*****  To join INSNA, visit  *****

Dear all

A quesiton about a measure of heterogeneity:

I have data from ego networks and want to develop a measure of heterogeneity 
of alters.  thought of IQV but not quite sure if it fits data well (although 
sure it can be used perhaps in variation form).

Data as follows:

egos give: age (category); sex (male/female); education, location and role 
in organisation (these two so varied that not plausible to match them in 
meaningful way between egos).

alters give same(1)/different(2) to ego on: age; sex; education; location; 
role (in work); position (in work- i.e.  'same', 'higher', 'lower' but 
recoded to same =1 different = 2)

Ego then has 'attributes' in terms of alters' characteristics.  I.e. ego has 
x% same age alters, y% different age alters (variable is % same).

Would like, though, to find heterogeneity measure between alters of 
individual egos.  So, thought version of IQV should work but not quite sure 
what the components are etc.  Simplest explanation i found has following 
link   (pdf file so cannot cut and paste into message as equations lose 
necessary elemants.  e.g. sigma becomes a square inverted U and the infinity 
symbol disappears.) (google iqv)

Suggestions very welcomed and thanks in advance

(Additional note:  this work is from psychological perspective and focussed 
on individual differences.  Thus sample is not from within a single 
organisation but from individuals that may not have overlapping networks.)

thanks to all

bill fear 

SOCNET is a service of INSNA, the professional association for social
network researchers ( To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.