Part SOCIAL And HeALTH fACTORS THAT ImPACT AgIng | 3 | 178 Chapter as what control dimension is measured, the ages and other characteristics of the sample, the specific domains examined, and the study designs. many of the studies about control and aging have used cross- sectional data, in which age and cohort differences are confounded, thus limiting conclusions about direct age-related changes. Longitudinal studies are helpful to move beyond the limitations of cross-sec- tional designs, and there is a good deal of evidence for longitudinal stability in perceived control into old age ( gatz & Karel, 1993; grover & Hertzog, 1991; Lachman, 1985, 1986 ). However, findings have been mixed, as some studies report declines in perceived control late in life (Rodin & Langer, goal relevance maybe an important contributing factor for the maintenance of control beliefs into old age. for example, Brandtstädter and Rothermund (1994)
proposed a model where sense of control is maintained in later adulthood through shifts in the subjective importance of developmental goals. The degree to which perceptions of control within a particular goal domain affected an individual’s general sense of control depended on the personal importance of that domain, and losses of control within a goal domain affected general perceptions of control to a lesser degree if the importance of the respective domain was downscaled within the same longitudinal interval. Within Brandtstädter’s (1990) model, assimilative forms of control decline with age, whereas accommodative forms of control show increases with aging. These patterns are similar to those found by Wrosch and colleagues (2006) guided by the lifespan theory of control. They reported that primary control strategies remain relatively stable across adulthood, but are more likely to be replaced by secondary control strategies in later life when older adults are faced with greater obstacles to goal attainment.
Other longitudinal investigations have documented a mix of gains and losses in control beliefs. for example, Lachman et al. (2009) found evidence for changes in the sense of control overtime in a national sample of adults in the United States studied over a year period. The year, cross-sectional differences mapped directly onto the year period change data for many of the control dimensions. Average patterns of change showed both gains and losses across different dimensions and domains of control, and these patterns also varied by age cohort group. Those in midlife looked particularly strong in terms of reporting the lowest levels of perceived constraints and greatest declines in perceived constraints overtime. In contrast, those in later life not only experienced increases in perceived constraints but also declines in health control. Thus, adulthood is characterized by a combination of ups and downs in the sense of control across different domains of life. Those who had a more adaptive personality profile (e.g., high in agreeableness, low in neuroticism, better quality of social relationships, better health, and higher cognitive functioning were more likely to maintain or increase control beliefs in general and in multiple domains. It is desirable to maintain a favorable balance of gains to losses in perceived control across life domains ( Baltes et al., 2006 ). Or, as suggested by Krause (2007) , what maybe important is to maintain control in the domains that are most meaningful or central for the individual.
In addition to mean levels, considering and modeling intraindividual variability and within-person change has received increased attention as evidence of the processes involved in psychological adaptation (
nesselroade & Salthouse, 2004; Sliwinski et al., 2003 ). Based on repeated assessments within days or across days, intraindividual variability over the short term has important predictive value for aging-related outcomes (e.g., fluid intelligence, mortality martin & Hofer, 2004 ). Some studies suggest that older adults show greater intraindividual (within-person) variability in cognitive performance domains ( Hultsch et al., 2002; nesselroade & Salthouse, 2004 ), but less in the affective domains ( Röcke et al., Although much of the work on intraindividual variability has focused on cognitive and affective functioning, a few studies have shown that locus of control operates not just as a stable individual difference variable, but also has an important dynamic aspect ( eizenman et al., 1997; Roberts & nesselroade, 1986 ). The degree of consistency of control beliefs is as important, if not more so, than the level of the beliefs ( eizenman et alas variability in control beliefs was found to predict mortality to a greater degree than level of control. eizenman et al. (examined weekly fluctuations in general control beliefs over 25 occasions for seven months in a sample of older adults. Significant within-person fluctuations in control beliefs were found, and, importantly, these fluctuations were associated with mortality five years later. more work is needed to examine how variability in control is linked with behavioral and physiological outcomes. One way of accomplishing this would be through the use of measurement burst designs (
nesselroade & Salthouse, 2004
). These designs gather estimates of intraindividual variability within a longitudinal design by nesting daily diary measurements of control beliefs (e.g., assessing control beliefs each day fora series of consecutive days) to capture intra-individual variability within long-term longitudinal assessments of behavioral and physiological outcomes. This approach has great potential for representing the dynamics of the aging individual and addressing the relationships between fluctuations in daily control beliefs and long-term behavioral and health outcomes as people age. given the benefits of a high sense of control for affect and action, whether or not veridical Thompson, 1999 ), a decline in perceived control
SOCIAL And HeALTH fACTORS THAT ImPACT AgIng Chapter THe ReLeVAnCe Of COnTROL BeLIefS fOR HeALTH And AgIng |