Researching grandparenting

Interest in grandparenting has increased considerably in recent years, particularly in Britain and the US. However, the quality of grandparent research has often been the subject of sharply critical comment. For example, Allen et al (2000) expressed concern about research that did not take account of “the total support network” (p 916). Studies which investigate the relationships between a grandparent and a favourite grandchild, or which ask college students to recount their childhood memories of their grandparents (Kornhaber and Woodward, 1981; Matthews and Sprey, 1984; Creasey and Koblewski, 1991), need to be treated cautiously since they do not fully recognise the importance of the part that parents, their children and other members of the wider family play in defining grandparents’ roles.

Allen et al (2000) also criticised research that assumed that the nature of family relationships could be captured in a single snapshot, and thus placed reliance upon cross-sectional designs rather than longitudinal studies. Amato and Keith (1991) were critical of studies that did not ensure that “social class and other variables” were controlled (p 32). They urged researchers to avoid “relying on parents’ ratings as their sole source of information” (p 33) and described convenience samples of volunteer subjects as “troubling” (p 31). Uhlenberg and Hammill (1998, p 276) commented on the fact that past research has not provided consistent findings about the variables that account for differences in grandparent behaviour. Szinovacz (1998, p 282) also criticised cross-sectional studies and the “prevalence of surveys and interviews” but seemed to approve of studies that were restricted to the study of grandparent–grandchild dyads. However, Aquilino’s (1999) study of parent–child closeness, contact, control and conflict concluded that findings are significantly affected by whether or not the information is given by the first, the second or the third generation.

Studies that fail to comply with the logic of experiments risk the accusation that their findings have been built on shaky foundations. At the same time, although critics demand large representative samples and sophisticated measures, few of them mention the problems of persuading families to take part in the research (discussed in Chapter One of this book).