J Am Acad Audiol 26:220-228 (2015)

The Effect of Background Babble on Working Memory in Young and Middle-Aged Adults DOI: 10.3766/jaaa.26.3.3 Michelle T. Neidleman* Use Wambacq* Joan Besing* Jaclyn B. Spitzer*t Janet Koehnke*

Abstract Background: Background noise has been found to negatively affect working memory. Numerous stud­ ies have also found that older adults perform more poorly on working memory tasks than young adults (YA). Hearing status has often been a confounding factor in older individuals. Therefore, it would be beneficial to investigate working memory functions in adverse listening conditions early in the aging process (i.e., middle-age), when hearing function is relatively unaffected. Purpose: The focus of this study was to determine the influence of background babble on working mem­ ory in YA and middle-aged adults (MA) with normal hearing. Research Design: Before testing was begun, we established that all participants could correctly identify words in a degraded experimental testing environment with 100% accuracy. Then, the participants lis­ tened to lists composed of five pairs of words in quiet and in 20-talker babble. After the final word pair, the participants were cued with the first word of one of the previous five word pairs. The participants were required to write down the second word of the pair. The percent correct scores for each of the five serial positions were analyzed comparing the two listening conditions for YA and MA. Ten YA and ten MA with normal hearing between 250-8000 Hz and a score of at least 26/30 on the Mini-Mental State Examination participated in the study. As different cognitive processes are used for initial, middle, and final serial posi­ tions, averaged scores were obtained for Positions 2 and 3 and for Positions 4 and 5. Subsequently, repeated-measures analyses of variance (ANOVAs) were conducted on mean scores of correctly recalled word pairs with serial positions (initial, middle, and final) and listening condition (quiet, babble) as the within-participant variables and age group (YA, MA) as the between-participant independent variable. This OMNIBUS repeated-measures ANOVA was then followed up with separate repeated-measures ANOVAS for the initial, middle, and final positions. Results: Correct recall scores were lower for early positions compared with the latter positions, irrespec­ tive of listening condition. For Position 1, YA— but not MA— performed significantly better in babble than in quiet. For the middle positions (Positions 2 and 3), MA performed significantly more poorly than the YA irrespective of listening condition. For the final positions (Positions 4 and 5), no age differences or effects of listening condition were found. Conclusions: The results indicate that both YA and MA have trouble recalling earlier pieces of infor­ mation in quiet and in babble. However, MA exhibited significantly poorer recall scores than YA in babble for Position 1, which suggest that cognitive processes related to memory encoding and retrieval are differ­ ent in background babble for MA and YA. Key Words: Cognition, working memory, middle-aged, noise/adverse effects, aging/psychology, task performance and analysis Abbreviations: ANOVA = analysis of variance; MA = middle-aged adults; SBR = signal-to-babble ratio; YA = young adults

‘ Montclair State University, Montclair, NJ; -(Columbia University, New York, NY Michelle T. Neidleman, 354 Conklintown Road, Ringwood, NJ 07456; Phone: 973-747-2891; E-mail: [email protected]

SSQ

Effect o f Background Babble on Working Memory/Neidleman et al

INTRODUCTION orking memory is what enables humans to store, process, and manipulate information. Working memory is necessary to perform more complex tasks of reasoning, language comprehen­ sion, and learning, as it is the first stage in transferring and encoding information into long-term memory stor­ age. Hence, working memory is used in everyday situa­ tions by individuals of all ages. This study was designed to investigate age-related differences in working mem­ ory by evaluating the effects of background babble on an auditory working memory task. We included middleaged adults (MA) and compared their working memory performance with th a t of young adults (YA). It is known th at in degraded listening conditions, even subtle tem­ poral and/or spectral changes in the peripheral and central auditory system, may result in an increased number of errors in speech perception, even when pure-tone audio­ metric results are within normal limits (e.g., Ruggles et al, 2012). In turn, these errors can negatively influence higher-order cognitive processing such as memory. To minimize the confounding impact of auditory decline on the study of auditory working memory, we ascertained that experimental stimuli presented in the background babble were perfectly identifiable by all participants. In the present study, we tested working memory by m easuring serial position curves, which use a list of items an individual needs to recall at a later time. Typ­ ically, the last few serial positions are recalled with greater accuracy (e.g., Serial Positions 4 and 5 in the experimental paradigm used in this study). This im­ provement in performance has been dubbed the recency effect, which refers to the increased ability of an individual to recall the most recently presented stimuli (Madigan and McCabe, 1971). It is believed that the last pieces of information are remembered with increased accuracy and frequency because they are still stored in short-term memory (Atkinson and Shiffrin, 1968; Murphy et al, 2000). On the other hand, there are times when the first item of a fist is recalled with greater accuracy (e.g., Serial Position 1 in this study). This has been called the primacy effect. This is thought to occur because the individual has spent more time rehearsing and encoding the informa­ tion into short-term memory storage to facilitate its transference into long-term memory storage (e.g., Serial Position 1 in the current study) (Atkinson and Shiffrin, 1968; Murphy et al, 2000). In contrast, the information immediately following the first item(s) (e.g., Serial Posi­ tions 2 and 3 in the current study) is recalled less fre­ quently because there is little space left in short-term memory storage to rehearse the items (Pasterino and Doyle-Portillo, 2012). This often causes the information from the middle of the list to be forgotten. Environmental, physiological, and psychological factors may negatively af­ fect one or all of the components associated with memory

W

encoding and the transfer of information for use with working memory (Reuter-Lorenz and Sylvester, 2005; Atkinson and Shiffrin, 1968; Glanzer, 1972; Brickman and Stern, 2009). In the current study, we will evaluate separately the response patterns for initial (Serial Posi­ tion 1), middle (Serial Positions 2 and 3), and final serial positions (Serial Positions 4 and 5), as they each involve distinct cognitive processes. Background noise is a common environmental factor, which can affect an individual’s memory performance. Numerous studies have demonstrated th a t background noise, such as babble, negatively affects memory abil­ ities and other cognitive abilities by increasing response times and decreasing accuracy (e.g., Neath, 2000; Baddeley and Salame, 1986). This effect is known as the irrelevant speech effect (Neath, 2000). It is im portant to recognize and study the impact of background babble on working memory because working memory is used on a daily basis and very often in the presence of background noise (Tremblay et al, 2000). In the current study, we will evaluate the effect of 20-talker babble as a factor affect­ ing serial position recall. Age is another factor th at appears to affect cognitive abilities, such as working memory. Studies have shown th at working memory declines with age (e.g., Murphy et al, 2000; Reuter-Lorenz and Sylvester, 2005). Some researchers believe th a t this is caused by the aging indi­ vidual’s inability to inhibit irrelevant stimuli, as well as slower processing skills (Lustig et al, 2007; Bell and Buchner, 2007). Others have found th a t the declining sensory function, as in temporal processing, of older adults may contribute to their cognitive declines in areas such as memory or processing speed (Baldwin and Ash, 2011; Ruggles et al, 2012). Lindenberger and Baltes (1994) attributed 93% of age-related differ­ ences in intelligence to auditory and visual perception. This is particularly important to keep in mind when evaluating the role of aging on working memory be­ cause the differences noted may be the result of sensory declines, cognitive changes, or a combination of factors (Murphy et al, 2000). In fact, when efforts are made to equate the difficulty level of the listening condition for older adults and YA, observed aging declines for higher cognitive skills, such as memory, diminish or disappear (e.g., Murphy et al, 2006). Interestingly, preserved cog­ nitive skills, such as semantic processing, have been found to counteract age-related declines in processing (Light and Singh, 1987; Pichora-Fuller, 2008). In the current study, we evaluated working memory skills while ascertaining th at speech perception accuracy in background 20-talker babble is equal between the age groups tested. Murphy et al (2000) completed a number of serial position studies looking at the aging and noise effects on memory. As the signal-to-babble ratio (SBR) decreased, the percent correct of the first three serial positions

S S I

Journal of the American Academy o f Audiology/Volume 26, Number 3, 2015

decreased for both YA and older adults. The recall of the final two serial positions was not affected by the babble, which suggested th at babble did not affect the recency effect in either age group. Both groups demonstrated a greater decline in memory for the first few items of information when compared with the final two serial positions. Young adults performed significantly better than the older adults in quiet for the first three serial positions, but the groups performed similarly in the final two serial positions. Lastly, the performance of the YA in babble was similar to th at of the older adults in quiet. Murphy et al (2000) suggested th at the effect of noise in the YA was similar to th at of aging in the older adult group. Both factors negatively affected the encod­ ing ability and, in turn, the recall ability of the partic­ ipants. These findings imply degradation in working memory capacity with age. In summary, research indicates th at background noise such as babble negatively affects the performance of older adults and YA on working memory tasks (Surprenant, 2007; Murphy et al, 2000). However, the effects tend to be greater for older adults whose cog­ nitive abilities begin to degrade as part of the normal aging process (Murphy et al, 2000; Reuter-Lorenz and Sylvester, 2005). Although research in this area has commonly studied older adults (age >60 yr), early aging effects on higher cognitive tasks such as memory, reasoning, and spatial visualization also occur in MA (Salthouse, 2009). Difficulties perceiving speech in background noise have been documented in MA with essentially normal hearing acuity (Heifer and Vargo, 2009). The participants in their study had hearing within normal limits from 250-4000 Hz, with mild hear­ ing loss at 6000 and/or 8000 Hz. Speech recognition dif­ ferences in noise were partially attributed to temporal processing changes. The question remains whether these perceived difficulties could be attributed to a presenescent decline in working memory. It is unclear whether background babble affects MA performance on working memory tasks to a greater extent than it does on YA per­ formance. Both age groups in our study had audiometric results within normal limits from 250-8000 Hz, and we established th at all participants were able to identify spoken words perfectly in our most difficult experimental listening condition. As a result, we were able to study the impact of cognitive aging independent of sensory pro­ cessing abilities. Typical learning environments and everyday listen­ ing conditions involve working memory tasks in quiet and background noise; therefore, the present study was conducted in quiet and in 20-talker babble (+6 dB SBR). We wanted to determine whether a break­ down in working memory performance would occur in babble. As in the study by Murphy et al (2000), stimuli in­ cluded lists of five pairs of unassociated words to eliminate the possible use of semantic context as a compensatory

222

strategy, which would put the MA at an advantage (Light and Singh, 1987; Pichora-Fuller, 2008). We hypothesized that the performance of both groups would be best in the final serial positions (Positions 4 and 5), followed by the first position (Position 1) and then worst in the middle serial positions (Position 2 and 3). On the basis of the results reported by Murphy et al (2000), we predicted that the presence of background babble would negatively affect working memory performance for both YA and MA. However, the effect would be greater in the MA when com­ pared with the performance of the YA. Serial position per­ formance curves in quiet were predicted to be comparable between the age groups. In the presence of background babble, the middle-aged curves, like those of the older adults in previous studies, were expected to show a greater decline in performance for the first and middle serial posi­ tions before rising to performance levels such as those of the YA in the final serial positions (Murphy et al, 2000).

METHODS Participants A total of 10 YA (YA group) between 22-30 yr old (mean age = 25 yr) and 10 MA (MA group) between 46-60 yr old (mean age = 55.1 yr) participated in this study. There were 2 males and 8 females in the YA group and 10 females in the MA group. All participants had hearing thresholds equal to or less than 25 dB HL from 250-8000 Hz bilaterally and were native English speakers. None of the participants had a history of neu­ rological, language, or auditory processing disorders. To exclude the possible confounder of declining sensory processing skills, we tested word recognition ability by using 25 words from the serial position working memory task originally developed by Murphy et al (2000). Words were presented in the sound field at +6 dB HL SBR and were spoken by the talker used during experimental testing. Background babble was obtained from Auditec (AudiTec 20-talker babble). Target words and background babble were presented from the same loudspeaker at 0° azimuth, as it has been shown that presentation of background babble and target speech from the same spatial location is more difficult than spatially separated sound sources (e.g., Dubno et al, 2002). All participants obtained perfect word recognition scores. Furthermore, a score of 26 or higher on the Mini-Mental State Examina­ tion (Folstein et al, 1975) was obtained by all partici­ pants. All participants provided informed consent in accordance with the Montclair State University Institu­ tional Review Board.

Stimuli A total of 200 word pairs, consisting of unassociated bisyllable words, were recorded by a female speaker

Effect of Background Babble on W orking Memory/Neidleman et al

(23 yr old) of Standard American English using Adobe Audition. These word pairs were also used by Murphy et al (2000) and Heinrich et al (2008). The materials were recorded in a double-walled, sound-attenuated booth, using a single-channel setting with a sampling rate of 44, 100 Hz, and 16-bit resolution. Words were edited using Adobe Audition and were equated for aver­ age root mean square amplitude. Before presentation, the level of the stimuli was cali­ brated at ear level at the levels used during participant testing. Sound level measures were taken in the sound field, a t 0° azimuth, 1.4 m from the listener and at ear level using a 1000 Hz calibration tone with the same root mean square amplitude values as the working memory word stimuli. Stimuli were presented at an average intensity of 66 dB A from a loudspeaker located at 0° azimuth, 1.4 m from the listener. To ensure running babble throughout the duration of the degraded listening condition, a 20-person babble previously recorded by AUDITEC was concatenated digitally onto CDs. The output of two CD players (SONY HDMI) was sent to two power amplifiers (Crown XLS202) before being presented at +6 dB SBR (i.e., 60 dB A) from four loudspeakers (Realistic Minimus), which were located at ear level, 1.4 m from the listener a t 45°, 90°, 270°, and 315° azimuth. Procedure In the sound-attenuated booth, each participant listened to lists composed of five pairs of words. A computer monitor placed in front of the participant indicated the start of each list with the word “READY.” Two practice runs corresponding to the two listening conditions were given at the start of the test. The prac­ tice runs consisted of two lists with five pairs of words th at were not used during actual testing. Within each list, pairs of words were separated by a 4,000 msec

gap and the words in each pair were separated by a 100 msec gap. Following the last pair of words in a list, there was a 500 msec gap before a 1000 Hz warning tone (1,000 msec) was presented. The warning tone was fol­ lowed by a 500 msec gap. Participants were then cued with the first word of one of the previous five word pairs and were required to write down the second word of that word pair. After writing down their response, the par­ ticipants pressed a button on the controller to move onto the next trial. No feedback was provided to the partic­ ipants. The same procedure was used during actual testing, when participants listened to a total of 40 lists consisting of five pairs of words (Table 1); 20 lists were presented in quiet and 20 lists in the 20-talker babble at +6 dB SBR. We created a single sequencing file for stimulus pre­ sentation to all participants. First, the 40 lists were divided into eight blocks consisting of 5 lists each. Within each block, we pseudo-randomly selected the sequence of serial position testing, with the restriction th at each serial position (1, 2, 3, 4, or 5) was tested only once. We then created a pseudo-randomized sequence of listening conditions and used this sequence to assign each block to a listening condition. No more than two consecutive blocks were presented with the same listen­ ing condition. When a block was presented in babble, the background babble was played continuously through­ out the block. Blocks were presented in the same order for all participants; however, listening conditions were counterbalanced. For example, if Participant 1 listened to block 1 in quiet, Participant 2 listened to block 1 in babble. Correct responses were those th at matched the target word. However, responses th a t were homophones of the target word, like “kernel” and “colonel,” and those th at added an additional morphological ending to the target word, like “trouser” and “trousers,” were also marked as correct. The total test time was approximately 25 min.

Table 1. Illustration of the Study Paradigm

Word Pairs Presented

m

Baker

100 ms gap

Problem

4000 ms gap

m

Climate

100 ms gap

Kindness

40000 ms gap

rjl

Custom

100 ms gap

Dealer

40000 ms gap

H

Courtroom

100 ms gap

Payment

4000 ms gap

Wrinkle

100 ms gap

Raider

4000 ms gap

m

500 ms gap Word Pair Tested

|2 |

Climate

1000 ms 1000 Hz tone

500 ms gap

Patient Response Time (Correct Answer: Kindness)

Notes: Five word pairs, corresponding to Serial Positions 1-5, were presented. After the last word pair, a 1,000 msec 1000 Hz tone was presented before one of the first words of the five previously heard words pairs was presented. The participant was expected to provide the second word of the word pair.

SS3

Journal of the American Academy of Audiology/Volume 26, Number 3, 2015

Figure 1. Mean score of correctly recalled word pairs as a func­ tion of age. Each serial position was tested four times, which is reflected on the y-axis.

RESULTS n order to evaluate whether there was a performance difference between the two age groups, the mean score of correctly recalled word pairs was collapsed across lis­ tening conditions (quiet and + 6 dB SBR) and was sepa­ rated out by serial position as a function of age in Figure 1. As hypothesized, the serial position curves for both YA and MA followed a similar trend of reduced performance in the middle positions (Serial Positions 2 and 3) com­ pared with the initial position (Serial Position 1), before rising to greater accuracy in the final positions (Serial Positions 4 and 5) (Fig. 1). Visual observation of the curves in Figure 1 shows that the MA had a greater decline in performance for the initial and middle positions than the YA, irrespective of listening condition. The effect of listening condition was evaluated in Fig­ ure 2. This figure reveals the difference in correct recall scores across the initial, middle (the average of Serial Positions 2 and 3), and final (the average of Serial Posi­ tions 4 and 5) serial positions as a function of age and listening condition (quiet and +6 dB SNB). As we had predicted, the shape of the serial position curves sup­ ports the recency effect in th a t the last two items of information were recalled with greater accuracy than those in the beginning or middle of the sequence. Regard­ less of listening condition, the MA recall scores were lower than the YA recall scores for both listening condi-

I

tions in the middle positions (Serial Positions 2 and 3). Interestingly, the two age groups performed similarly in quiet for first position, but significantly different in babble. The YA performed significantly better than the MA in babble in Position 1. We conducted an OMNIBUS repeated-m easures analysis of variance (ANOVA) on the correct recall scores with position (initial, middle, and final) and lis­ tening condition (quiet, babble) as w ithin-participant variables; age group served as the between-participant variable. For the middle and final positions, correct recall scores were averaged between Positions 2 and 3 (for mid­ dle position) and between Positions 4 and 5 (for final posi­ tion). A Huynh-Feldt correction was applied for factors with more than two levels. We observed significant main effects of position [F(2 ,3 6 ) = 16.581, p < 0.001] and age group [F(1;18) = 7.054, p = 0.016]. We also found a significant interac­ tion effect of position by age group LF(2 ,3 6 ) = 5.686, p = 0.007], Interaction effects of background by age group [F(lii8) = 3.620, p = 0.073] and background by position by age group [F(2,29.295) = 2.122, p - 0.081] approached significance. Subsequently, we conducted separate repeated-measures ANOVAs on the correct recall scores for initial, middle, and final positions with age as the between-participant variable, and listening condition as the within-participant variable. For Position 1, there was a main effect of group [F(i,i8) = 6.538,p = 0.02], as well as an interaction effect of babble by group [F(iil8) = 5.158, p = 0.036] (Fig. 3). Post hoc independent-samples two-tailed f-tests comparing the YA versus MA revealed no significant differences in quiet, but a significant difference in babble (f = 3.226, df = 18, p = 0.003). The YA and MA performed similarly in quiet, but in babble the YA performed significantly better than the MA. A post hoc paired-samples f-test comparing performance in quiet versus babble for each age group showed a significant difference in the performance of the YA (t = -2.689, df = 9, p = 0.025), but there was no significant difference in the performance of the

- • “ MA Quiet -O -M A Babble - * - Y A Quiet -A -Y A Babble

YA

MA Age Group

Figure 2. Mean score of correctly recalled word pairs as a func­ tion of age and condition. Each serial position was tested four times, which is reflected on the y-axis.

SS4

Figure 3. Comparison of age and listening condition for the first position. Serial Position 1 was tested four times in each listening condition, which is reflected on the y-axis.

Effect of Background Babble on Working Memory/Neidleman et al

Figure 4. Comparison of age and listening condition for middle positions (Serial Positions 2 and 3). Each serial position was tested four times in each listening condition, and the average score is reflected on the y-axis.

MA. YA performance improved in background babble for Position 1. For the middle position (averaged scores for Positions 2 and 3), the repeated-measures ANOVA revealed a main effect of group [F(1>18) = 9.740, p = 0.006] (Fig. 4), and no effect of listening condition. The MA performed worse than the YA, irrespective of listening condition. For the final position (averaged scores for Positions 4 and 5), no significant findings were obtained (Fig. 5). The results illustrate th at the YA were better at recalling earlier pieces of information stored in long-term memory than the MA independent of listening condition. This was true when encoding was enabled, as in Posi­ tion 1, as well as when encoding was not enabled, as in the middle positions (Serial Positions 2 and 3). However, when the information was stored in short-term memory, as in the final positions (Serial Positions 4 and 5), both age groups performed equally well.

DISCUSSION reviously, it has been shown th at many of the lis­ tening difficulties experienced by older listeners may be an effect of cognitive aging and/or a sensory loss (Murphy et al, 2000; Cevera et al, 2009). Studies have

P

YA

MA Age Group

Figure 5. Comparison of age and listening condition for the final positions (Serial Positions 4 and 5). Each serial position was tested four times in each listening condition, and the average score is reflected on the y-axis.

shown a regression in working memory ability with age, but factors such as background noise and hearing loss have been found to exacerbate working memory capacity (Murphy et al, 2000). We aimed to determine if MA with normal-hearing sensitivity would have poorer perfor­ mance than YA on a serial position working memory task and if the background babble would further affect their working memory capacity. The results of this study included both expected and unexpected findings. As predicted, both age groups per­ formed best in the final serial positions followed by the first position and then, lastly, the middle positions. The YA outperformed the MA for initial and middle serial positions, but performed similarly in the final serial positions, independent of listening condition. Surpris­ ingly, the YA performed significantly better in Position 1 in babble than they did in quiet. They also performed significantly different from the MA in babble. Another unexpected finding was th at the MA did not exhibit a significant difference in their performance in babble and in quiet. In the present study, we did not find evidence of a pri­ macy effect. Both age groups displayed more difficulty remembering the earlier word pairs in the first and middle positions. In the first and middle positions, there was no effect of babble, but there was an effect of age. As discussed earlier, the first position is rehearsed without interruption before entering long-term memory stor­ age, whereas the middle positions have less short-term memory space and time to rehearse and encode the information into long-term memory storage. MA scores were significantly lower than YA scores, suggesting th at MA may not have been as effective as YA in storing, transferring, or retrieving the information from short-term and long-term memory storage. Although the MA per­ formed significantly more poorly than the YA in the initial and middle serial positions, the performance in the final positions improved similarly for the MA and YA Both age groups exhibited the recency effect because they re­ called the latter pieces of information with greater accuracy. The last few items were likely easier to access because they were still stored in short-term memory (Murphy et al, 2000). The lack of an age effect for the last serial positions is consistent with the literature, which indicates th at the recency effect is fairly resistant to the effects of aging (Craik, 1994). On the other hand, earlier information, as in the first position or in the middle positions, requires retrieval from long-term memory storage and is affected by aging (Craik, 1968; Glanzer, 1972). There are various theories as to why long-term memory declines with age. The most cited theory suggests that the decline is related to a reduction in processing resources available during the encoding, storage, and/or retrieval stage (Brickman and Stem, 2009). When evaluating the interaction effect of listening condition by age group for each serial position, we only

225

J o u rn a l o f th e A m erica n A cad em y o f A u d io lo g y /Volume 26, Number 3, 2015

obtained a significant difference in performance between the YA and MA for Position 1. For Position 1, YA performed significantly better in babble than in quiet, but this improvement in performance in babble was not observed in the MA. The storage load for Position 1 is minimal, as no other positions have been rehearsed and encoded into short- or long-term memory. It is known th at difficult listening situations increase the amount of listening effort (Pichora-Fuller, 2003). For example, Gisselgard et al (2004) found that irrelevant speech caused increases in prefrontal cortex activity dur­ ing working memory tasks. Hence, it is plausible that the YA used greater cognitive resources and working memory strategies to overcome the presence of babble for word pairs in Position 1. The increased listening effort, however, seemed most apparent for Position 1, as no differences in performance were observed between babble and quiet lis­ tening conditions for the middle or final positions. Both Hamilton et al (1977) and Daee and Wilding (1977) presented findings in support of enhanced pro­ cessing in difficult listening conditions. This may have been because the stress of the noise amplified the par­ ticipants’ ability to store and recall the information or because they rehearsed more in noise than they did in quiet (Hamilton et al, 1977; Daee and Wilding, 1977). It is known th a t items in Position 1 of a serial position working memory task are rehearsed more than the sub­ sequent items. Therefore, it is likely th a t the YA in our study relied more heavily on rehearsal strategies in babble than in quiet, which led them to perform better in babble than in quiet in Position 1. It is of note th at the system is taxed more when there is extra information to maintain and rehearse (Atkinson and Shiffrin, 1968). Hence, after Position 1, the rehearsal process becomes more difficult. The YA still outperformed the MA, but the advantage of listening in babble observed in YA for Position 1 was not observed in subsequent posi­ tions. Smith (1985) found a similar effect in his study when his participants performed better in noise. He sug­ gested that the presence of noise forced the participants to use a dominant strategy of recall, which they continued to use in easier conditions, such as quiet. So, it is believed that the improved performance of the YA in babble for Position 1 in the present study was the result of both an increased listening effort and an enhanced rehearsal process. It is also possible that the YA continued the use of these strategies in the middle positions, leading to con­ tinued better performance than the MA, albeit irrespec­ tive of listening condition. It appears that the MA either did not make greater use of this strategy in background babble or, if they tried, were not as successful at encoding and/or retrieving the information. As stated earlier, no effects of listening conditions were found for the middle or final serial positions. This finding, and the high accuracy of recall for final posi­ tions, suggests th at the babble did not degrade the sen­

22B

sory perception of the words. The absence of a listening condition effect on the remaining serial positions may be the result of the SBR that was used in this study. The +6 dB SBR may not have been sufficiently difficult to affect the working memory abilities of either age group. Surprenant (2007) found that at a -10 dB signal-to-noise ratio level, performance differences and declines begin to occur from ages 30-80 yr. A more negative SBR may have resulted in significant differences between the quiet and babble conditions in our study. Murphy et al (2000) also found an effect of listening condition with negative SBRs. A negative effect of noise on serial position performance for Positions 1-3 was found in the YA when the researchers used -5 and -10 dB SBRs. Working memory ability may decline and the amount of mental effort required to im­ prove performance may increase, as the listening environ­ ment becomes more challenging. The results of this study do not enable us to deter­ mine whether the MA were simply not expending as much effort as the YA when performing the working memory task in babble, whether the MA were expending as much effort as the YA, but simply not as successful, or whether they did not have the necessary cognitive resources to perform with greater accuracy in babble. The effect of performance on age in Position 1 suggests that the MA have greater difficulty rehearsing, encoding, and/or transferring information into long-term memory storage in babble. Neuroimaging studies on working mem­ ory have found that YA and older adults are, at times, using different areas of their brain for these tasks, as well as expending different levels of activation (Reuter-Lorenz and Sylvester, 2005; Grady and Craik, 2000). This trend implies that the older adults are recruiting additional cog­ nitive resources in order to complete a task, but even so, they may not be able to perform as well as YA. This cog­ nitive change may be beginning in middle age, which could partially explain the performance differences seen in this study. These results are of clinical im portance because many audiology patients are MA, who also present with complaints of difficulty understanding speech in noisy listening environments. However, audiological evaluations for these individuals often reveal hearing within normal limits. It is important to understand whether these com­ plaints may be associated with cognitive declines because of the normal cognitive aging process or other sensoiy declines related to hearing, such as that of temporal process­ ing, which have not been tested yet via conventional audio­ metric testing (Anderson et al, 2012; Ruggles et al, 2012). In summary, the results of the present study reveal th at cognitive changes and working memory declines occur in middle age. The cognitive components respon­ sible for these differences were not evaluated individually in this study. However, other studies have attributed the differences in working memory ability to factors such as slowed processing, auditory processing declines, impaired

Effect of Background Babble on Working Memory/Neidleman et al

attentional resources, and reduced inhibitory control (Murphy et al, 2000; Surprenant, 2007; Grady and Craik, 2000). Future studies should focus on determining whether the differences in the YA and MA working mem­ ory performance were the result of the encoding process, retrieval process, or both processes. Analysis of reaction times may also be used to reveal differences in processing speed, whereas the inclusion of another task or distracter may be used to measure attentional differences. The results from these future studies can provide insight into which mechanisms are responsible for the decline in audi­ tory working memory with age and perhaps suggest what may be done to retain working memory capacity through­ out the aging process.

CONCLUSIONS s hypothesized, the YA performed better than the MAin the first and middle serial positions and sim­ ilarly for the final serial positions. The background babble affected the performance of the MA in only Position 1, but the babble did not affect their performance in the middle or final serial positions. In comparison, the performance of the YA improved in babble for Position 1, but no dif­ ferences in listening condition performance were ob­ served for the middle or final listening positions. Both age groups had difficulty remembering earlier items, but the MA performed more poorly, which suggests that the MA are beginning to show performance differences related to working memory. This decline is independent of hearing loss, as determined by a hearing and word rec­ ognition test. It is likely that a sensory loss, such as that of hearing, would further compound the difficulty of both age groups to encode, store, and recall information. Al­ though the results of this study indicate a difference in MA and YA working memory performance, the exact cause remains unknown. It may be due to listening effort, speech perception, or other cognitive differences that occur throughout the aging process. Future studies will have to identify which factors are influencing this working memory performance difference, as well as what can be done to preserve working memory as individuals age.

A

REFERENCES Anderson S, Parbery-Clark A, White-Schwoch T, Kraus N. (2012) Aging affects neural precision of speech encoding. J Neurosci 32(41): 14156-14164.

Bell R, Buchner A. (2007) Equivalent irrelevant-sound effects for old and young adults. Mem Cognit 35(2):352-364. Brickman AM, Stern Y. (2009) Aging and memory in humans. In: Squire LR, ed. Encyclopedia o f Neuroscience Vol. 1. New York, NY: Academic Press, 175-180. Cevera TC, Soler MJ, Dasi C, Ruiz JC. (2009) Speech recog­ nition and working memory capacity in young-elderly listen ­ ers: effects of hearing sensitivity. Can J Exp Psychol 63(3): 216-226. Craik FIM. (1968) Two components in free recall. J Verbal Learn Verbal Behav 7(6):996-1004. Craik FIM. (1994) Memory changes in normal aging. Curr Dir Psy­ chol Sci 3(5):155-158. Daee S, Wilding JM. (1977) Effects of high intensity white noise on short-term memory for position in a list and sequence. Br J Psy­ chol 68(3):335-349. Dubno JR, Ahlstrom JB, Horwitz AR. (2002) Spectral contribu­ tions to the benefit from spatial separation of speech and noise. J Speech Lang Hear Res 45(6):1297-1310. Folstein MF, Folstein SE, McHugh PR. (1975) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12(3):189-198. G isselgard J, P etersson KM, Ingvar M. (2004) The irrelev an t speech effect and working memory load. Neuroimage 22(3): 1107-1116. Glanzer M. (1972) Storage mechanisms in recall. In: Bower GH, ed. The Psychology o f Learning and Motivation: Advances in Research and Theory Vol. 5. New York, NY: Academic Press. Grady CL, Craik FIM. (2000) Changes in memory processing with age. Curr Opin Neurobiol 10(2):224—231. Hamilton P, Hockey B, Rejman R. (1977) The place of the concept of activation in human information processing theory: An integra­ tive approach. In: Dornic S, ed. Attention and Performance Vol. 6. San Diego, CA: Academic Press. Heinrich A, Schneider BA, Craik FI. (2008) Investigating the influence of continuous babble on auditory short-term memory performance. Q J Exp Psychol (Hove) 61(5):735-751. Heifer KS, Vargo M. (2009) Speech recognition and tem poral processing in middle-aged women. J Am Acad A udiol 20(4): 264-271. Light LL, Singh A. (1987) Implicit and explicit memory in young and older adults. J Exp Psychol Learn Mem Cogn 13(4):531-541. Lindenberger U, Baltes PB. (1994) Sensory functioning and intelligence in old age: a strong connection. Psychol Aging 9(3):339-355.

Atkinson RC, Shiffrin RM. (1968) Human memory: A proposed system and its control processes. Psychol Learn Motiv 11:249.

Lustig C, Hasher L, Zacks RT. (2007) Inhibitory deficit theory: Recent developments in a “new view”. In: Gorfein DS, MacLeod CM, eds. The place of Inhibition in Cognition. Washington, DC: American Psychological Association, 145-162.

Baddeley A, Salame P. (1986) The unattended speech effect: per­ ception or memory? J Exp Psychol Learn Mem Cogn 12(4): 525-529.

Madigan SA, McCabe L. (1971) Perfect recall and total forgetting: A problem for models of short-term memory. J Verbal Learn Verbal Behav 10:101-106.

Baldwin CL, Ash IK. (2011) Impact of sensory acuity on auditory working memory span in young and older adults. Psychol Aging 26(1):85-91.

Murphy DR, Craik FIM, Li KZH, Schneider BA. (2000) Comparing the effects of aging and background noise on short-term memory performance. Psychol Aging 15(2):323-334.

227

J o u r n a l o f th e A m erican A cadem y o f Audiology/Volume 26, Number 3, 2015

Murphy DR, Daneman M, Schneider BA. (2006) Why do older adults have difficulty following conversations? Psychol Aging 21(1):49—61.

tive and Cerebral Aging. London, UK: Oxford U niversity Press, 186-217.

N eath I. (2000) Modeling the effects of irrelevant speech on mem­ ory. Psychon Bull Rev 7(3):403—423.

Ruggles D, Bharadwaj H, Shinn-Cunningham BG. (2012) Why middle-aged listeners have trouble hearing in everyday settings. Curr Biol 22(15):1417-1422.

Pasterino E, Doyle-Portillo S. (2012) What is psychology?: Essen­ tials. 2nd ed. Belmont, CA: Wadsworth Publishing.

Salthouse TA. (2009) When does age-related cognitive decline begin? Neurobiol Aging 30(4):507-514.

Pichora-Fuller MK. (2003) Cognitive aging and auditory informa­ tion processing. Int J Audiol 42(Suppl 2):S26-S32.

Smith AP. (1985) The effects of noise on recall of lists of associated words. Curr Psychol 4(1):17-21.

Pichora-Fuller MK. (2008) Use of supportive context by younger and older adult listeners: balancing bottom-up and top-down inform ation processing. In t J A udiol 47(Suppl 2):S72-S82.

Surprenant AM. (2007) Effects of noise on identification and serial recall of nonsense syllables in older and younger adults. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn 14(2):126-143.

Reuter-Lorenz PA, Sylvester C-YC. (2005) The cognitive neuro­ science of working memory and aging. In: Cabeza R, Nyberg L, P ark DC, eds. Cognitive Neuroscience o f Aging: L inking Cogni­

Tremblay S, Nicholls AP, Alford D, Jones DM. (2000) The irrele­ vant sound effect: does speech play a special role? J Exp Psychol Learn Mem Cogn 26(6):1750-1754.

SS8

Copyright of Journal of the American Academy of Audiology is the property of American Academy of Audiology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.