effects of age and auditory degradation on memory - Semantic Scholar

3 downloads 0 Views 62KB Size Report
Partially masking words by a babble noise has an age-dependent, adverse effect on memory. On the other hand, temporally distorting the words, so that they are.
EFFECTS OF AGE AND AUDITORY DEGRADATION ON MEMORY Antje Heinrich University of Toronto [email protected]

Bruce Schneider University of Toronto [email protected]

Abstract The study compares the effects of auditory masking and auditory distortion on memory for words in younger and older adults. Partially masking words by a babble noise has an age-dependent, adverse effect on memory. On the other hand, temporally distorting the words, so that they are equally-often misheard as when they were partially masked, has no effect on memory unless the words are presented at high-intensity levels. It is argued that masking places a greater demand on processing resources than does temporal distortion, and that it is this greater load on processing resources that is responsible for poorer recall when words are presented at moderate sound pressure levels. Presenting temporally-distorted words at higher intensity levels may increase processing demands because the combination of temporal distortion and high intensity produce a rollover effect.

A number of studies have shown that perceptual and cognitive processes tend to be largely correlated in older adults (Lindenberger & Baltes, 1994; Baltes & Lindenberger, 1997). This leads to the suggestion that changes in perception may cause changes in cognition. Some studies seem to support this notion by showing that when younger and older adults are equated for perceptual differences, cognitive differences tend to decline or even vanish completely. Evidence for this can be found in a study by Humes and Christopherson (1991) who adjusted the level of masking noise at each frequency for their young listeners such that their threshold was elevated to match that of presbycusics. By doing so they found that their young subjects performed equally to older subjects on a monaural word identification task. Further support for this notion also comes from Schneider, Daneman, Murphy, and Kwong See (2000) who presented stories or lectures in quiet or noise and asked their subjects to answer questions about their content. When the listening situation was altered to produce a comparable perceptual stress in both age groups (equal difficulty in hearing individual words), age-related performance differences in cognition were minimized. However, equating for perceptual differences does not always result in comparable memory performance. Murphy, Craik, Li, and Schneider (2000) tested young and old participants in a paired-associate memory task after making it equally difficult for both groups to hear the actual words. In this study, they were not able to eliminate age-related performance differences

completely. Hence, adjusting hearing conditions across age results in a similar memory A verage P ercentage of C orrectly R ecalled W ords as a Function of A ge and P erceptual M anipulation

Average Percentage Correct

100

Young Q uiet

90

Young N oise

80

O ld Q uiet

70

O ld N oise

60 50 40 30 20 10 0 1

2

3

4

5

Serial Position

Figure 1. Memory recall as a function of age and perceptual manipulation; the vertical bars depict standard error of the means. Note. From “ Comparing the effects of aging and background noise on short-term memory performance”, by Murphy, D.R., Craik, F.I.M., Li, K.Z.H., and Schneider, B.A., 2000, Psychology and Aging, 15.

performance on some but not on all cognitive tasks. The extend to which cognitive rather than perceptual changes are responsible for remaining differences across tasks needs to be experimentally explored. The study presented here is based on Murphy et al. (2000). The experimental task was a paired-associates memory task. On a trial five word pairs were presented aurally to the subject. Following a warning tone the first word of one of the pairs was presented again and the participant was asked to retrieve the second word of the pair. Murphy et al. manipulated perception by presenting the word pairs either in quiet or in background babble. They tested both young and older adults on the memory task. Murphy et al. found that in the quiet condition, where no explicit perceptual stress was applied, younger participants showed a significantly higher memory performance for the first three word pairs as compared to older subjects (see Figure 1). When perceptual stress in form of a twelve speaker multi-talker babble was added to the task, memory performance of young adults declined to look very similar to that to old adults in quiet. However, when the equivalent background babble was added to the word presentation of the older adults, their memory performance further declined to even lower levels of retention. In order to explain the results, the model shown in Figure 2 is used. Before information can be recalled, it has to be perceived and encoded. Suppose, perception and cognition depend on

a common pool of processing resources, and there are several conditions that may delete those resources. Research results suggest that aging may be one of the factors that severely restrict processing resources. This Perception would explain the age-related drop in memory performance (Noise, Jitter) depend on a frequently seen in these kinds of common pool of experiments because fewer resources resources are available for encoding. In the Murphy et al. Encoding into Memory study, young adults tested in noise exhibited a drop in Aging memory performance very similar to that seen accompanying aging. Here, noise as a perceptual factor drew Recall on the same pool of processing resources, hence withdrawing Figure 2. Perception and Cognition in a memory task resources from encoding and so draw on the same pool of processing resources. leading to a similar recall Aging may further restrict resources. performance as seen under aging. The link to the experiments presented here was our wish to model more realistically, age-related changes in auditory function to see whether this would affect memory performance in a way similar to babble. One of the changes frequently reported to occur with age is an increase in temporal distortion to the auditory signal. This is believed to affect speech recognition (Humes & Christopherson, 1991; Dreschler & Plomp, 1985). In order to simulate temporal distortion we introduced a random time delay to the word stimuli. The time delay, called jitter, was quantified in RMS levels. An argument in favor of the hypothesis that we should be able to replicate the results found using background noise by using jitter is the fact that we added an adequate amount of jitter to the words that resulted in the same decrease in word identification as found under noise. Controlling for perception was possible because Murphy et al. had measured perceptual performance alone by playing the words embedded in noise to their participants, asking them to repeat the words immediately, and measuring the percentage of correctly repeated words. Following the same procedure with temporally altered words, we found the amount of jitter that resulted in the same percentage of correctly perceived words. Thus, if the memory deficit depended solely on the accuracy of word identification, we should yield similar results which both methods. However, there are also profound differences between both kinds of perceptual manipulation, which leaves the possibility that they will not affect memory in a similar way. Whereas in noise the subject has to pick out the signal from a competing background, there is no background distraction in the jitter condition, rather, the signal itself is distorted. Method A 2x2 independent group design was used. In each condition fifteen young and fifteen older subjects were tested. Our young participants (19-25 years of age, all with normal

Average Percentage of Correctly Recalled Words as a Function of Age and Jitter 90

Young 40 dB Old 40 dB

80

Average Percentage Correct

Young 50 dB 70

Old 50 dB

60

50

40

30

20

10

0 1

2

3

4

5

Serial Position

audiograms) were undergraduate students from the University of Toronto at Mississauga, the older subjects (65 years and older, all with good hearing for their age group) were volunteers Figure 3. Memory recall as a function of age and presentation level; the vertical bars depict standard error of the means

from the local community. Two groups of fifteen older and fifteen younger adults, respectively, were tested on presentation levels of 40 and 50 dB above individual speech threshold. The speech thresholds were determined by obtaining each participants babble threshold. Moreover, both age groups were equated for perceptual stress by determining the amount of temporal distortion for each group that led to the same percentage of correctly perceived words. The extent of temporal distortion that equated perceptual performance for both age groups was higher for young participants on both presentation levels. Results and Discussion Figure 3 displays the results obtained for the paired-associate memory task using temporally distorted words for both age groups and two presentation levels. Conducting a repeated measures analysis of variance (ANOVA) for all conditions (Serial Position (5) X Age (2) X Presentation Level (2)) confirms a pronounced serial position effect (F = 46.76, p = .0001) such that the forth and fifth word pair were remembered more easily than the first three pairs. There also was a significant age effect (F = 20.01, p = 0.0001) confirming the observation that young adults remembered the word pairs better than did older adults. There was no main effect of

presentation level (F < 1) but the interaction between age group and presentation level reached significance (F = 3.9, p = 0.05) indicating that the manipulation did not have any effect on the memory performance of young participants. For older participants, on the other hand, presenting the word pairs on a higher level led to a marked drop in memory performance. The 50dB SL presentation level condition is of special interest because this is the level used in the Murphy et al. study. As can be seen, both perceptual manipulations did not affect memory in the same way; in the noise condition, young adults memory performance dropped to the level of that of older adults' whereas in the jitter condition, memory performance did not show this drop for young subjects. This result leads to the conclusion that pulling out a word from a competing background has a more profound effect on young adults' memory than distorting the signal itself. Hence, it was not possible to mimic old participants' memory performance by distorting the words themselves. Apparently, it is not the difficulty in correctly identifying the word itself that penalizes participants in the memory task. One possible explanation could be that attentional processes are needed when words are presented in a background babble but not when presented in quiet. Therefore, presenting temporally-distored words does not draw on attentional resources and therefore does not affect memory as much as when attentional resources have to be deployed to identify the words in a background noise. The persistence of an age effect even when word identification difficulty is controlled suggests that there is a memory component that declines independently of perceptual processes and thus cannot be recovered by equating for perceptual stress. The other interesting outcome of this study is the drop in memory performance for temporally-distorted words in older adults when these words are presented at a higher sound pressure level. This result is most likely due to a rollover effect, a syndrome that is sometimes observed in patients with central auditory disorders. Word identification is normal for these individuals for intermediate sound pressure levels but decreases at higher levels. We would expect rollover effects in patients with temporal processing disorders. To see how this could happen, recall that sound information is encoded in at least two ways. First, the number of active neurons and their rate of firing conveys the information contained in the sound envelope (e.g., the cadence of speech). The higher the intensity level of the incoming stimulus the more neurons fire. Unfortunately, this mechanism works properly only up to a certain intensity level. When the intensity of the stimulus reaches a certain level, the neurons will saturate so that information conveyed by the envelope of the sound is lost. The second important mechanism is frequency coding. The frequency of the stimulus is, at least in part, encoded by phase locking. Hence, the firing of the neurons is locked to the frequency of the auditory stimulus such that the frequency of the stimulus and the frequency of the neural firings are synchronized with the degree of synchrony declining with increasing frequency. Now assume that in this study both frequency and intensity coding were compromised in older adults. Temporal coding was compromised as a result of age-related changes to the auditory system (which results in a higher internal jitter level), and because in addition to that we temporally distorted the stimulus. Intensity coding was reduced because the overall presentation level of the speech material for old adults averaged 73 dB SPL, which was about 8 dB higher than the presentation level for young adults. Thus, given that intensity coding works only up to a certain level, the auditory neurons of the older subjects may have been saturated to a higher degree than those of younger subjects so that they could not encode intensity modulation as well younger participants. Thus, saturation effects combined with temporal distortion could have increased the processing load for the older participants at the higher sensation level, thereby reducing memory performance.

Looking at future research there are two interesting facts to explore. First of all, it would be interesting to see whether we are able to replicate the influence of the rollover effect on memory for young adults. Secondly, having seen that jitter did not draw on processing resources in the same way babble did leaves us with the question whether we can find some other manipulation that helps us mimic older adults' memory performance in younger adults. Another common way of accessing processing resources is by using a divided attention paradigm. It would be interesting to see whether we could mimic the assumed age-related reduction in processing resources by dividing younger adults' attention in the memory task. References Baltes, P.B. & Lindenberger, U. (1997). Emergence of a powerful connection between sensory and cognitive functions across the adult life span: a new window to the study of cognitive aging? Psychology and Aging, 12, 12-21. Dreschler, W.A. & Plomp, R. (1985). Relations between psychological data and speed perception for hearing impaired subjects. Journal of the Acoustical Society of America, 78, 1261-1270. Humes, L.E. & Christopherson, L. (1991). Speech identification difficulties of hearing-impaired elderly persons: The contributions of auditory-processing deficits. Journal of Speech and Hearing Research, 34, 686-693. Lindenberger, U. & Baltes, P.B. (1994). Sensory functioning and intelligence in old age: A strong connection. Psychology and Aging, 9, 339-355. Murphy, D.R., Craik, F.I.M., Li, K.Z.H., & Schneider, B.A. (2000). Comparing the effects of aging and background noise on short-term memory performance. Psychology and Aging, 15, 1-12. Schneider, B.A., Daneman, M., Murphy, D.R. & Kwong See, S. (2000). Listening to discourse in distracting settings: The effects of aging. Psychology and Aging, 15, 110-125.