Text-Based Levels

Most of the work conducted in aging and discourse processing has examined memory at text-based lev­els. Because texts are constructed with information at these different hierarchical levels of importance, a key question is whether there are age differences in memory for these different levels. Answering this question amounts to looking for age differences between memory for main ideas and memory for details. The literature on this issue is large and com­plex (Johnson, 2003; Wingfield & Stine-Morrow,

2000) . In general, the data indicate two important points. First, when text is clearly organized, with emphasis on structure and the main ideas, older adults are similar to younger adults in recalling more main ideas than less important details.

However, in addition to these competencies in text processing, older adults exhibit observed defi­cits under specific conditions. Studies show that older adults are adversely affected by rapid presenta­tion, highly unpredictable or unorganized material, and material that is dense in propositions (Hartley, Stojack, Mushaney, Annon, & Lee, 1994; Stine, Wingfield, & Lindfield, 1995). Older adults may be at a disadvantage when presented with text at speeds geared to younger adults. Indeed, when the speed- of-presentation variable is removed and participants can pace themselves, age differences are eliminated (Wingfield & Lindfield, 1995; Stine-Morrow et al.,

2001) . In addition, the more predictable and mean­ingful the text, the better the older adults perform

Attention and Memory 209

even at rapid rates (Gordon-Salant & Fitzgibbons, 1997; Wingfield, 1996). Finally, high verbal ability reduces age differences in overall recall during rapid presentation (Wingfield & Stine-Morrow, 2000). Another area of difficulty for older adults appears when the text becomes propositionally dense (an increase in the number of propositions) and less familiar. In this case older adults have more difficulty than younger adults do in identifying or inhibiting less important details (Hartley, 1993; Wingfield & Stine, 2000).

An important aspect of text variables is that they may interact with personal characteristics of the reader to create differences in memory for discourse. An example of this interaction involves situations in which people hold opinions or have knowledge about a topic and are then presented additional written information. In terms of people’s ability to remember this new information, does it matter whether the new information agrees or dis­agrees with what people already know?

To answer this question, Rice and Okun (1994) tested older adults’ memory for accurate informa­tion about osteoarthritis that either contradicted their previously held false beliefs or affirmed their previously held accurate beliefs. All of the par­ticipants reported having osteoarthritis for at least 2 years. Rice and Okun found that older adults recall and recognize information that contra­dicts their previously held beliefs less accurately than they recall and recognize information that affirms their initial beliefs. These findings point out that individuals who educate older adults must be careful to identify misconceptions that they may bring into the situation. However, Rice and Okun also found that if the disconfirm — ing material was stated explicitly, older adults were better at remembering the information later. Thus, explicitly stating that certain information is often misunderstood, and then stating the correct information, may help older adults remember the accurate information.