This article was downloaded by: [UQ Library] On: 20 November 2014, At: 00:06 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Experimental Aging Research: An International Journal Devoted to the Scientific Study of the Aging Process Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uear20
Visual and auditory modality and suffix effects in young and elderly adults a
Susan Karp Manning & Joan Greenhut-Wertz
a
a
Hunter College and the Graduate School of the City University of New York and New York University Medical Center Published online: 27 Sep 2007.
To cite this article: Susan Karp Manning & Joan Greenhut-Wertz (1990) Visual and auditory modality and suffix effects in young and elderly adults, Experimental Aging Research: An International Journal Devoted to the Scientific Study of the Aging Process, 16:1, 3-9, DOI: 10.1080/03610739008253868 To link to this article: http://dx.doi.org/10.1080/03610739008253868
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
3
Experimental Aging Research, Volume 16, Number 1. 1990, ISSN 0734-0664 01990 Beech Hill Enterprises Inc.
Visual and Auditory Modality and Suffix Effects in Young and Elderly Adults SUSANKARP MANNINGAND JOANGREENHUT-WERTZ
Downloaded by [UQ Library] at 00:06 20 November 2014
Hunter College and the Graduate School of the City University of New York and New York University Medical Center Research has shown that elderly as compared with young adults show relative deficits both in processing visually as compared with auditorily presented stimuli and in tasks having attentional components. In this study, visual and auditory presentation was compared in young and elderly adults using the suffix paradigm in which the control condition involves immediate serial recall and the experimental condition, a suffix, a not-to-be-remembered final item. The standard finding in this paradigm is called the modality effect, superior auditory as compared with visual performance in the control condition which is localized at the end of the sequence. Generally, auditory suffixes following auditory sequences reduce the modality efect while visual suffixes following visual sequences do not. The results showed generally standard modality and suffix effects for both age groups. Relatively inferior performance was present in the elderly in the visual as compared with the auditory control conditions suggesting recoding deficits in this group. Auditory sufixes following auditory sequences had a relatively greater performance effect on the elderly than the young, while visual suffixes following visual sequences did not impair the performance of either group. This suggests a modality specific attentional deficit in the elderly. Rank order correlations suggest that individuals within both age groups showing large differences in performance between auditory and visual control conditions may have relative recoding difficulties for their age. Additionally, increased susceptibility to auditory interference for elderly as compared with young subjects may be a marker of aging, while relative within-group susceptibility to auditory interference may be a deficit on the part of young subjects.
M
any studies of young normal adult subjects have shown superior performance in immediate serial recall tasks when stimuli are presented auditorily rather than visually. (e.g., Cooley & McNulty, 1967; Crowder & Morton, 1969; Penney, 1975). A small group of studies has shown this modality superiority to be relatively greater in elderly as compared with young subjects. For example, McGhie, Chapman, and Lawson (1965) found relative deficits in the elderly when subjects were asked to report repeated digits in sequences when presentation was visual as compared with auditory. Le Breck & Baron (1987) show relative auditory superiority in a recognition memory paradigm while Arenberg (1968, 1977) and Taub (1975) demonstrate this modality difference with other tasks. Although the relative performance superiority when stimuli are presented auditorily as compared with visually has been a consistent finding in the small number of studies performed with elderly subjects, there has been
relatively little interpretation of its possible meaning. One possible explanation is that the elderly show a relative deficit in recoding stimuli. A considerable body of research has suggested that visual stimuli are processed in such a way that they are recoded (see Manning, Koehler, & Hampton [1990] for a discussion of different types of recoding) into a representation which contains auditory components, while auditory stimuli produce representations more similar to the physical signal. Support for this view comes from classic work such as that of Conrad (1964), which shows that normal adults frequently make auditory confusion errors with visually-presented stimuli while the reverse confusions may be relatively rare (see, however, Manning, 1977). A more recent view consistent with this explanation has been proposed by Shand and Klima (1981). They suggest that in normal subjects the auditory modality is a Primary Linguistic Modality which is defined as a modality or format in which subjects can process presented information
We wish to thank doctoral student Lisa Amoruso Mehr for help in running some of the subjects at Hunter College. Special thanks are also due to doctoral student Joan Mackell for her conscientious work at both Hunter and New York University Medical Center. We also wish to express our appreciation to Steven Ferris of the Department of Psychiatry for making the subjects and facilities at New York University Medical Center’s Milhauser Laboratory available to us. This paper was presented at the meeting of the Eastern Psychological Association, Arlington, Virginia, April, 1987. Some support for this project were provided by a grant from the City University of New York PSC-CUNY research award to Susan Karp Manning. Correspondence regarding this article and offprint requests should be sent to Susan Karp Manning, Department of Psychology, Hunter College, 695 Park Avenue, New York, NY 10021, U.S.A., or bitnet to SKMHC@CUNYVM.
MANNINGIGREENHUT-WERTZ
Downloaded by [UQ Library] at 00:06 20 November 2014
4
without transforming the representation into another format. The idea of auditory representations involving less recoding and being processed more directly than visual representations might be hypothesized to lead to the relative auditory superiority in the elderly. Anecdotal evidence for such a deficit in the elderly is provided by observations that a higher proportion of older as compared with younger adults seem to move their lips and/or read aloud. This effect is similar to that with children learning to read who first read aloud and then mouth words before learning to read silently. This suggests the recoding of visual stimuli may be less automatic for both elderly and children. With elderly these effects may be hypothesized to result from the breakdown of learned visual recoding with age, while with children they may be hypothesized to result from the early phases of learning to readily form appropriate visual representations. Another line of research with elderly populations is relevant to this research. Experimentation has shown that tasks such as digit span remain relatively unimpaired as compared with those involving greater attentiona1 demands such as response inhibition. For example, Freund and Witte (1986) found that older subjects perform less well in assessing the frequency of repeatedly presented items than younger subjects, and interpret these data as evidence for greater susceptibility to distraction from the non-repeated stimuli. However, in spite of the evidence for greater relative distractability in the elderly, whether this relative distractabilty is modality specific has never been tested. There is considerable data showing young normals to be more disrupted by even minimal auditory distractors in immediate auditory recall of auditory stimuli as compared with visual distractors in the recall of visual stimuli (e.g., Cooley & McNulty, 1967; Crowder & Morton, 1960). What should be predicted for the elderly? If the previously discussed data regarding the greater distractability of the elderly suggests a general problem with distractors, it would be predicted that the elderly would show relatively greater interference from any distractor than the young subjects. However, if recoding difficulties per se are the source of the relative visual deficit, a not-readily recoded visual stimulus could exert minimal effects while a more easily recoded auditory stimulus would lead to greater relative distractability in the elderly. Thus there would be little or no difference between relative performance for young and elderly subjects with visual distractors and a relatively greater difference in the effect of auditory distractors between the two groups. The distractability might then be strictly a function of the inability to ignore a stimulus requiring minimal recoding in order to be processed. Thus performance was compared for normal older and younger subjects using the suffix paradigm, which contains several components that test the above hypothesis and differentiate the two age groups. In the control condition, items are presented sequentially for immediately recall. When control sequences are presented auditorily
as compared with visually, recency, the superior recall of a final item as compared with middle items, tends to be present and/or of greater magnitude. This greater auditory as compared with visual recency is known as the modality effect. A suffix is a not-be-be-remembered item appended onto a sequence of items to-be-recalled, and thus is a minimal distractor. Sequences with suffixes are compared to controls to determine the effect of the suffix on recall. The suffix effect refers to the reduction or elimination of recency, when a suffix is present. In serial recall paradigms, suffix effects are found frequently with alphanumeric auditory stimuli and suffixes, but are virtually absent with comparable visual stimuli and suffixes (e.g., Crowder & Morton, 1969; Manning & Schreier, 1988).
For years, the dominant explanation for these effects was Crowder and Morton’s (1969) sensory trace theory which proposed longer lasting auditory than visual sensory traces resulting in extra auditory information that is useful in the recall of final auditory but not visual items. This was hypothesized to result in auditory but not visual recency and the modality effect. The auditory suffix effect was hypothesized to occur because the suffix prevented access to the information in the auditory sensory traces. Campbell and Dodd (1980), Spoehr and Corin (1978), Greene and Crowder (1984), and others report that stimuli that are lipread or mouthed, and thus not auditory, show both within-format recency and suffix effects, and show cross-format effects in reducing recency in auditorily presented stimuli. Glenberg (e.g., 1984) and others using the continuous distractor paradigm demonstrated auditory recency when recall was delayed and auditory interference was present both between presented items and between the final item and recall. Both the auditorylike effects and the results with delayed recall present a serious challenge to sensory trace views. Although theories have been proposed to explain the above exceptions to the usual auditory-visual distinctions (e.g., Campbell & Dodd, 1980; Greene & Crowder, 1984; Manning & Robininson, 1989; Nairne, 1988; Penney & Butt, 1986; Shand & Klima, 1981) or the results in paradigms involving delayed recall (e.g., Glenberg, 1984), none explain all the data. Nevertheless, in spite of exceptions, when comparing serial recall of standard auditory and visual alphanumeric stimuli, recency usually is greater for auditorily presented stimuli which leads to the modality effect. Large end-of-sequence suffix effects appear with auditory but not with visual stimuli. Further work with this paradigm has shown that a suffix may lead to a preterminal effect - one involving a decrement in performance on items preceding the final item. (See Baddeley & Hull, 1979; Balota & Engle, 198 1, for studies using auditory stimuli, and Manning, 1980; Manning & Gmuer, 1984, for studies using visual stimuli.) Engle and coworkers (e.g., Balota & Engle, 1981; Engle, 1980; Greenberg & Engle, 1983) hypothesize that, in the auditory modality, preterminal suffix effects may occur
5
Downloaded by [UQ Library] at 00:06 20 November 2014
MODALITY AND SUFFIX EFFECTS
in situations involving “strategic factors,” defined as strategies either used by the subject, implicit in the experimental task, or attentionally based. Thus the suffix paradigm should be ideal for clarifying the sources of differences in the processing of visual and auditory stimuli for young and elderly subjects. Specifically, the control condition requires immediate recall, while the suffix condition both requires response inhibition and increases attentional demands. What then is predicted about the results for the elderly as compared with the young in the standard suffix paradigm? The same pattern of end-of-sequencesuperiority for auditory as compared with visual sequences is expected to be present, resulting in recency and the modality effect for both groups. However, a general relative acrossthe-sequence visual inferiority is also expected - a result of the hypothesized recoding difficulties which lead to greater attentional demands in processing visually presented stimuli. As to the effects of suffixes, if the greater effect of interference on the elderly as compared with the young is not modality specific, greater suffix effects are predicted for the former group in both modalities at both terminal and preterminal parts of the sequence. If, however, the effects of interfering items on the elderly are modalityspecific, fewer relative effects of visual than auditory interference would be predicted. If the elderly are more likely to fully process auditory than visual stimuli, it is predicted that auditory suffixes will cause relatively greater performance decrements at both terminal and preterminal serial positions in this group as compared with the younger group. Method
Subjects Subjects in both the young and elderly groups were well-matched and above average with respect to both educational level and general functioning. All subjects had graduated from high school or the equivalent, and many had advanced degrees or were currently in college. All subjects were able to see and hear the stimuli with no difficulty. No subjects were using medication or had known medical problems that might impair cognitive functioning. The young were 13 female and 3 male students from Hunter College and New York University Medical School ( N W ) , ranging in age from 20-29 years with a mean age of 22.9 years. All were currently studying and some had advanced degrees. The elderly consisted of 14 female and 2 male subjects ranging in age from 60-84 years with a mean age of 68.6 years. The elderly subjects were either Hunter College students, part of a normal control group being studied at NYU, recruited from the NYU neighborhood, or referred by other subjects. All the elderly scored 27 or above on the Mini-Mental Status test (Folstein, Folstein, & McHugh, 1975) - well within the normal range. The ex-
periment was performed at both Hunter College and NYU.
Stimuli and Experimental Design The experiment consisted of four conditions: visual control, visual suffii, auditory control, and auditory suffix. Each condition required serial recall of eight sequences of the same six consonents, which, along with the suffix Y, were chosen to minimize interitem auditory and visual similarity. (See Manning, 1977, for ratings of auditory and visual similarity.) The letters J, Q, R, S, L, and M were arranged to form 32 different random sequences each containing every letter presented once in random order. The 32 sequences were divided into four lists of eight sequences. Subjects received all lists. Order assignment was such that the experiment formed four balanced Latin squares with respect to presentation order. Half of the subjects in each group received the two auditory (or visual) conditions first and the other two conditions second. Within each modality, half of the subjects received the suffix condition first and the control condition second.
Apparatus and Procedure Auditory stimuli were presented binaurally through headphones on tapes made by a female reader. The sound level was adjusted so as to be comfortable for each subject. Visual stimuli were shown at Hunter on a TRS-80 Model 3 computer and at NYU on a TRS Model 100 computer. No computer specific performance differences were apparent. Sequences were presented at a rate of one per second with a 30-second recall period during which subjects wrote the lists in strict serial order without backracking or making corrections. In the auditory control, the visual control, and the suffix conditions, respectively, a click, a dash, and the letter Y were described as not-to-be-recalled recall signals. Finally, to test a new technique, subjects on the last trial of each block of trials were asked to rate how sure they were of the correctness of each answer. Since these data will not be discussed further in this paper, they are only being mentioned to provide a complete procedure. Results
The data in Table 1 were scored by counting an item as correct if and only if it was both correct and in the correct serial position. Statistics reported had a probability of Type I error of .05 or less unless otherwise stated. The control conditions were analyzed alone by using an Age (2) x Modality (2; visual versus auditory) x Serial Position (6) analysis of variance (ANOVA). The main effects of serial position (F [5,150] = 14.34, MS. =1.17),age(F[1,30] = 5.38,MS. = 10.46),andmodality (F [1,30] = 4.94, MS,= 4.75) were significant. These main effects reflected serial position effects, superior
6
MANNING/GREENHUT- WERTZ
TABLE
1
Mean Number of Items Recalled in Control Condition as a Function of Age Group of Subjects, Presentation Modality, and Serial Position Serial Position
Elderly Auditory Visual Young Auditory Visual
Downloaded by [UQ Library] at 00:06 20 November 2014
Note:
1
2
3
4
5
6
7.fi9 7.(H)*
6.75 6.12*
6.44 5.50s
6.25 5.25*
6.44 5.31*
7.25 6.19*
7.~1 7.75
7.06 7.38
6.50 7.31
6.75 6.69
7.06 6.38*
1.75 6.94*
represents a significant Dunnett test, C - E = 0.58. Maximum score at each serial position is 8.00.
recall of the young as compared to the elderly, and superior performance on the auditory as compared to the visual control condition. The Modality x Serial Position (F[5,150] = 2.98, MS, = .96) and the Modality x Age (F[ 1,301 = 3.42, MS. = .96, p < .lo) interactions were the only other reliable effects, with the latter being marginally significant, They were due, respectively, to the standard modality effect being present for both age groups, and the relatively greater difference between auditory and visual performance for the elderly across the early parts of the serial recall curve. Preplanned Dunnett tests were performed, using the Age x Modality x Serial Position interaction as the error term; C - E = .58. These comparisons show significant superiority for auditory presentation at all serial positions within the elderly group. These effects increase in size at the later serial positions. For the young, serial position effects were limited to the final two positions. Thus the elderly show a generalized relative performance decrement for visual stimuli in addition to a standard modality effect. Young subjects exhibit only a standard modality effect. To obtain more information about the nature of the modality differences in the young and elderly, a further series of preplanned comparisons were made. In order to test for the relative visual inferiority found by previous researchers, matched t tests were performed comparing overall auditory to overal auditory visual performance within both the young imd the elderly groups. The results showed significant overall performance superiority to be present for the elderly but not for the young subjects; ts (15) = 2.48 and .75, respectively. Additionally, a score, to be called the discrepancy score (DS), was calculated for all subjects by subtracting overall performance in the visual condition from overall performance in the auditory condition. As expected there was a significantly greater difference between DSs for the elderly than the young, t (30) = 1.84. In an effort to find out more about modality differences in the young and elderly, correlations were done within age groups between the auditory and visual con-
trol performance and the DS, respectively. To explain the procedure used, the Pearson product moment correlation (r) between auditory (a) and visual (v) control performance can be referred to as row,the correlation between a and DS (which is equal to M aminus M.) can be referred to as rat.-”),and the correlation between v and DS can be referred to as T ~ ( . - ~ )Because . using Pearson r1to correlate a and/or v with DS leads to values which are both a function of r., and of 0. and u., rank order correlations, r,, were used. These r, provide information about the relative loading of the auditory and visual components on DS, independent of 0. and 0”. The results showed r,o(o-v) and r,,.-,, to be + .26 and - .73, and + .36 and - .63 for the elderly and young, respectively. Only r-(.-”)was significant for both groups. These patterns are similar and indicate the DS in both groups to have a heavier loading on visual as contrasted with auditory performance.2 To further measure age differences, an independent I test comparing auditory and visual performance at the terminal position showed no significant age differences to be present in the size of the modality effect; t (30) = .40.
Thus both elderly and young subjects show patterns of recency and associated modality effects which do not differ in any major way at the final position. When compared with the young, however, the elderly show both overall poorer performance and the predicted relatively greater differential between auditory and visual recall at the preterminal positions. A second ANOVA was performed to determine the effect of the suffix. The results of this Age (2) x Modality (2) X Suffix (2) x Serial Position (6) analysis are shown ‘The actual formula for r .,.-” is o. - r.”o, / \Jd+ o; - 2r.v0.0, The formula for rv(.-”)may be derived similarly.
,
2Such a pattern of correlations would not have to exist. For example, if the DSs were approximately the same size for all subjects, it would be possible for overall auditory superiority to exist in the absence of negative correlations between DS and visual performance. Other possible patterns also exist.
7
MODALITY AND SUFFIX EFFECTS
TABLE 2 Mean Number of Items Recalled in Suffix and Control Conditions as a Function of Age Group of Subjects, Presentation Modality, and Serial Position
Downloaded by [UQ Library] at 00:06 20 November 2014
Serial Position
Elderly Auditory Control Suffix Visual Control Suffix Young Auditory Control Suffix Visual Control Suffix
1
2
3
4
5
6
7.69 7.19
6.75 5.69*
6.44 4.88*
6.25 4.62*
6.44 4.62*
7.25 5.75*
7.00 7.38
6.12 6.06
5.50 5.19
5.25 4,88
5.31 4.88
6.19 5.44*
7.81 7.81
7.06 7.12
6.50 6.50
6.75 5.56*
7.06 5.69*
7.75 6.38*
7.75 7.81
7.38 7.06
7.31 7.19
6.69 6.69
6.38 6.56
6.94 7.00
Note: * represents a significant Dunnett test, C - E = 0.58. Maximum score at each serial position is 8.00.
in Table 2. Significant main effects of age (8' [1,30] = 7.94, MS. = 24.19), suffix (F [1,30] = 17.47, MS, = 3.54), and serial position (F [5,150] = 29.60, MS, = 1.79) were present. These were due, respectively, to the superior performance for the young as compared with the elderly, the superior performance on the no-suffix as compared with the suffix condition, and the usual serial position effects. Remaining significant effects included the interactions of Modality x Suffix (F [1,30] = 13.83, MS, = 2.53), and Serial Position x Suffix (F[5,150] = 5.06, MS. = 0.85). These were the result of the greater effects of the suffix on the auditory modality, and of the suffix at the end of the serial position curve. Preplanned Dunnett tests were performed using the Age x Modality x Suffix x Serial Position interaction as the error term; C = E = 0.58. This analysis showed significant suffix decrements for the auditory modality at Positions 4-6 for the young and 2-6 for the elderly. For the visual modality, the Dunnett tests showed a small significant decrement to be present for elderly subjects at Position 6. Since end-of-sequence suffix effects are considered to be different from preterminal effects, they were analyzed separately. Independent t tests showed no significant differences in the size of terminal suffix effects for elderly and young subjects, for either auditory or visual stimuli; ts (30) = -21 and 1.17, respectively. As for preterminal suffix effects, matched t tests comparing the overall size of the suffix decrement over all positions (in order to bias the test against our hypothesis)
showed reliably greater decrements for auditory as compared with visual stimuli for both elderly and young groups; ts (15) = 2.92 and 2.74, respectively. In a manner similar to the way the differential performance between the auditory and visual modality (the DS) was analyzed, the size of the decrement produced by an auditory suffix - to be called the auditory interference score (AIS) - was examined. AIS was calculated for each subject by subtracting the score on the auditory suffix condition from that on the auditory control condition. As with DS, rank order correlations were performed between the auditory control condition and AIS for the age groups separately. The pattern of results differed such that the young showed a significant negative correlation (rr.(nir)= - 32) and the elderly showed a small nonsignificant positive correlation (rsa(ois) = + . l 1). This is strongly suggestive that the overall greater susceptibility to auditory interference on the part of the elderly is age specific and not related to immediate memory of auditory stimuli in the absence of interference. In the young, however, the group shows less overall interference; it appears that individuals showing relatively greater interference may also show relatively poorer overall auditory performance in the control condition. Further tests showed the overall decrement produced by the auditory suffix on auditory stimuli to be significantly greater for the elderly than for the young; t (30) = 2.93. The overall dcecrement produced by the visual suffix on visual stimuli was not different for young and elderly; I(30) = 3 6 .
8
MANNINGIGREENHUT-WERTZ
Downloaded by [UQ Library] at 00:06 20 November 2014
Discussion Several conclusions may be drawn from this experiment, which we believe to be the first in which the suffix paradigm was used to compare visual and auditory stimulus presentation with young and elderly subjects. First, performance on certain aspects of the paradigm differs for the two age groups. As predicted, in the control condition, the elderly show superior auditory over visual recall. It was hypothesized that the source of these difference might be greater difficulty in stimulus recoding for elderly subjects. If visual stimuli need to be recoded into an auditory representation, while auditory stimuli can be processed more directly from the physical signal, this recoding effect would contribute to the relative modality differences between age groups. To explain the suffix data, the effects of auditory and visual suffixes will be considered first. Although both groups show both terminal and nonterminal auditory suffix effects, the elderly show preterminal auditory suffix effects that are significantly greater than those shown by the young. These effects extend further back in the sequence for the elderly. Although a very small atypical visual suffix effect is present for the elderly at the terminal position, a finding not unknown in the young (see, for example, Hitch [1975]), neither group shows visual suffix effects of any size. These modality differences are of theoretical importance and indicate the 'deficits of the elderly are not attentional in a general sense. Instead they are modality specific. This then contradicts the hypothesis that the requirement of exclusion of a distractor always results in relatively worse performance in the elderly than in the young. What then might explain the pattern of results that were obtained here? As before, we hypothesize that recoding difficulties in the elderly in combination with an inability to exclude irrelevant stimuli when they are presented in a mode that is easy to process can explain the entire pattern of results in this experiment. This explanation is offered for both the relative auditory superiority and the greater susceptibility to preterminal interference from an auditory suffix found in the elderly. As for other information obtained in this study, the large negative correlations of the DS with visual performance combined with thle small positive correlations with auditory performance suggest that a large DS may relate to generalized coding difficulties. Subjects in both groups with relatively large DSs show relatively poor recall of visual stimuli for their age. For the elderly, this may be a new situation related to aging or it may be an exaggerated manifestation of a lifelong difficulty. For the young, it may represent a lifelong deficit. Regardless, however, of the source, because of the differential correlations of DS with the visual and auditory control conditions, DS is worthy of further study as a possible marker of coding deficits. Unlike DS, AIS shows differential patterns of results for young and elderly subjects. It is virtually uncorrelated
with performance in the auditory control condition in the elderly, but is negatively correlated in the young. This lack of correlation, combined with the overall larger size of AIS, suggests that susceptibility to auditory interference is a more common deficit in the elderly. In the young, however, the high negative correlation between AIS and the auditory control is suggestive of a deficit that is also of interest to investigate. Before concluding, there are three further issues to mention. First, although there is a general decrement and some previously described relative differences in the performance of elderly as compared with young subjects, most standard effects in this paradigm are present and d o not interact with age. This is consistent with previous findings that the modality effect and the auditory endof-sequence suffix effect are extremely robust over many conditions. This experiment then adds the normal elderly - a group which may have attentional and possibly other deficits - to those exhibiting standard findings. Second, this experiment used short sequence lengths which were easy for most of the subjects in both the young and elderly groups. It is of importance to see whether the same findings with respect to similarities and differences in modality and suffix effects between the two groups hold with longer sequences. Regardless, however, of any differences in the data patterns which may appear in other experimental situations, these data add to the literature in this area. Finally, there is an important set of data to be accounted for. Parkinson and Perey (1980) performed a study using auditory stimuli in the suffix paradigm which compared young and elderly subjects. These researchers report data showing that both the shape of the serial position curve and the type of suffix effect obtained depend on the interaction between sequence length and digit span for individual subjects. Of greatest importance for the current study were findings that, when digit spans were not equated, performance was poorer for elderly subjects. However, when span was equated by use of subjects with five and six item spans, the size of the auditory suffix was the same for both age groups. Parkinson and Perey conclude that it is a difference in immediate memory span that leads to the differential performance between young and elderly when digit span is not equated. We question these conclusions. Control performance in the suffix paradigm is similar to a span condition and is certainly a measure of immediate memory. The only way that our results could have been obtained under the Parkinson and Perey hypothesis is if the young subjects were showing large variability in their memory span while the elderly were not. Examination of our auditory control condition indicates this is definitely not the case. Both the mean performance and the standard deviations are very similar for the elderly and young, respectively, Ms = 6.80 and 7.16, SDs = .92 and .71. A second problem with the Parkinson and Perey view is that they used young subjects with unusually short auditory spans. Given our negative correlation between
MODALITY AND SUFFIX EFFECTS
AIS and the control condition in the young and the lack of such an effect in the elderly, it may be that Parkinson and Perey chose young subjects who were unusually greatly affected by auditory interference and compared them to elderly subjects where AIS was not correlated with immediate memory. Thus we conclude that the findings here open up a variety of areas of research with respect to the effects of aging on immediate recall. They also suggest some areas of work of possible importance in detecting recoding deficits in young and older subjects.
Downloaded by [UQ Library] at 00:06 20 November 2014
References Arenberg, D. (1968). Input modality in short-term retention of old and young adults. Journal of Gerontology, 23, 462-465. Arenberg, D. (1977). The effects of auditory augmentation on visual retention for young and old adults. Journal of Gerontology, 32, 192-195. Baddeley, A., & Hull, A. (1979). Prefix and suffix effects: Do they have a common basis? Journal of Verbal Learning and Verbel Behavior, 18, 129-140. Balota, D.A., & Engle, R.E. (1981). Structural and strategic factors in the stimulus suffix effect. Journal of Verbal Learning and Verbal Behavior, 20, 346-357. Campbell, R., & Dodd, B. (1980). Hearing by eye. Quarterly Journal of Experimental Psychology, 32, 85-99. Conrad, R. (1964). Acoustic confusions in immediate memory. British Journal of Psychology, 55, 75-84. Cooley, R.K., & McNulty, J.A. (1967). Recall of CCC trigrams over short intervals as a function of mode of presentation. Psychonomic Science, 9, 543-544. Crowder, R.G., & Morton, J. (1969). Precategorical acoustic storage (PAS). Perception and Psychophysics, 5, 365-373. Engle, R.W. (1980). The suffix effect: How many positions are involved? Memory and Cognition, 8, 247-252. Folstein, M.F., Folstein, S.E., & McHugh, P.R. (1975). Mini Mental State: A practical method for grading the cognitive state for the clinician. Journal of Psychiatric Research, 12, 189- 198. Freund, J.S., & Witte, K.L. (1986). Recognition and frequency judgments in young and elderly adults. American Journal of Psychology, 99, 91-102. Glenberg, A.M. (1984). A retrieval account of the longterm modality effect. Journal of Experimental Psychology:Learning, Memory, and Cognition, 10, 16-31. Greenberg, S.N., & Engle, R.W. (1983). Voice change in the stimulus suffix effect: Are the effects structural or strategic? Memory and Cognition, 11, 551-556. Greene, R.L., & Crowder, R.G. (1984). Modality and suffix effects in the absence of auditory stimulation.
9
Journal of Verbal Learning and VerbalBehavior, 23, 371-382.
Hitch G.J. (1975). The role of attention in visual and auditory suffix effects. Memory and Cognition, 3, 501-595.
Le Breck, D.B., & Baron, A. (1987). Age and practice effects in continuous recognition memory. Journal of Gerontology, 42, 89-91. McGhie, A., Chapman, J., & Lawson, J.S. (1965). Changes in immediate memory with age. British Journal of Psychology, 56, 69-75. Manning, S.K. (1977). Ratings of the auditory and visual similarity of consonants: Implications for research. Behavior Research Methods and Instrumentation, 9, 495-498.
Manning, S.K. (1980). Tactual and visual suffix effects. Quarterlj Journal of Experimental Psychology, 32, 257-267.
Manning, S.K. (1985). Suffix effects on visual sequences presented with the Optacon. A test of sensory, attentional, and changing state theories. Bulletin of the Psychonomic Society, 23, 1-4. Manning, S.K., Koehler, L., & Hampton, S. (1990). The effects of recoding and presentation format in recency suffix effects. Memory and Cognition, 18, 164-173. Manning, S.K., & Robinson, 1.1. (1989). Recency and suffix effects as a function of auditory confusability and set size. American Journal of Psychology, 102, 495-5 10.
Manning, S.K., & Schreier, H. (1988). Recency and suffix effects in pictures as a function of recall method. American Journal of Psychology, 101, 97-109. Nairne, J. (1988). A framework for interpreting recency effects in immediate serial recall. Memory and Cognition, 16, 385-393. Parkinson, S.R., & Perey, A. (1980). Aging, digit span, and the stimulus suffix effect. Journal of Gerontology, 35, 736-742. Penney, C.G. (1975). Modality effects in short-term verbal memory. Psychological Bulletin, 82, 68-84. Penney, C.G., &Butt, A.K. (1986). Within- and betweenmodality associations in probed recall: A test of the separate-streams hypothesis. Canadian Journal of Psychology, 40, 1- 11. Shand, M.A., & Klima, E.S. (1981). Non auditory suffix effects in congenitally deaf signers of American Sign Language. Journal of Experimental Psychology: Human Learning and Memory, 7, 464-474. Spoehr, K.T., & Corin, W.J. (1978). The stimulus suffix effect as a memory coding phenomenon. Memory and Cognition, 6, 583-589. Taub, H.A. (1975). Mode of presentation, age, and shortterm memory. Journal of Gerontology, 50, 56-59.
Experimental Aging Research: An International Journal Devoted to the Scientific Study of the Aging Process Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uear20
Visual and auditory modality and suffix effects in young and elderly adults a
Susan Karp Manning & Joan Greenhut-Wertz
a
a
Hunter College and the Graduate School of the City University of New York and New York University Medical Center Published online: 27 Sep 2007.
To cite this article: Susan Karp Manning & Joan Greenhut-Wertz (1990) Visual and auditory modality and suffix effects in young and elderly adults, Experimental Aging Research: An International Journal Devoted to the Scientific Study of the Aging Process, 16:1, 3-9, DOI: 10.1080/03610739008253868 To link to this article: http://dx.doi.org/10.1080/03610739008253868
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
3
Experimental Aging Research, Volume 16, Number 1. 1990, ISSN 0734-0664 01990 Beech Hill Enterprises Inc.
Visual and Auditory Modality and Suffix Effects in Young and Elderly Adults SUSANKARP MANNINGAND JOANGREENHUT-WERTZ
Downloaded by [UQ Library] at 00:06 20 November 2014
Hunter College and the Graduate School of the City University of New York and New York University Medical Center Research has shown that elderly as compared with young adults show relative deficits both in processing visually as compared with auditorily presented stimuli and in tasks having attentional components. In this study, visual and auditory presentation was compared in young and elderly adults using the suffix paradigm in which the control condition involves immediate serial recall and the experimental condition, a suffix, a not-to-be-remembered final item. The standard finding in this paradigm is called the modality effect, superior auditory as compared with visual performance in the control condition which is localized at the end of the sequence. Generally, auditory suffixes following auditory sequences reduce the modality efect while visual suffixes following visual sequences do not. The results showed generally standard modality and suffix effects for both age groups. Relatively inferior performance was present in the elderly in the visual as compared with the auditory control conditions suggesting recoding deficits in this group. Auditory sufixes following auditory sequences had a relatively greater performance effect on the elderly than the young, while visual suffixes following visual sequences did not impair the performance of either group. This suggests a modality specific attentional deficit in the elderly. Rank order correlations suggest that individuals within both age groups showing large differences in performance between auditory and visual control conditions may have relative recoding difficulties for their age. Additionally, increased susceptibility to auditory interference for elderly as compared with young subjects may be a marker of aging, while relative within-group susceptibility to auditory interference may be a deficit on the part of young subjects.
M
any studies of young normal adult subjects have shown superior performance in immediate serial recall tasks when stimuli are presented auditorily rather than visually. (e.g., Cooley & McNulty, 1967; Crowder & Morton, 1969; Penney, 1975). A small group of studies has shown this modality superiority to be relatively greater in elderly as compared with young subjects. For example, McGhie, Chapman, and Lawson (1965) found relative deficits in the elderly when subjects were asked to report repeated digits in sequences when presentation was visual as compared with auditory. Le Breck & Baron (1987) show relative auditory superiority in a recognition memory paradigm while Arenberg (1968, 1977) and Taub (1975) demonstrate this modality difference with other tasks. Although the relative performance superiority when stimuli are presented auditorily as compared with visually has been a consistent finding in the small number of studies performed with elderly subjects, there has been
relatively little interpretation of its possible meaning. One possible explanation is that the elderly show a relative deficit in recoding stimuli. A considerable body of research has suggested that visual stimuli are processed in such a way that they are recoded (see Manning, Koehler, & Hampton [1990] for a discussion of different types of recoding) into a representation which contains auditory components, while auditory stimuli produce representations more similar to the physical signal. Support for this view comes from classic work such as that of Conrad (1964), which shows that normal adults frequently make auditory confusion errors with visually-presented stimuli while the reverse confusions may be relatively rare (see, however, Manning, 1977). A more recent view consistent with this explanation has been proposed by Shand and Klima (1981). They suggest that in normal subjects the auditory modality is a Primary Linguistic Modality which is defined as a modality or format in which subjects can process presented information
We wish to thank doctoral student Lisa Amoruso Mehr for help in running some of the subjects at Hunter College. Special thanks are also due to doctoral student Joan Mackell for her conscientious work at both Hunter and New York University Medical Center. We also wish to express our appreciation to Steven Ferris of the Department of Psychiatry for making the subjects and facilities at New York University Medical Center’s Milhauser Laboratory available to us. This paper was presented at the meeting of the Eastern Psychological Association, Arlington, Virginia, April, 1987. Some support for this project were provided by a grant from the City University of New York PSC-CUNY research award to Susan Karp Manning. Correspondence regarding this article and offprint requests should be sent to Susan Karp Manning, Department of Psychology, Hunter College, 695 Park Avenue, New York, NY 10021, U.S.A., or bitnet to SKMHC@CUNYVM.
MANNINGIGREENHUT-WERTZ
Downloaded by [UQ Library] at 00:06 20 November 2014
4
without transforming the representation into another format. The idea of auditory representations involving less recoding and being processed more directly than visual representations might be hypothesized to lead to the relative auditory superiority in the elderly. Anecdotal evidence for such a deficit in the elderly is provided by observations that a higher proportion of older as compared with younger adults seem to move their lips and/or read aloud. This effect is similar to that with children learning to read who first read aloud and then mouth words before learning to read silently. This suggests the recoding of visual stimuli may be less automatic for both elderly and children. With elderly these effects may be hypothesized to result from the breakdown of learned visual recoding with age, while with children they may be hypothesized to result from the early phases of learning to readily form appropriate visual representations. Another line of research with elderly populations is relevant to this research. Experimentation has shown that tasks such as digit span remain relatively unimpaired as compared with those involving greater attentiona1 demands such as response inhibition. For example, Freund and Witte (1986) found that older subjects perform less well in assessing the frequency of repeatedly presented items than younger subjects, and interpret these data as evidence for greater susceptibility to distraction from the non-repeated stimuli. However, in spite of the evidence for greater relative distractability in the elderly, whether this relative distractabilty is modality specific has never been tested. There is considerable data showing young normals to be more disrupted by even minimal auditory distractors in immediate auditory recall of auditory stimuli as compared with visual distractors in the recall of visual stimuli (e.g., Cooley & McNulty, 1967; Crowder & Morton, 1960). What should be predicted for the elderly? If the previously discussed data regarding the greater distractability of the elderly suggests a general problem with distractors, it would be predicted that the elderly would show relatively greater interference from any distractor than the young subjects. However, if recoding difficulties per se are the source of the relative visual deficit, a not-readily recoded visual stimulus could exert minimal effects while a more easily recoded auditory stimulus would lead to greater relative distractability in the elderly. Thus there would be little or no difference between relative performance for young and elderly subjects with visual distractors and a relatively greater difference in the effect of auditory distractors between the two groups. The distractability might then be strictly a function of the inability to ignore a stimulus requiring minimal recoding in order to be processed. Thus performance was compared for normal older and younger subjects using the suffix paradigm, which contains several components that test the above hypothesis and differentiate the two age groups. In the control condition, items are presented sequentially for immediately recall. When control sequences are presented auditorily
as compared with visually, recency, the superior recall of a final item as compared with middle items, tends to be present and/or of greater magnitude. This greater auditory as compared with visual recency is known as the modality effect. A suffix is a not-be-be-remembered item appended onto a sequence of items to-be-recalled, and thus is a minimal distractor. Sequences with suffixes are compared to controls to determine the effect of the suffix on recall. The suffix effect refers to the reduction or elimination of recency, when a suffix is present. In serial recall paradigms, suffix effects are found frequently with alphanumeric auditory stimuli and suffixes, but are virtually absent with comparable visual stimuli and suffixes (e.g., Crowder & Morton, 1969; Manning & Schreier, 1988).
For years, the dominant explanation for these effects was Crowder and Morton’s (1969) sensory trace theory which proposed longer lasting auditory than visual sensory traces resulting in extra auditory information that is useful in the recall of final auditory but not visual items. This was hypothesized to result in auditory but not visual recency and the modality effect. The auditory suffix effect was hypothesized to occur because the suffix prevented access to the information in the auditory sensory traces. Campbell and Dodd (1980), Spoehr and Corin (1978), Greene and Crowder (1984), and others report that stimuli that are lipread or mouthed, and thus not auditory, show both within-format recency and suffix effects, and show cross-format effects in reducing recency in auditorily presented stimuli. Glenberg (e.g., 1984) and others using the continuous distractor paradigm demonstrated auditory recency when recall was delayed and auditory interference was present both between presented items and between the final item and recall. Both the auditorylike effects and the results with delayed recall present a serious challenge to sensory trace views. Although theories have been proposed to explain the above exceptions to the usual auditory-visual distinctions (e.g., Campbell & Dodd, 1980; Greene & Crowder, 1984; Manning & Robininson, 1989; Nairne, 1988; Penney & Butt, 1986; Shand & Klima, 1981) or the results in paradigms involving delayed recall (e.g., Glenberg, 1984), none explain all the data. Nevertheless, in spite of exceptions, when comparing serial recall of standard auditory and visual alphanumeric stimuli, recency usually is greater for auditorily presented stimuli which leads to the modality effect. Large end-of-sequence suffix effects appear with auditory but not with visual stimuli. Further work with this paradigm has shown that a suffix may lead to a preterminal effect - one involving a decrement in performance on items preceding the final item. (See Baddeley & Hull, 1979; Balota & Engle, 198 1, for studies using auditory stimuli, and Manning, 1980; Manning & Gmuer, 1984, for studies using visual stimuli.) Engle and coworkers (e.g., Balota & Engle, 1981; Engle, 1980; Greenberg & Engle, 1983) hypothesize that, in the auditory modality, preterminal suffix effects may occur
5
Downloaded by [UQ Library] at 00:06 20 November 2014
MODALITY AND SUFFIX EFFECTS
in situations involving “strategic factors,” defined as strategies either used by the subject, implicit in the experimental task, or attentionally based. Thus the suffix paradigm should be ideal for clarifying the sources of differences in the processing of visual and auditory stimuli for young and elderly subjects. Specifically, the control condition requires immediate recall, while the suffix condition both requires response inhibition and increases attentional demands. What then is predicted about the results for the elderly as compared with the young in the standard suffix paradigm? The same pattern of end-of-sequencesuperiority for auditory as compared with visual sequences is expected to be present, resulting in recency and the modality effect for both groups. However, a general relative acrossthe-sequence visual inferiority is also expected - a result of the hypothesized recoding difficulties which lead to greater attentional demands in processing visually presented stimuli. As to the effects of suffixes, if the greater effect of interference on the elderly as compared with the young is not modality specific, greater suffix effects are predicted for the former group in both modalities at both terminal and preterminal parts of the sequence. If, however, the effects of interfering items on the elderly are modalityspecific, fewer relative effects of visual than auditory interference would be predicted. If the elderly are more likely to fully process auditory than visual stimuli, it is predicted that auditory suffixes will cause relatively greater performance decrements at both terminal and preterminal serial positions in this group as compared with the younger group. Method
Subjects Subjects in both the young and elderly groups were well-matched and above average with respect to both educational level and general functioning. All subjects had graduated from high school or the equivalent, and many had advanced degrees or were currently in college. All subjects were able to see and hear the stimuli with no difficulty. No subjects were using medication or had known medical problems that might impair cognitive functioning. The young were 13 female and 3 male students from Hunter College and New York University Medical School ( N W ) , ranging in age from 20-29 years with a mean age of 22.9 years. All were currently studying and some had advanced degrees. The elderly consisted of 14 female and 2 male subjects ranging in age from 60-84 years with a mean age of 68.6 years. The elderly subjects were either Hunter College students, part of a normal control group being studied at NYU, recruited from the NYU neighborhood, or referred by other subjects. All the elderly scored 27 or above on the Mini-Mental Status test (Folstein, Folstein, & McHugh, 1975) - well within the normal range. The ex-
periment was performed at both Hunter College and NYU.
Stimuli and Experimental Design The experiment consisted of four conditions: visual control, visual suffii, auditory control, and auditory suffix. Each condition required serial recall of eight sequences of the same six consonents, which, along with the suffix Y, were chosen to minimize interitem auditory and visual similarity. (See Manning, 1977, for ratings of auditory and visual similarity.) The letters J, Q, R, S, L, and M were arranged to form 32 different random sequences each containing every letter presented once in random order. The 32 sequences were divided into four lists of eight sequences. Subjects received all lists. Order assignment was such that the experiment formed four balanced Latin squares with respect to presentation order. Half of the subjects in each group received the two auditory (or visual) conditions first and the other two conditions second. Within each modality, half of the subjects received the suffix condition first and the control condition second.
Apparatus and Procedure Auditory stimuli were presented binaurally through headphones on tapes made by a female reader. The sound level was adjusted so as to be comfortable for each subject. Visual stimuli were shown at Hunter on a TRS-80 Model 3 computer and at NYU on a TRS Model 100 computer. No computer specific performance differences were apparent. Sequences were presented at a rate of one per second with a 30-second recall period during which subjects wrote the lists in strict serial order without backracking or making corrections. In the auditory control, the visual control, and the suffix conditions, respectively, a click, a dash, and the letter Y were described as not-to-be-recalled recall signals. Finally, to test a new technique, subjects on the last trial of each block of trials were asked to rate how sure they were of the correctness of each answer. Since these data will not be discussed further in this paper, they are only being mentioned to provide a complete procedure. Results
The data in Table 1 were scored by counting an item as correct if and only if it was both correct and in the correct serial position. Statistics reported had a probability of Type I error of .05 or less unless otherwise stated. The control conditions were analyzed alone by using an Age (2) x Modality (2; visual versus auditory) x Serial Position (6) analysis of variance (ANOVA). The main effects of serial position (F [5,150] = 14.34, MS. =1.17),age(F[1,30] = 5.38,MS. = 10.46),andmodality (F [1,30] = 4.94, MS,= 4.75) were significant. These main effects reflected serial position effects, superior
6
MANNING/GREENHUT- WERTZ
TABLE
1
Mean Number of Items Recalled in Control Condition as a Function of Age Group of Subjects, Presentation Modality, and Serial Position Serial Position
Elderly Auditory Visual Young Auditory Visual
Downloaded by [UQ Library] at 00:06 20 November 2014
Note:
1
2
3
4
5
6
7.fi9 7.(H)*
6.75 6.12*
6.44 5.50s
6.25 5.25*
6.44 5.31*
7.25 6.19*
7.~1 7.75
7.06 7.38
6.50 7.31
6.75 6.69
7.06 6.38*
1.75 6.94*
represents a significant Dunnett test, C - E = 0.58. Maximum score at each serial position is 8.00.
recall of the young as compared to the elderly, and superior performance on the auditory as compared to the visual control condition. The Modality x Serial Position (F[5,150] = 2.98, MS, = .96) and the Modality x Age (F[ 1,301 = 3.42, MS. = .96, p < .lo) interactions were the only other reliable effects, with the latter being marginally significant, They were due, respectively, to the standard modality effect being present for both age groups, and the relatively greater difference between auditory and visual performance for the elderly across the early parts of the serial recall curve. Preplanned Dunnett tests were performed, using the Age x Modality x Serial Position interaction as the error term; C - E = .58. These comparisons show significant superiority for auditory presentation at all serial positions within the elderly group. These effects increase in size at the later serial positions. For the young, serial position effects were limited to the final two positions. Thus the elderly show a generalized relative performance decrement for visual stimuli in addition to a standard modality effect. Young subjects exhibit only a standard modality effect. To obtain more information about the nature of the modality differences in the young and elderly, a further series of preplanned comparisons were made. In order to test for the relative visual inferiority found by previous researchers, matched t tests were performed comparing overall auditory to overal auditory visual performance within both the young imd the elderly groups. The results showed significant overall performance superiority to be present for the elderly but not for the young subjects; ts (15) = 2.48 and .75, respectively. Additionally, a score, to be called the discrepancy score (DS), was calculated for all subjects by subtracting overall performance in the visual condition from overall performance in the auditory condition. As expected there was a significantly greater difference between DSs for the elderly than the young, t (30) = 1.84. In an effort to find out more about modality differences in the young and elderly, correlations were done within age groups between the auditory and visual con-
trol performance and the DS, respectively. To explain the procedure used, the Pearson product moment correlation (r) between auditory (a) and visual (v) control performance can be referred to as row,the correlation between a and DS (which is equal to M aminus M.) can be referred to as rat.-”),and the correlation between v and DS can be referred to as T ~ ( . - ~ )Because . using Pearson r1to correlate a and/or v with DS leads to values which are both a function of r., and of 0. and u., rank order correlations, r,, were used. These r, provide information about the relative loading of the auditory and visual components on DS, independent of 0. and 0”. The results showed r,o(o-v) and r,,.-,, to be + .26 and - .73, and + .36 and - .63 for the elderly and young, respectively. Only r-(.-”)was significant for both groups. These patterns are similar and indicate the DS in both groups to have a heavier loading on visual as contrasted with auditory performance.2 To further measure age differences, an independent I test comparing auditory and visual performance at the terminal position showed no significant age differences to be present in the size of the modality effect; t (30) = .40.
Thus both elderly and young subjects show patterns of recency and associated modality effects which do not differ in any major way at the final position. When compared with the young, however, the elderly show both overall poorer performance and the predicted relatively greater differential between auditory and visual recall at the preterminal positions. A second ANOVA was performed to determine the effect of the suffix. The results of this Age (2) x Modality (2) X Suffix (2) x Serial Position (6) analysis are shown ‘The actual formula for r .,.-” is o. - r.”o, / \Jd+ o; - 2r.v0.0, The formula for rv(.-”)may be derived similarly.
,
2Such a pattern of correlations would not have to exist. For example, if the DSs were approximately the same size for all subjects, it would be possible for overall auditory superiority to exist in the absence of negative correlations between DS and visual performance. Other possible patterns also exist.
7
MODALITY AND SUFFIX EFFECTS
TABLE 2 Mean Number of Items Recalled in Suffix and Control Conditions as a Function of Age Group of Subjects, Presentation Modality, and Serial Position
Downloaded by [UQ Library] at 00:06 20 November 2014
Serial Position
Elderly Auditory Control Suffix Visual Control Suffix Young Auditory Control Suffix Visual Control Suffix
1
2
3
4
5
6
7.69 7.19
6.75 5.69*
6.44 4.88*
6.25 4.62*
6.44 4.62*
7.25 5.75*
7.00 7.38
6.12 6.06
5.50 5.19
5.25 4,88
5.31 4.88
6.19 5.44*
7.81 7.81
7.06 7.12
6.50 6.50
6.75 5.56*
7.06 5.69*
7.75 6.38*
7.75 7.81
7.38 7.06
7.31 7.19
6.69 6.69
6.38 6.56
6.94 7.00
Note: * represents a significant Dunnett test, C - E = 0.58. Maximum score at each serial position is 8.00.
in Table 2. Significant main effects of age (8' [1,30] = 7.94, MS. = 24.19), suffix (F [1,30] = 17.47, MS, = 3.54), and serial position (F [5,150] = 29.60, MS, = 1.79) were present. These were due, respectively, to the superior performance for the young as compared with the elderly, the superior performance on the no-suffix as compared with the suffix condition, and the usual serial position effects. Remaining significant effects included the interactions of Modality x Suffix (F [1,30] = 13.83, MS, = 2.53), and Serial Position x Suffix (F[5,150] = 5.06, MS. = 0.85). These were the result of the greater effects of the suffix on the auditory modality, and of the suffix at the end of the serial position curve. Preplanned Dunnett tests were performed using the Age x Modality x Suffix x Serial Position interaction as the error term; C = E = 0.58. This analysis showed significant suffix decrements for the auditory modality at Positions 4-6 for the young and 2-6 for the elderly. For the visual modality, the Dunnett tests showed a small significant decrement to be present for elderly subjects at Position 6. Since end-of-sequence suffix effects are considered to be different from preterminal effects, they were analyzed separately. Independent t tests showed no significant differences in the size of terminal suffix effects for elderly and young subjects, for either auditory or visual stimuli; ts (30) = -21 and 1.17, respectively. As for preterminal suffix effects, matched t tests comparing the overall size of the suffix decrement over all positions (in order to bias the test against our hypothesis)
showed reliably greater decrements for auditory as compared with visual stimuli for both elderly and young groups; ts (15) = 2.92 and 2.74, respectively. In a manner similar to the way the differential performance between the auditory and visual modality (the DS) was analyzed, the size of the decrement produced by an auditory suffix - to be called the auditory interference score (AIS) - was examined. AIS was calculated for each subject by subtracting the score on the auditory suffix condition from that on the auditory control condition. As with DS, rank order correlations were performed between the auditory control condition and AIS for the age groups separately. The pattern of results differed such that the young showed a significant negative correlation (rr.(nir)= - 32) and the elderly showed a small nonsignificant positive correlation (rsa(ois) = + . l 1). This is strongly suggestive that the overall greater susceptibility to auditory interference on the part of the elderly is age specific and not related to immediate memory of auditory stimuli in the absence of interference. In the young, however, the group shows less overall interference; it appears that individuals showing relatively greater interference may also show relatively poorer overall auditory performance in the control condition. Further tests showed the overall decrement produced by the auditory suffix on auditory stimuli to be significantly greater for the elderly than for the young; t (30) = 2.93. The overall dcecrement produced by the visual suffix on visual stimuli was not different for young and elderly; I(30) = 3 6 .
8
MANNINGIGREENHUT-WERTZ
Downloaded by [UQ Library] at 00:06 20 November 2014
Discussion Several conclusions may be drawn from this experiment, which we believe to be the first in which the suffix paradigm was used to compare visual and auditory stimulus presentation with young and elderly subjects. First, performance on certain aspects of the paradigm differs for the two age groups. As predicted, in the control condition, the elderly show superior auditory over visual recall. It was hypothesized that the source of these difference might be greater difficulty in stimulus recoding for elderly subjects. If visual stimuli need to be recoded into an auditory representation, while auditory stimuli can be processed more directly from the physical signal, this recoding effect would contribute to the relative modality differences between age groups. To explain the suffix data, the effects of auditory and visual suffixes will be considered first. Although both groups show both terminal and nonterminal auditory suffix effects, the elderly show preterminal auditory suffix effects that are significantly greater than those shown by the young. These effects extend further back in the sequence for the elderly. Although a very small atypical visual suffix effect is present for the elderly at the terminal position, a finding not unknown in the young (see, for example, Hitch [1975]), neither group shows visual suffix effects of any size. These modality differences are of theoretical importance and indicate the 'deficits of the elderly are not attentional in a general sense. Instead they are modality specific. This then contradicts the hypothesis that the requirement of exclusion of a distractor always results in relatively worse performance in the elderly than in the young. What then might explain the pattern of results that were obtained here? As before, we hypothesize that recoding difficulties in the elderly in combination with an inability to exclude irrelevant stimuli when they are presented in a mode that is easy to process can explain the entire pattern of results in this experiment. This explanation is offered for both the relative auditory superiority and the greater susceptibility to preterminal interference from an auditory suffix found in the elderly. As for other information obtained in this study, the large negative correlations of the DS with visual performance combined with thle small positive correlations with auditory performance suggest that a large DS may relate to generalized coding difficulties. Subjects in both groups with relatively large DSs show relatively poor recall of visual stimuli for their age. For the elderly, this may be a new situation related to aging or it may be an exaggerated manifestation of a lifelong difficulty. For the young, it may represent a lifelong deficit. Regardless, however, of the source, because of the differential correlations of DS with the visual and auditory control conditions, DS is worthy of further study as a possible marker of coding deficits. Unlike DS, AIS shows differential patterns of results for young and elderly subjects. It is virtually uncorrelated
with performance in the auditory control condition in the elderly, but is negatively correlated in the young. This lack of correlation, combined with the overall larger size of AIS, suggests that susceptibility to auditory interference is a more common deficit in the elderly. In the young, however, the high negative correlation between AIS and the auditory control is suggestive of a deficit that is also of interest to investigate. Before concluding, there are three further issues to mention. First, although there is a general decrement and some previously described relative differences in the performance of elderly as compared with young subjects, most standard effects in this paradigm are present and d o not interact with age. This is consistent with previous findings that the modality effect and the auditory endof-sequence suffix effect are extremely robust over many conditions. This experiment then adds the normal elderly - a group which may have attentional and possibly other deficits - to those exhibiting standard findings. Second, this experiment used short sequence lengths which were easy for most of the subjects in both the young and elderly groups. It is of importance to see whether the same findings with respect to similarities and differences in modality and suffix effects between the two groups hold with longer sequences. Regardless, however, of any differences in the data patterns which may appear in other experimental situations, these data add to the literature in this area. Finally, there is an important set of data to be accounted for. Parkinson and Perey (1980) performed a study using auditory stimuli in the suffix paradigm which compared young and elderly subjects. These researchers report data showing that both the shape of the serial position curve and the type of suffix effect obtained depend on the interaction between sequence length and digit span for individual subjects. Of greatest importance for the current study were findings that, when digit spans were not equated, performance was poorer for elderly subjects. However, when span was equated by use of subjects with five and six item spans, the size of the auditory suffix was the same for both age groups. Parkinson and Perey conclude that it is a difference in immediate memory span that leads to the differential performance between young and elderly when digit span is not equated. We question these conclusions. Control performance in the suffix paradigm is similar to a span condition and is certainly a measure of immediate memory. The only way that our results could have been obtained under the Parkinson and Perey hypothesis is if the young subjects were showing large variability in their memory span while the elderly were not. Examination of our auditory control condition indicates this is definitely not the case. Both the mean performance and the standard deviations are very similar for the elderly and young, respectively, Ms = 6.80 and 7.16, SDs = .92 and .71. A second problem with the Parkinson and Perey view is that they used young subjects with unusually short auditory spans. Given our negative correlation between
MODALITY AND SUFFIX EFFECTS
AIS and the control condition in the young and the lack of such an effect in the elderly, it may be that Parkinson and Perey chose young subjects who were unusually greatly affected by auditory interference and compared them to elderly subjects where AIS was not correlated with immediate memory. Thus we conclude that the findings here open up a variety of areas of research with respect to the effects of aging on immediate recall. They also suggest some areas of work of possible importance in detecting recoding deficits in young and older subjects.
Downloaded by [UQ Library] at 00:06 20 November 2014
References Arenberg, D. (1968). Input modality in short-term retention of old and young adults. Journal of Gerontology, 23, 462-465. Arenberg, D. (1977). The effects of auditory augmentation on visual retention for young and old adults. Journal of Gerontology, 32, 192-195. Baddeley, A., & Hull, A. (1979). Prefix and suffix effects: Do they have a common basis? Journal of Verbal Learning and Verbel Behavior, 18, 129-140. Balota, D.A., & Engle, R.E. (1981). Structural and strategic factors in the stimulus suffix effect. Journal of Verbal Learning and Verbal Behavior, 20, 346-357. Campbell, R., & Dodd, B. (1980). Hearing by eye. Quarterly Journal of Experimental Psychology, 32, 85-99. Conrad, R. (1964). Acoustic confusions in immediate memory. British Journal of Psychology, 55, 75-84. Cooley, R.K., & McNulty, J.A. (1967). Recall of CCC trigrams over short intervals as a function of mode of presentation. Psychonomic Science, 9, 543-544. Crowder, R.G., & Morton, J. (1969). Precategorical acoustic storage (PAS). Perception and Psychophysics, 5, 365-373. Engle, R.W. (1980). The suffix effect: How many positions are involved? Memory and Cognition, 8, 247-252. Folstein, M.F., Folstein, S.E., & McHugh, P.R. (1975). Mini Mental State: A practical method for grading the cognitive state for the clinician. Journal of Psychiatric Research, 12, 189- 198. Freund, J.S., & Witte, K.L. (1986). Recognition and frequency judgments in young and elderly adults. American Journal of Psychology, 99, 91-102. Glenberg, A.M. (1984). A retrieval account of the longterm modality effect. Journal of Experimental Psychology:Learning, Memory, and Cognition, 10, 16-31. Greenberg, S.N., & Engle, R.W. (1983). Voice change in the stimulus suffix effect: Are the effects structural or strategic? Memory and Cognition, 11, 551-556. Greene, R.L., & Crowder, R.G. (1984). Modality and suffix effects in the absence of auditory stimulation.
9
Journal of Verbal Learning and VerbalBehavior, 23, 371-382.
Hitch G.J. (1975). The role of attention in visual and auditory suffix effects. Memory and Cognition, 3, 501-595.
Le Breck, D.B., & Baron, A. (1987). Age and practice effects in continuous recognition memory. Journal of Gerontology, 42, 89-91. McGhie, A., Chapman, J., & Lawson, J.S. (1965). Changes in immediate memory with age. British Journal of Psychology, 56, 69-75. Manning, S.K. (1977). Ratings of the auditory and visual similarity of consonants: Implications for research. Behavior Research Methods and Instrumentation, 9, 495-498.
Manning, S.K. (1980). Tactual and visual suffix effects. Quarterlj Journal of Experimental Psychology, 32, 257-267.
Manning, S.K. (1985). Suffix effects on visual sequences presented with the Optacon. A test of sensory, attentional, and changing state theories. Bulletin of the Psychonomic Society, 23, 1-4. Manning, S.K., Koehler, L., & Hampton, S. (1990). The effects of recoding and presentation format in recency suffix effects. Memory and Cognition, 18, 164-173. Manning, S.K., & Robinson, 1.1. (1989). Recency and suffix effects as a function of auditory confusability and set size. American Journal of Psychology, 102, 495-5 10.
Manning, S.K., & Schreier, H. (1988). Recency and suffix effects in pictures as a function of recall method. American Journal of Psychology, 101, 97-109. Nairne, J. (1988). A framework for interpreting recency effects in immediate serial recall. Memory and Cognition, 16, 385-393. Parkinson, S.R., & Perey, A. (1980). Aging, digit span, and the stimulus suffix effect. Journal of Gerontology, 35, 736-742. Penney, C.G. (1975). Modality effects in short-term verbal memory. Psychological Bulletin, 82, 68-84. Penney, C.G., &Butt, A.K. (1986). Within- and betweenmodality associations in probed recall: A test of the separate-streams hypothesis. Canadian Journal of Psychology, 40, 1- 11. Shand, M.A., & Klima, E.S. (1981). Non auditory suffix effects in congenitally deaf signers of American Sign Language. Journal of Experimental Psychology: Human Learning and Memory, 7, 464-474. Spoehr, K.T., & Corin, W.J. (1978). The stimulus suffix effect as a memory coding phenomenon. Memory and Cognition, 6, 583-589. Taub, H.A. (1975). Mode of presentation, age, and shortterm memory. Journal of Gerontology, 50, 56-59.
