= .5*((d-m)^2 - (d+m));

putting our values give df = -2 (violation of the condition 2)

My understanding of factor analysis says that, I can extract up to 5 factors sicne I have 5 variables (number of factors = number of variables). I am just wondering whether there is a bug in the code or I am doing anything wrong. Please help>

Thanks

Answer 1

[supovic]

I understand that degrees of freedom should be positive number but I did not get why is df = ((d - m)^2 - (d + m)) / 2 ?