Interviewing cognitively impaired older adults: How useful is a Cognitive Interview?

Abstract

This research examined whether an Enhanced Cognitive Interview (ECI) and a Modified Cognitive Interview (MCI) improved 75- to 96-year-old adults’ recall for a video-taped event. Specifically, we examined the effectiveness of these interviews in relation to a Structured Interview (control), and compared the performance of older adults who scored high or low on the Mini-Mental State Examination (MMSE), which is a test to assess cognitive functioning. Participants with low MMSE scores recalled fewer correct details, and were less accurate than those with high scores. Nevertheless, both high and low MMSE groups reported substantially more correct information about Action, Person, Object, and Surrounding details with the MCI and ECI than with the SI. No corresponding increase in the amount of incorrect and confabulated information was found. Furthermore, recall accuracy for both groups was greater with the ECI and MCI than with the Structured Interview. Overall, findings suggest that older individuals, including those with reduced cognitive ability can benefit from the use of Cognitive Interview protocols.

Notes

¹Several versions of the MMSE exist. One key difference in these versions is whether individuals are asked to spell the word “world” backwards, perform serial subtraction using the number 7, or perform both of these activities (see Ganguli et al., 1990, for more information about these differences). The MMSE used in the present study incorporated the “world” option.

²Several other explanations for educational differences in MMSE scores may exist (Crum et al., 1993). For example, a limited education (and the lifestyle factors that often accompany it) can increase the risk of cognitive impairment in old age (Baker et al., 2002; Bassett & Folstein, 1991; Crum et al., 1993; DeRonchi et al., 1998; Jorm, Scott, Henderson, & Kay, 1988). It is also possible that inherent cognitive deficits or disorders are responsible for both educational attainment and MMSE performance. In this case, poor performance on the MMSE does not necessarily result from low education. Instead, low education and poor MMSE performance are both due to an underlying disorder (Baker et al., 2002). The practice of using education level to adjust MMSE scores is also criticised on the grounds that formal education does not necessarily reflect overall educational attainment (Baker et al., 2002).

³If cells in a factorial design contain unequal numbers of scores, main and interaction effects will not be independent of one another (Howell, 1997). To adjust for the presence of unequal Ns in conditions, it is necessary to base Tukey's HSD post hoc tests on estimated marginal means and their associated standard error values (Tabachnick & Fidell, 1996). In addition, the formula for determining the critical difference with Tukey's HSD was altered such that the harmonic mean was used in place of N (cf. Hayes & Delamothe, 1997; Howell, 1997).

⁴According to Norman and Shallice's (1986) model of the control of action, the supervisory attentional system is involved in controlling of non-routine, “intentional” actions that require “higher-order” cognitive control. The contention scheduling system, on the other hand, explains how schemas are stored and activated for routine tasks.

⁵Shizotypy refers to a non-clinical manifestation of biological and cognitive factors that, when seen in a more extreme form, cause clinical psychotic symptoms (e.g., delusions, paranoia; Langdon & Coltheart, 2001).