The LSN Network Study
Principal investigator: Theo van Tilburg
Introduction
As a part of the "Living arrangements and social networks of older adults" reserach program, 673 of the respondents of the Main Study participated in a Network Study. This study, a so-called multi-actor study, was constituted as the data collection of a research program into "Reciprocity of social support in the network of personal relationships of older adults".
The aim of the Network Study is to collect more data about the features of the networks of older adults, and to conduct research into changes in the characteristics and the functions of the network. Three differences between the Network Study and the Main Study are striking. First, the questionnaire of the Network Study is more detailed about the supportive exchanges between older adults and their network members and about other characteristics of their relationships. Second, the Network Study focusses on changes in a fixed selected part of the network, while the LSN Main Study and the successive LASA survey focus on a changing personal network. Third, (a selection of the) network members participated in the Network Study as respondents. Although the network is still ego-centric, adding information obtained from network members makes it possible to analyze the structural features of a "full" personal network, in addition to structural characteristics like the network size that can be assessed from "star" networks.
Figure 1 gives an example of a "star" network. The network consists of the "anchor" of the network, eight persons and their relationships with anchor. The anchor is the respondent of the NESTOR-LSN main survey and graphically represented as the middle of a star. In the interview of the Main Study the anchor of this example mentioned at least eight persons with whom he is in touch regularly and who are important to him: his wife, his son, who is a member of the household, his daughter, one of his brothers, a neighbor, a colleague, someone who is known from an organization and a friend (see LSNa047). If he mentioned more than eight persons, the eight network members were selected with whom contact is most frequent (see LSNa047). When we gather information about the supportive content of these eight relationships, we have essentially the same information as the data in LSNa047.
Figure 2 gives an example of a "full" network. The network consists of the "anchor" of the network, eight persons and their relationships with anchor, and their mutual relationships. The respondent was asked cooperation for the Network Study, and if permission was obtained, the respondent was asked to give the full names and addresses of the selection of the network members. The existence of the relationships between network members was assumed (e.g. between anchor's wife and anchor's daughter) or was asked in the interview (e.g. between a daughter and a friend of the anchor) (see LSNa056). The nine persons ("points") and the 22 relationships between them ("lines") are referred to as the network, although it is a sampled part of the whole network when the anchor has more than eight network members. When we investigate the supportive exchanges in these 22 relationships, we have more data than is present in LSNa047: we have the supportive exchanges in a full network (we assume that non existing relationships have no supportive content). In this network, the respondent of the face-to-face survey has on the one hand the same position as the other network members, and, on the other hand, is the anchor of the network: the network is delineated by anchor. Furthermore, we have information about the content of each the relationships from two sources: for each relationship pair we have the answers of both participants in the relationship about the characteristics of the relationship.
Another way to present the network is with the help of an adjacency matrix (Table 1). The first column gives the names of the nine persons in the network, and the second gives a short description of the type of the relationship between the eight members and anchor. In the matrix in the right part of the table, the existence of a relationship is presented as 1 (equal to a line in figure 2), the absence of a relationship as 0, while the diagonal is empty.
Table 1. Adjacency matrix of the example network
| Name | Description | # | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
| Mr. J. Jones | anchor | 0 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | |
| Mrs. A. Jones-Robertson | wife | 1 | 1 | 1 | 1 | 1 | 1 | 0 | 1 | 1 | |
| Mr. B. Jones | son | 2 | 1 | 1 | 1 | 1 | 1 | 0 | 0 | 1 | |
| Mrs. S. Jones | daughter | 3 | 1 | 1 | 1 | 1 | 1 | 0 | 0 | 1 | |
| Mr. D. Jones | brother | 4 | 1 | 1 | 1 | 1 | 0 | 0 | 0 | 1 | |
| Mr. R. Merchant | neighbor | 5 | 1 | 1 | 1 | 1 | 0 | 0 | 0 | 0 | |
| Mrs. M. Flowers-Clark | colleague | 6 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | |
| Mr. P. Barnes | organization | 7 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | |
| Mrs. Y. Doll-Boot | friend | 8 | 1 | 1 | 1 | 1 | 1 | 0 | 0 | 0 |
All nine persons in this network received a questionnaire by mail. The anchor is asked to answer questions about his relationships with eight network members, his wife about seven relationships, and so on. In total in 1992-1993 we sent questionnaires to 4,264 respondents with 17,396 mutual relationships (Actually, half of these relationships exist, the other half concerns the mirrored relationship). There are at least three reasons to ask questions not only to the anchor, but to the network members too.
The first reason is that we are extremely interested in the circumstances of the people in the network. They are important if we are to understand why some elderly people give or receive so much support, and others so much less. If for example a friend gives a great deal of help, it might be because he is retired and has plenty of time to spare. If he also happens to live in the neighborhood, then he has all the more chance to give help and support. A daughter can be in a situation where she can not give much help at all, for example because she has a job and young children at home. If she also lives an hour drive away, it will be even harder for her to visit regularly and be much of a help. In the discussion now being conducted in society, a daughter is sometimes all too quickly assumed to be always able to help. We hope the data gathered here will provide a more discerning view of the options open for example to daughters to give help and support.
It is also important for us not to work from the assumption that it is always the older adult who needs help. This is why we persistently ask about the help and support the olderadults give to the people in their network as well as the help and support they receive. We did this in the face-to-face interview as well as in the questionnaire we sent by mail.
We also posed a number of questions about what network members think about giving and receiving support. For example, how important they feel it is to give support. We asked if they felt it was only important to give help and support to their relatives, or also felt it was important to give help and support to their friends, whether male or female. We think people have different ideas about this, which might be relevant to the help and support that is given. It is clear that data about the ideas of the network members can not be gathered from the elderly respondents; we had to ask the people themselves.
A second reason to send network members a questionnaire is that we also want to know whether they know the other people in the network, and whether they ever help these other people. The pattern of giving and receiving help might be far more complicated than we tend to think. Most people are familiar with the following situation involving a son and daughter-in-law: "If I (daughter-in-law of the elderly respondent) get the housework finished now, then you (son of the elderly respondent) can go see your father." We do not really know how often this occurs, or how often it is relevant to the help given to the elderly person, in this case the father. Barely any research has been conducted into this kind of pattern of indirect help from the daughter-in-law via the son to the father. Of course patterns of this kind do not only occur with an elderly person and his son and daughter-in-law. They can also occur in the relationships between an elderly person and various of his neighbors (see Figure 3). We are very curious as to whether we will be able to uncover information of this kind by way of this study.
A third reason to send questionnaires to the people in the networks is that each individual has his own idea of what happens in a personal relationship. If for example we ask how often one person helps another, a modest helper will be quick to say it isn't that often, "I like doing it, and I am helping out a bit." However, for the person who is on the receiving end, even the most modest amount of help might be extremely valuable. In that case, there is a chance that if we ask how much help was received, the amount might be overestimated. In a pilot study we conducted in 1991, there did indeed appear to be sizable differences between the amount of help one person reported giving and the amount the other person reported receiving (van Tilburg, 1992a). These findings confirm the results of an earlier study by Antonucci and Israel (1986). By comparing the information we got from the two persons, we are now better able to estimate how much help was "really" given. By comparing the responses given by the two persons, we hope to better understand the data gathered about the help given and received within personal relationships.
Futhermore, the availability of data from two sources (both persons in the relationship) makes it possible to compare the proxy data on demographic charcteristics of network members as given by the anchor of the network and these data from the network members. This offers opportunities for methodological research into the reliability of survey data (see for examples Pfenning, Pfennig & Mohler, 1991; Schenk, Mohler & Pfennig, 1992).
Updated 06-12-2000
