No abstract

Using classical test theory, 6 cognitive scales were constructed to address the need for measurement instruments that are reliable and have adequate representation of the difficulty range for both healthy older adults and patients with dementia. Internal consistency reliability estimates for the scales ranged from .92 to .98 in a probable Alzheimer's disease sample (n = 46), .79 to .95 in a healthy older adult sample (« = 131), and .93 to .99 in the combined sample. Examination of item characteristics and scale scores in the 2 samples yielded 4 findings that support the need for such specially constructed scales.Most neuropsychological measures that are used in the evaluation of dementia fall into two broad groups: (a) measures of discrete cognitive abilities that were developed for use with a normal population (e.g.,

Recognition and source memory were explored in healthy older adults, adults diagnosed with very mild dementia of the Alzheimer type (DAT), and adults diagnosed with mild DAT. Two sentence-completion tasks were used. In Task 1, half of the sentences were completed (clozed) by the participant, and half by the experimenter. In Task 2, half were participant clozed, and half were participant read (already clozed). Recognition of the cloze words and accuracy of categorizing them as participant generated or experimenter generated (Task 1) and participant generated or participant read (Task 2) were measured (source discrimination). Contrary to previous reports, the DAT groups showed the generation effect, that is, better recognition for participant-generated words than experimenter-generated words (Task 1) or read words (Task 2). Source discrimination was disproportionately impaired in the DAT groups.

The authors investigated the influence of test format on the source-memory performance of older adults (N = 128). Each participant viewed a picture and wrote a description of the scene. Then half of the participants (control group) read a text that accurately described the scene; the other half (misled group) read a text that contained misinformation. After writing another scene description, the participants were given a surprise memory test. Half were given a yes/no recognition test, and half were given a source-monitoring test. The misled yes/no participants mistakenly indicated more often than the control yes/no participants that misleading-text items were in the picture (suggestibility effect). There was no suggestibility effect for source-monitoring participants. The data are discussed in terms of the source monitoring framework.

This study enhances understanding of how self-efficacy motivates participation in physical activity, as noted in previous studies, as well as provides a new understanding of what motivates participation in social activities. The high reliability of the new measures used in this study provides evidence for further use of these measures in other contexts. It is important to note that this study further supports the use of Bandura's theory of self-efficacy for cross-cultural applications.