Contents - Index

NETWORK>EGO NETWORKS>HONEST BROKER>LONGTITUDINAL

PURPOSE Measurers change in brokerage at two time points.

DESCRIPTION Given a directed network at two distinct time periods this routine looks at the changes in the brokerage of trust-like relations over time. 

PARAMETERS 
(Input) Network dataset
Name of file to be analysed.This must be a multirealtional dataset in which the first level will be time T1 and the second level will be time T2. Data type: Multirelational Digraph.
 
(Output) Simple statistics
Name of UCINET datafile containing scores as described in the first table below. Default = <filename>-SimpHBI

(Output) Deatiled statistics
Name of UCINET datafile containing scores as described in the second table below. Default = <filename>-DetHBI

Ego must send tie to alters as well as receive
If this box is checked then ego networks are defined by reciprocated ties otherwise the ego network consists of in-coming ties only.

LOG FILE Two tables giving statistical summaries of the changes over the time period. The first table gives the following for each ego:

T1Size: number of alters at time T1
T1Opps: number of pairs of alters
T1Brok: number of brokerage ties that is number of edges between alters
T1Pct: T1Brok divided by T1Opps as a percentage

These are then calculated and reported at time T2 

The last four columns of the first table give the differences by taking the scores at T1 and subtracting off the scores at T2.

The second table gives the following:

OppsKept: number of pairs of alters present at time T1 and T2
OppsLost: number of pairs of alters present at T1 but absent at T2
OppsGain: number of pairs of alters absent at T1 but present at T2
OppsNeve: number of pairs of alters not present at time T1 nor T2 
BrokKept number of alter alter ties present at time T1 and T2
BrokLost number of alter alter ties present at T1 but absent at T2
BrokGain number of alter alter ties absent at T1 but present at T2
BrokNeve number of alter alter ties absent at T1 and T2

The last 4 columns give each of the Kept, Lost, Gained and Never as a percentage by taking the Brok score and dividing this by the corresponding Opps score and multiplying by 100. 



TIMING O(N^3)

COMMENTS None

REFERENCES  Ahjay