This dissertation investigated different factors contributing to age differences in working memory (WM) performance. Younger and older adults participated in five experiments, four on visuospatial WM (VSWM) and one on verbal WM. All addressed methodological issues that may differentially lower older adults’ performance. Experiments 1a and 1b manipulated the administration of a VSWM span task, with participants performing the task under either an ascending format (shortest sets presented first), or an interference-reducing descending format (longest sets presented first). Older adults’ performed significantly better in the descending compared to ascending format, consistent with an age-related susceptibility to proactive interference (PI). By contrast, younger adults did better in the ascending compared to descending condition, possibly due to their ability to benefit from practice and strategy use when easier trials are presented first. Experiment 2 considered how the similarity of task material influenced the build-up of PI and whether or not the combination of two PI-reducing manipulations (i.e., descending format and distinct trials) would further improve older adults’ performance. Distinctiveness helped older adults on the ascending format; however, combining distinct trials and a descending format provided no additional benefit. Experiment 3 considered whether or not synchronizing a VSWM task with an individual’s circadian arousal pattern would moderate interference effects, with the task administered at a peak or off-peak time of day. Peak-time administration improved older adults’ performance on the descending, but not the ascending, condition. Experiment 4 investigated the possibility that the serial order requirement of many WM tasks contributes to age differences in performance. Younger and older adults participated in a verbal WM span task – Operation Span - under either serial order or free recall instructions. Typical age differences were found when order but not free recall was required. Further analyses of the order condition data revealed that older adults were, in fact, recalling the items just as well as young adults, only not in the correct order. Taken together, the findings strongly suggest that age differences found on typical WM span tasks are influenced by numerous factors, such as task presentation, individual circadian arousal patterns, material similarity, and recall instructions.

The processing of information within the visual system is limited by several cognitive and neural bottlenecks. One critical bottleneck occurs in visual working memory (VWM), as the amount of information that can be maintained on-line is limited to three to four items. While numerous theories have addressed this limited capacity of VWM, it is unclear how processing bottlenecks in the initial selection and perception of visual information affect the number or precision of representations that can be maintained in VWM. The purpose of this dissertation was to examine whether early competition for resources within the visual system limits the number or precision of representation that can be maintained in VWM. To establish whether competitive interactions affect VWM, Chapters 1 – 4 tested whether performance on VWM tasks was related to the distance between memory items. The results of these experiments reveal that when objects are presented close together in space, VWM performance is impaired relative to when those same objects are presented further apart. Using a three-component model of continuous responses in a recall task, Chapters 3 – 4 demonstrated that the distance between objects primarily affects the precision of responses, and increases the number of non-target errors. Chapter 5 extended these findings to distractors, demonstrating that multiple distractors affect the precision and accuracy of VWM responses. Chapters 6 – 7 tested how attentional selection can bias memory representations, revealing that objects that are given high attentional priority were reported with greater precision. Finally, Chapters 8 and 9 examined bias-signals as a potential source of individual differences in VWM performance, revealing that high-performers have more precise representations of sub-capacity representations than low-performers. Together, these results reveal that VWM performance is limited by competition for representation within the visual system, and that attention plays a critical role in resolving competition and consequently, determining the contents of VWM.

This study compares the role of cognitive processes in children diagnosed with learning disabilities (LD) through the traditional aptitude-achievement discrepancy model with students diagnosed on the basis of their low achievement alone. Historically, in North American settings, LD has been diagnosed when an individual’s achievement on standardized tests in reading, mathematics, or written expression is substantially lower than the expected level for age, schooling, and level of intelligence (American Psychiatric Association, 2000). As this conceptualization has come under increasing scrutiny, alternate identification methods such as the low achievement/non-discrepant method have been gaining support in the literature (e.g. Siegel, 1999; Stanovich, 2005). A secondary objective of this study is to determine whether identifiable differences exist between the cognitive profiles (WISC-IV) of students diagnosed with reading disability (RD) and mathematics disability (MD). This study also addresses whether the WISC-IV Working Memory Index can be used to differentiate between various categories of students with LD. The findings of this study indicate that the discrepant (DLD) and non-discrepant (NDLD) learning disability (LD) groups could not be distinguished by the WISC-IV Working Memory Index (WMI). Amongst the overall sample of students with LD, those with average or above working memory scores (high) could be differentiated from those with below average working memory scores (low) on the WISC-IV Perceptual Reasoning Index (PRI). Students with LD who had low WMI scores could also be differentiated from those with high WMI scores on four WIAT-II subtests. WMI scores could not be used to differentiate students with Reading Disability (RD), Mathematics Disability (MD) or Generalized Learning Disability (GLD). However, differences between these three LD groups were found on the WISC-IV Verbal Comprehension Index (VCI), Perceptual Reasoning Index (PRI), and marginally on the Processing Speed Index (PSI). Finally, the four WISC-IV Index scores were able to correctly predict group membership in the RD, MD, and GLD groups approximately 70% of the time. / Psychological Studies in Education

Different versions of complex, simple, and running tests of immediate memory span were compared in their ability to predict fluid intelligence (gF). Conditions across memory tasks differed in terms of whether or not a secondary cognitive task was interleaved between to-be-remembered items (complex versus other span tasks), whether or not more items were presented than were ultimately to-be-remembered (running versus other span tasks), and whether presentation rate was relatively fast or slow (running and simple span tasks). Regressions indicated that up to 42.6% of variance in general fluid gF was explained by the memory span measures entered in different combinations. Across comparisons, shared relationships among span tasks accounted for a plurality of total variance in gF. Results indicate that in spite of procedural differences and resulting intra-individual variance in memory performance, the present memory tasks captured largely the same inter-individual variance in working memory capacity, insofar as this is important for higher-order cognition.

Traumatic brain injury (TBI) is frequently associated with both cognitive and behavioural executive dysfunction. Assessment of executive dysfunction has traditionally been measured using tests that were not based on theory and this has been postulated as one reason why there are poor predictive relationships between performance on traditional executive tasks and functional outcome following TBI. Baddeley�s (1996) working memory model may offer a theoretical basis with which to design new executive measures and possibly improve prediction of outcome. Baddeley�s theory has made recent advances in identifying core central executive (CE) processes that are likely to be relevant to TBI. The research presented in the thesis used Baddeley�s proposed coordinative sub function of the CE (and it related dual tasking measure) to assess: (a) whether this theoretically based dual tasking test would be sensitive to TBI injury status and show a significant relationship with functional outcome, (b) whether the theoretically based test would be more sensitive to the above relationships than executive measures that have been traditionally available, and (c) whether more ecologically relevant dual tasks could be developed that reflect the coordinative construct and improve predictive relationships between task performance and real life functional ability. This study found that Baddeley�s Dual Task Test (DTT) measure was sensitive to TBI injury and was related to functional outcome following injury. A questionnaire assessing everyday dual tasking ability was developed as was an in vivo conversation and motor based dual task. The dual tasking questionnaire reflected Baddeley�s coordinative construct, however, applications to an in vivo conversation and motor based dual task were less successful. These ecologically relevant tasks showed a relationship between language dysfluencies and dual tasking and identified avoidance of dual tasking in everyday settings as an important predictor of functional outcome following injury. Overall the DTT and the dual tasking questionnaire showed greater injury sensitivity and stronger relationships with outcome than three executive measures traditionally used in clinical practice. Regression analysis confirmed that the dual based tasks were helpful in predicting a variety of outcomes following TBI, and implications for rehabilitation planning are discussed. Further increases in the predictive power of the dual tasking construct are likely to be achieved when mechanisms of action involved in both laboratory-based and real-life dual tasking are identified. Findings from the current study suggest a range of mechanisms could be involved in dual tasks and these are discussed.

The effective operation of auditory verbal working memory processes is essential to both understanding and communicating in everyday life. These processes are particularly vulnerable to disruption in a wide range of clinical conditions that affect brain functioning. Problems with working memory are frequently the primary reason why people with brain dysfunction are referred for neuropsychological assessment; consequently, verbal memory has been ranked as the second most frequently assessed cognitive ability. However, current psychometric approaches, predominantly serial recall, paired-associate- and list-learning tasks have been designed mainly to detect the presence (and severity) or absence of memory impairment rather than identifying intact or impaired processes. Performance is usually measured by scores of free recall and recognition, with testing procedures that are often lengthy and time-consuming for both patients and clinicians. When assessing working memory, the span tasks have been viewed as “a complete assessment of the function of phonological short-term memory” (Vallar & Papagno, 2002). Digit span, in particular, has come to be regarded as the principal neuropsychological methodology for evaluating the capacity of auditory working memory (Andrewes, 2001), despite doubts about its lack of sensitivity (Lezak, Howieson, & Loring, 2004) and potential to overestimate capacity (Andrewes, 2001). Theoretical advances regarding the underlying component processes of encoding and retrieval have only slowly been incorporated into clinical practice but have not been applied in any systematic or coherent manner. Concepts of working memory have been extended to include semantic, phonological and temporal aspects. One potentially useful approach examined the role of phonological and semantic codes and cues and their susceptibility and immunity to interference effects (Tehan & Humphreys, 1995). Alternatively, another study investigated patients’ ability to accurately retrieve visual information by keeping track of the current episode, differentiating it from previous similar episodes that had the potential to reduce effective performance because of interference effects or poor discrimination (Parkin, Leng, & Hunkin, 1990). The aim of the research reported in this thesis was to develop a more compact cuedretrieval procedure to directly measure the phonological, semantic and temporal aspects of working memory and to examine its clinical usefulness by administering the task to brain injury rehabilitation patients. The conceptual section of the introductory chapter commences with a discussion on memory systems, particularly working memory, and is followed by an examination of some of the core processes involved in successful and unsuccessful remembering (forgetting). The applied section examines traditional and current approaches to memory assessment in clinical practice and their limitations in addressing the clinical needs of patients with memory impairments. Finally, an alternative approach that focuses on specific encoding and cueing processes rather than overall recall, is proposed. Empirical studies are reported in the following four chapters; Chapter 2 described the development and implementation of the initial studies with non-clinical participants, while Chapter 3 described the refinement of the procedure. Semantic and phonological coding and cueing effects were examined within a temporal context. The findings clearly demonstrated that semantic and phonological processes acted in quite distinct and consistent ways, irrespective of the degree of prior exposure to the word stimuli. Participants had more difficulty keeping track of the current episode when the codes and cues were phonological. Following refinements to the stimuli and procedure, replication of the previous results confirmed that cued-retrieval using semantic, phonological and temporal cues was a legitimate approach to memory assessment. Furthermore, differential effects were still found even after significantly shortening the procedure, thus adding support for its potential usefulness in clinical settings. The next two chapters examined the clinical application of this alternative approach in brain injury rehabilitation patients. The ability of non-clinical participants and brain injury rehabilitation patients to effectively use codes and cues to retrieve verbal information was compared. Quite distinct profiles indicated that the cued-retrieval procedure clearly differentiated clinical and non-clinical participants. Brain injury patients obviously performed more poorly on both semantic and phonological tasks but they also demonstrated a reduced ability to effectively use phonological codes and cues in retrieving information, as their performance declined across successive episodes due to increasing interference effects. Finally, the cued-retrieval procedure was administered to brain injury patients in the context of routine, post-injury, comprehensive neuropsychological assessment to examine the comparative value of this process approach in relation to other currently used clinical tests of verbal and working memory. The results clearly indicated that cued-retrieval was more sensitive in detecting memory impairments than most of the other tests in current use and, due to its brevity, was able to provide this information relatively quickly. The closing chapter summarized and integrated the empirical findings of these research studies and highlighted the implications for future clinical practice and further research. The preliminary evidence clearly indicated that this brief and sensitive method of examining the underlying processes of verbal working memory has clinical and theoretical potential.

In New Zealand, approximately 10% of births are considered premature, that is less than 37 weeks gestation. With advances in medical technology, young infants are surviving gestation periods as few as 23 weeks. It is expected that many of these severely premature infants will demonstrate some problem in their academic, or cognitive function including language functioning. It is agreed that children who are born severely premature often present with language problems, the nature of the difficulties are not clear. Research examining language abilities that involve cognitive functions such as inference generation have demonstrated that children born prematurely exhibit difficulties with phonologic short-term memory and executive function. Language tasks such as inference understanding require children to integrate real-world knowledge with the linguistic information to generate and produce language that is more complex. The aim of this study was to discover if the language of children born severely premature differs from that of language-age matched peers. This study examined high-level language abilities of school-age children born severely prematurely, specifically, language tasks that involved executive functions including working memory, story inferencing, and recognising absurdities. Six children who were born less than 28 weeks gestation participated in this study. Their results on the above measures were compared to a language-aged matched comparison group, determined by performance on a standardised test. It was hypothesised that the children born severely premature would not differ from their language-age matched peers on measures of general language ability but differences would exist on measures of language processing and inferencing. The findings overall showed little difference between the preterm group and their language-age matched peers on measures except for the measure of chronological age. Although no group difference was found for the measure of working memory, a larger variance on this measure was observed in the preterm group.

This thesis explored two main research questions, firstly investigating whether musical expertise offers a performance advantage in working memory for sequences of tones that vary in pitch, and secondly whether any advantage of musical expertise may be present in older as well as younger individuals. Previous research on working memory for tone sequences has reported that articulatory suppression interferes with temporary storage of verbal but not with tone sequences (Koelsch et al, 2009), suggesting a “tonal loop” within a musician’s working memory (Schulze, Zysset, Mueller, Friederici, & Koelsch, 2011) that complements the phonological loop for verbal material in musicians and non-musicians alike (e.g., Baddeley, 1986; 1992). The five experiments reported here explored detection of a pitch change between pairs of tone sequences with or without concurrent articulatory suppression or singing suppression. In Experiment 1, with pairs of auditory tonal (in a musical key) sequences to be compared, singing suppression impaired non-musicians significantly more than musicians, although both groups showed an impairment, whereas only non-musicians were affected by verbal articulatory suppression. In Experiment 2, conducted only with musicians who could read music, the first sequence of each pair was presented visually and the second sequence for comparison was presented aurally. Musicians were again impaired by singing suppression but not by articulatory suppression. For Experiment 3, for auditory atonal (no musical key) pairs of sequences, non-musicians performed at floor, and musicians were again significantly more impaired by singing suppression than by articulatory suppression. In contrast, for Experiment 4, only with musicians who could read music, for visually presented atonal sequences each followed by an auditory atonal sequence for comparison, musicians were significantly more impaired by articulatory suppression than singing suppression. These results suggest that for tonal sequences, musicians use their musical training and experience, coupled with subvocalised singing, but for atonal sequences, additional strategies involving phonological rehearsal may be used. Non-musicians may also rely on musical experience and subvocal singing for tonal sequences but seem to be unable to do so for atonal sequences. Results are consistent with the use by both musicians and non-musicians of a tonal loop for the rehearsal of tone sequences, which develops with musical training and may be used in addition to subvocal rehearsal. Previous research has suggested musical expertise may offer some protection against cognitive ageing (Hanna-Pladdy & MacKay, 2011; Amer, Kalender, Hasher, Trehub, & Wong, 2013). Experiment 5 in this thesis explored whether a lifetime of musical training and experience may offer the same advantages in old age for retaining tone sequences that had been found in Experiments 1 and 3 for younger musicians. This experiment also considered whether any advantage for older musicians on this task could be explained by the proposed “bilingual advantage” (e.g., Bialystok, Craik, Klein & Viswanathan, 2004), and what other aspects of cognition might be associated with tone sequence memory performance. A test battery was utilised with 74 older adults (60-80 years of age) to assess the influence of musical and language expertise, and cognitive abilities (attention, working memory capacity, self-reported prospective and retrospective memory) on the music-related pitch sequence comparison task from Experiments 1 and 3. Working memory capacity was found to predict individual differences in the ability to detect pitch changes between pairs of tone sequences, regardless of musical experience. Older musicians performed more poorly on the pitch change detection task overall than the younger musicians in the earlier experiments, but their performance on the task was significantly better than for age-matched non-musically trained peers who were close to floor for both tonal and atonal sequences, suggesting some benefit from a lifetime of musical experience.

abstract: Working memory (WM) and attention deficits have been well documented in individuals with aphasia (IWA) (e.g. Caspari et al., 1998; Erickson et al., 1996; Tseng et al., 1993; Wright et al., 2003). Research into these cognitive domains has spurred a theoretical shift in how aphasia is conceptualized - from a purely linguistic disorder to a cognitive-information processing account. Language deficits experienced by IWA may result from WM impairments or from an inability to allocate cognitive effort to the tasks. However, how language impacts performance on these tasks has not been readily investigated. Further, there is a need for a more direct measure of effort invested to language tasks. Heart rate variability (HRV) is a physiological measure of cognitive workload that has been used to measure effort in neurologically intact participants. Objectives of the study included: (1) determining the feasibility of using HRV as a measure of effort IWA invest into verbal compared with spatial WM tasks, (2) Comparing participants' performance on verbal and spatial WM tasks; and (3) determining the relationship among performance, perceived task difficulty, and HRV across verbal and spatial tasks. Eleven IWA and 21 age- and education-matched controls completed verbal and spatial n-back tasks at three difficulty levels. Difficulty ratings were obtained before and after each task. Results indicated spatial WM was relatively preserved compared with verbal WM for the aphasia group. Additionally, the aphasia group was better at rating task difficulty after completing the tasks than they were at estimating task difficulty prior to completing the tasks. Significant baseline-task differences in HRV were found for both groups. Relationships between HRV and performance, and HRV and task difficulty were non-significant. Results suggest WM performance deficits in aphasia may be primarily driven by their language deficit. Baseline-task differences in HRV indicate effort is being allocated to the tasks. Difficulty ratings indicate IWA may underestimate task demands for both verbal and spatial stimuli. However, the extent to which difficulty ratings reflect effort allocated remains unclear. Additional research is necessary to further quantify the amount of effort IWA allocate to verbal and non-verbal tasks. / Dissertation/Thesis / Ph.D. Speech and Hearing Science 2012

xiii, 155 p. : ill. A print copy of this thesis is available through the UO Libraries. Search the library catalog for the location and call number. / In the multiple object tracking (MOT) task, observers are presented with multiple identical objects, some of which are temporarily identified as targets. After a selection period, all objects move randomly and independently for several seconds. At the end the motion period, all objects stop and observers must identify the target objects again. This task has been used to study a variety of important cognitive questions from object-based attention to cognitive development, divided attention and the development of expertise. Yet, surprisingly little is known about the neural mechanisms that underlie the ability to track multiple targets independently. Although a number of researchers have used fMRI (functional magnetic imaging) to examine what areas are active during MOT, the current set of studies is the first to employ ERPs (event-related potentials) to examine the neural mechanisms of MOT. With excellent temporal resolution, the ERP methodology allows researchers to delineate the time course of different phases of a single task with millisecond precision, something not possible with fMRI. In Chapter II, we manipulated the number of targets and difficulty of tracking and observed a lateralized contralateral negativity that was sensitive to the number of targets but not difficulty of tracking. Chapter III examined the effect of irrelevant white probes flashed briefly throughout the trial while observers tracked. We observed modulations of early visual components that indicated that during tracking, spatial attention focused on targets but did not differentiate between distractors and empty space. Finally, in Chapter IV, we examined the relationship between visual working memory (VWM) and MOT by manipulating the presence or absence of task relevant motion. We found that the waveforms evoked by an MOT task in the absence of task-relevant motion were nearly identical to waveforms evoked by the VWM task, suggesting that VWM is an important part of the typical MOT task. This thesis includes previously published and unpublished material. / Committee in Charge: Edward Vogel, Chairperson, Psychology; Edward Awh, Member, Psychology; Ulrich Mayr, Member, Psychology; Paul van Donkelaar, Outside Member, Human Physiology