Scoring of Data Entered

This page provides the scoring associated with a network of friends and family members.  A positive score indicates a positive influence, a negative score indicates a negative influence.  The larger the value, the larger the extent of the influence. 

The procedure for scoring includes the following steps:

1. The total number of unique people ranked anywhere in the survey is calculated by examining the unique contacts ranked in response to any of the three components of the survey.  For example, if the client has ranked the relative importance of 3 unique friends, the relative frequency of contact with 5 new unique friends and listed the orientation of two new unique friends, then the total number of friends ranked any where in the survey is 3+5+2=10.  Note that in any component of the survey a different total number of friends are typically ranked.
2. Any individual not ranked in the survey components is assigned the rank equal to the total number of people ranked anywhere in the survey.  We are assuming that people not mentioned in the survey have the least rank order in the survey.
3. Any individual not listed in response to the orientation survey is assigned a value of 0 for orientation.  We are assuming individuals without a specific orientation have no influence.
4. The rank order of frequency of contacts and importance of contacts are turned into a probability estimate using the following table set up for rank orders assigned to up to 9 people.  In most responses to the survey clients do not assign or think through more than 9 people.  If the total number of friends exceeds nine, the rank order for 9 persons can be used as the individuals with higher ranks will have near zero probabilities.  The use of the table can be illustrated by noting that a person that ranks his friend as 2nd in relative order of contacts among  7 friends is assigning a probability of contact equal to 0.2276.
    

   Rank	One Plus Total Number of Friends Ranked
   	1	2	3	4	5	6	7	8	9	10
   1	1.0000	1.0000	0.7500	0.6111	0.5208	0.4567	0.4083	0.3704	0.3451	0.3269
   2		0.0000	0.2500	0.2778	0.2708	0.2567	0.2417	0.2276	0.2181	0.2114
   3			0.0000	0.1111	0.1458	0.1567	0.1583	0.1561	0.1546	0.1536
   4				0.0000	0.0625	0.0900	0.1028	0.1085	0.1123	0.1151
   5					0.0000	0.0400	0.0611	0.0728	0.0806	0.0862
   6						0.0000	0.0278	0.0442	0.0552	0.0631
   7							0.0000	0.0204	0.0340	0.0438
   8								0.0000	0.0159	0.0273
   9									0.0000	0.0128
   10										0.0000
   Table 1:  Converting Ranks to Probabilities

5. The net impact of an individual is calculated by multiplying the probability of contact by the probability of importance of the contact and by the orientation of the contact.  For example, if the contact is ranked first in frequency of contact among six people then the probability of contact is 0.4083.  If the same contact is ranked 5th in terms of relative importance then the probability of importance is 0.04.  If the contact's orientation is anti-drug use (orientation=1), then the impact of this person is estimated to be 0.4083 * 0.004 * 1 
6.  The net impact of the entire network is calculated as the sum of the impact of each contact.

To facilitate the scoring procedure, we have created a web page where information may be entered and the network impact calculated.  If you have already entered the necessary frequency of contact ranks, the relative importance ranks and the orientation of support score, then the network's impact on the index case is provided as below.  Please enter the code for the information you have provided so that it can be analyzed and results posted here:

An Example

The following shows how sample data are analyzed.  Table 1 provides the raw data.  Notice that for some individuals, information about frequency of contact, relative importance of contacts or orientation is missing

Table 2:  Original Data

In the next step, missing rank data are assigned the lowest rank in the responses.  Missing orientation values are assigned value of zero.

Table 3:  Estimation of Missing Ranks and Orientation Values

In the next step, data in Table 1 is used to transform ranks to probabilities. 

Table 4:  Estimation of Probabilities from Ranks

In Table 4 above, the impact of each person is calculated as the product of the importance of the person, the probability of contact with the person and the orientation of the person.  In the final analysis, the impact of the entire network on the person is calculated as the sum of the impact of individual persons (see Table 5). 

Table 5:  Overall Impact of the Network