Memory for Affective Prosody

Vivien Ormiston-mason (2009). Memory for Affective Prosody PhD Thesis, School of Psychology, The University of Queensland.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
s33221540_PhD_abstract.pdf s33221540_PhD_abstract.pdf application/pdf 13.02KB 0
s33221540_PhD_totalthesis.pdf Final Thesis Lodgement application/pdf 1.40MB 4
Author Vivien Ormiston-mason
Thesis Title Memory for Affective Prosody
School, Centre or Institute School of Psychology
Institution The University of Queensland
Publication date 2009-09
Thesis type PhD Thesis
Supervisor Associate Professor Ken McFarland
Professor Michael Humphreys
Total pages 184
Total black and white pages 184
Subjects 17 Psychology and Cognitive Sciences
Abstract/Summary This thesis presents a new tool for examining affective prosodic memory. The goals of the project were to demonstrate affective prosodic memory, and compare it with phonological memory using a paradigm specifically adapted for this purpose. Affective prosodic memory is a new area of study. The topic has a range of potential applications, including the eventual development of a brief, computerised test of affective prosodic memory for use in brain-injured populations. The tool beginning its development here would have the potential to assist not only the brain injured, but also their families, friends and caregivers, in understanding the communication and social difficulties experienced, and in guiding education and management programmes. A modified version of the Brown-Peterson task of working memory (Brown, 1958; Peterson & Peterson, 1959) was designed, incorporating new features to control for the effects of familiarity (adapted from Parkin, Leng & Hunkin, 1990). The Brown-Peterson task has been shown to be useful for examining and comparing a wide range of stimuli in experimental research, as well as in examining working memory in a large range of patient groups (with medical or neurological conditions). The proactive interference design offered the benefits of a second measure of performance in addition to overall accuracy. The basic experimental task was a computerised cued recall task. For each of 16 trials, participants heard four words, spoken in different emotional tones, and then read aloud four digits, in a distractor paradigm. They then saw a face which acted as a memory cue: In the prosodic task, the face showed an emotional expression matching one of the four words; while in the phonological task, the face showed a mouth position representing the ending-sound of one of the four words. In Experiment 1, participants were asked to state the word that matched (i.e., fit) the face. In Experiment 2, the instructions were to state the position of the word that matched the face. In Experiment 3, an affective distractor task was used to examine selective interference effects. Literature collected from neighbouring fields converged to suggest the prediction that affective and phonological memory would differ (Buchanan & Adolphs, 2002, 2004; Davidson & Irwin, 1999; Liberman & Mattingly, 1985; 1989; LeDoux, 2000; Levitin, 1999; Owren, Rendall & Bachorowski, 2005; Patel & Peretz, 1997; Peretz et al., 1994; Scott et al., 1997; Semal et al., 1996; Wurm et al., 2001; Zatorre, 2003; Zatorre et al., 1992). On the basis of existing studies examining memory for music and lyrics (Crowder, Serafine & Repp, 1990; Morrongiello & Roes, 1990; Samson & Zatorre, 1991; Serafine, Crowder & Repp, 1984; Serafine et al., 1986), it was hypothesised that phonological memory would exceed affective prosodic memory. Consistent with this, Experiment 1 found that in the current paradigm, affective prosodic memory was consistently recalled at greater than chance levels, and that it was considerably weaker than phonological memory in terms of overall accuracy. Some preliminary evidence was obtained that affective memory was more susceptible to proactive interference than its counterpart, phonological memory, with performance declining earlier and more sharply in the affective condition. Experiment 2 showed that changing the response requirement of the task (from a word to an item position) did not alter the main findings, which were replicated: Affective prosodic memory remained weaker than its counterpart in overall accuracy, and also suffered more from the effects of proactive interference. The results from Experiment 2 were consistent with both phonological memory and auditory affective memory being naturally processed in an ordered fashion. Experiment 3 showed that affective interference (in the form of a distractor task) had a detrimental effect on performance in the emotion condition in particular. This supports the case that what was being measured was indeed a form of affective memory, adding to construct validity. Validity and reliability were further explored by administering three independent measures thought to have elements in common with the primary experimental task, Emotion Evaluation Test of the TASIT, the Affective Auditory Verbal Learning Test, and a Spoonerisms task. Relationships were noted between the experimental task and related measures in most cases, increasing confidence that the task was tapping what it was designed to measure. The obtained findings provide evidence that affective prosodic memory exists. They are also consistent with the view that affective prosodic memory and phonological memory differ, and that affective prosodic codes are less transient than phonological codes, although future work would be required to corroborate these findings. Some limited converging support is available at the present time from a small group of studies that have implied the existence of affective prosodic memory by studying it indirectly (it was not required for successful task completion). The current methodological paradigm appeared to perform well and successfully detect overall accuracy differences as well as interference differences in the variables measured. The findings begin to pave the way for a standard procedure to examine prosodic memory. The general task format also shows promise for more general use in application to a wide variety of different research questions and classes of material. The current project has revealed some promising findings, both methodological and in terms of content matter, that can be used to guide future research. In the series of experiments presented here, affective memory was more susceptible to proactive interference than phonological memory. Performance declined more rapidly in the affective memory condition, consistent with the interpretation that proactive interference built up more quickly in the affective domain. This may reflect the complexity of the bindings necessary in memory, particularly where working memory tasks are concerned. This fits with what is known of the neural basis of emotion, as well as the functions of emotion in an evolutionary sense. The current findings appear to suggest a boundary condition for the emotional enhancement effect on memory. In this sense they join the work of Mather et al. (2006), Mitchell et al. (2006) and Perlstein, Elbert and Stenger (2002), who obtained a similar result. Future work should further refine the task developed here, pursue reliability testing, normative data gathering, and testing in clinical samples, toward creating a brief, computerised test of affective prosodic working memory. There are implications flowing from the current work in the areas of understanding memory, advancing psychometrics, and in terms of clinical neuropsychology. The project has added to understanding of the components of memory by identifying a new form of memory, affective prosodic memory. It has developed an effective assessment method with evidence suggestive of a sound validity and reliability base. It has also supplied both a framework and assessment method for clinical neuropsychologists to use with patients, one that is easy to administer.
Keyword Affective prosodic memory
working memory
neuropsychological assessment
emotional memory
Brown-Peterson task

Citation counts: Google Scholar Search Google Scholar
Created: Mon, 22 Nov 2010, 22:50:10 EST by Ms Vivien Ormiston-mason on behalf of Library - Information Access Service