FCOR - correlation of pooled subtypes

GoogleCodeExporter commented 9 years ago

Hello, 
I am using the FCOR program, and I wondered if there is an option to calculate 
the correlation "mother - offspring" (boys & girls together) in addition to the 
separate correlations: "mother-son", "mother-daughter". 
In other words, how can I pool together two specific subtypes and calculate 
their overall familial correlation?
Thank you,
Einat

Original issue reported on code.google.com by einatgra...@gmail.com on 4 Aug 2014 at 10:50

GoogleCodeExporter commented 9 years ago

Currently, FCOR dos not have the option to compute the correlation between a 
main type and a subtype.  In case your data consists of nuclear families only, 
you can set all offsprings' sexes to be the same, either female or male, then 
the result for mother-son (or mother-daughter) will be what you want.

Original comment by opensage...@gmail.com on 3 Sep 2014 at 9:21

GoogleCodeExporter commented 9 years ago

Hello,
I am running FCOR analysis on nuclear families (or pedigrees), and  I am
using the following script:

fcor, out=  "BMI"
{
trait="BMI"
interclass_weight = "uniform"
intraclass_weight =  "uniform"
standard_error = "true", conservative, pairs
type = "both", tabular
sex_name ="true"
homogeneity_test = "true", all
generation_limit = "2"
}

fcor, out=  "BMI"
{
trait="BMI"
interclass_weight = "pair"
intraclass_weight =  "pair"
standard_error = "true", conservative, pairs
type = "both", tabular
sex_name ="true"
homogeneity_test = "true", all
generation_limit = "2"
}

I expect to see a difference in the results because of the different
weighting, but I get an error:

%FCOR-I: Invalid parameter for fcor_analysis : interclass_weight
         Skipped...
%FCOR-I: Invalid parameter for fcor_analysis : intraclass_weight
         Skipped...

and the correlations are identical.

Where is my mistake?

I am using S.A.G.E. 6.3 Windows.

Thanks a lot,

Einat

Original comment by einatgra...@gmail.com on 26 Oct 2014 at 11:40

jcastib / opensage

FCOR - correlation of pooled subtypes #6