Abstract This study examined memory functioning from a female perspective, with the aim of determining factors that might impact performance and render the accuracy of memory measurement, particularly with advancing age, problematic. Factors investigated, among others, were the role of attention and/or engagement with the memory tasks administered, state affect (i.e., positive and negative arousal) at time of testing, subjective memory appraisal, particularly in the domain of perceived memory self-efficacy (MSE), and the use of hormone replacement therapy (HRT) by older post-menopausal women. Two experimental computer-based tests of episodic memory, Paired Associates (PA) and Serial Recall (SR), were administered to 181 female participants aged 18 to 86 years. The tasks were designed to emphasise components that make episodic memory especially difficult, and minimise the use of strategies that might assist recall. Thus, they varied the requirement for recall as opposed to recognition, the need to form an association between a pair of unrelated words, and the need to discriminate the most recent list from earlier list(s). Other measures used included a demographic survey administered to participants individually in an interview format, and a number of variables examined in this study derived from responses to items contained in this survey. The research battery also included psychometric measures of transient affective states, psychological well-being, alertness, in addition to measures of global cognitive status and metamemory (i.e., subjective memory appraisal). The overall aim was to examine a range of factors that might influence episodic memory performance in cognitively intact healthy women, and thus render the interpretation of age-related changes to memory functioning problematic. For analyses participants were assigned to three groups - young, middle-aged and older. There were 60 young adults aged 18 to 29 years, 60 middle-aged adults aged 49 to 60 years, and 61 older adults aged 61 to 86 years. Each participant was tested individually in a single session lasting approximately 3 ½ hours, with younger participants requiring less time to complete assessments. Order of test administration and instructions were standardised across the entire sample. Inferential statistics included correlation, t-test statistic, and analysis of variance (ANOVA) with Tukey post-hoc comparisons. Hierarchical multiple regression analyses were conducted to determine key correlates of memory performance outcomes. No significant differences between the cohorts were found in mean years of education. However, episodic memory recall differed significantly by age group. As expected, young adults recalled significantly more words in the memory tasks than their older counterparts, and middle-aged adults outperformed adults in the oldest cohort. Moreover, older adults’ performance deficits were more pronounced in the tasks requiring that they make an association between a pair of unrelated words. Across all cases, transient mood states were significantly related to memory scores; however, individuals in the oldest cohort were particularly vulnerable to mood fluctuations. This cohort experienced a significantly greater decline in positive affect and a significant greater increase in negative affect while undergoing memory testing, highlighting their greater vulnerability to stressors inherent in a memory testing situation. Although scores on the measure of attention were near ceiling, indicative of participants’ level of effort, motivation, and engagement with the memory tasks, the measure of attention discriminated between older and younger adults’ results, and was a key predictor of memory performance. Noteworthy is that attention scores significantly contributed to performance variability in younger and older adults but not in middle-aged adults. Across all cases, age, education, and attention were the key contributing factors to variability in memory scores. Although four lifestyle factors: (1) subjective sleep appraisal, (2) body mass index (BMI), (3) physical activity, and (4) caffeine intake were significantly associated with performance in the memory tasks, once the effects of these key variables were removed, lifestyle factor did not uniquely contribute to performance variability. Moreover, no association was found between hormone replacement therapy (HRT) and episodic memory performance across the broader sample. However, in a small subgroup of older women (n = 15, M age = 66 years), long-term users of this treatment, HRT had a significant effect on memory performance and was indicative of better recall on the memory tasks. The effect of subjective memory appraisal, MSE included, on objective performance outcomes was examined. The results showed that memory self-evaluations were not a significant contributing factor to episodic memory performance, confirming that memory self-appraisal is a poor predictor of actual memory performance, and thus does not pose a challenge to the measurement of age-related changes to memory abilities. Although there were commonalities, factors influencing memory performance differed by age cohort. For example, in young adults, positive mood, a perception of sleeping well, subjective health, and attention were significantly related to performance on the memory tasks. However, once the effect of attention was removed in the regression analysis, no other variable was predictive of episodic memory functioning in this cohort. In contrast, the single significant predictor of memory performance in middle-aged individuals was education, and neither attention, nor positive mood, or physical activity had a significant effect on this cohort’s performance. Similarly, having more years of formal education benefited older adults’ episodic memory functioning. However, high scores on global cognitive functioning and on the tasks measuring attention were equally important to episodic memory recall in this age group. In sum, the significant contribution of age to memory variability attested to the utility of the memory measures in detecting age-related changes to episodic memory functioning, which were independent of deficits in attention or level of education. Moreover, the effect of several factors (e.g., transient mood, lifestyle) on memory scores was explained by an effect on attention, and this has clear implication for the proper evaluation of long-term changes to memory functioning. Limitations of the study and suggestion for future research are discussed.

Abstract This study examined memory functioning from a female perspective, with the aim of determining factors that might impact performance and render the accuracy of memory measurement, particularly with advancing age, problematic. Factors investigated, among others, were the role of attention and/or engagement with the memory tasks administered, state affect (i.e., positive and negative arousal) at time of testing, subjective memory appraisal, particularly in the domain of perceived memory self-efficacy (MSE), and the use of hormone replacement therapy (HRT) by older post-menopausal women. Two experimental computer-based tests of episodic memory, Paired Associates (PA) and Serial Recall (SR), were administered to 181 female participants aged 18 to 86 years. The tasks were designed to emphasise components that make episodic memory especially difficult, and minimise the use of strategies that might assist recall. Thus, they varied the requirement for recall as opposed to recognition, the need to form an association between a pair of unrelated words, and the need to discriminate the most recent list from earlier list(s). Other measures used included a demographic survey administered to participants individually in an interview format, and a number of variables examined in this study derived from responses to items contained in this survey. The research battery also included psychometric measures of transient affective states, psychological well-being, alertness, in addition to measures of global cognitive status and metamemory (i.e., subjective memory appraisal). The overall aim was to examine a range of factors that might influence episodic memory performance in cognitively intact healthy women, and thus render the interpretation of age-related changes to memory functioning problematic. For analyses participants were assigned to three groups - young, middle-aged and older. There were 60 young adults aged 18 to 29 years, 60 middle-aged adults aged 49 to 60 years, and 61 older adults aged 61 to 86 years. Each participant was tested individually in a single session lasting approximately 3 ½ hours, with younger participants requiring less time to complete assessments. Order of test administration and instructions were standardised across the entire sample. Inferential statistics included correlation, t-test statistic, and analysis of variance (ANOVA) with Tukey post-hoc comparisons. Hierarchical multiple regression analyses were conducted to determine key correlates of memory performance outcomes. No significant differences between the cohorts were found in mean years of education. However, episodic memory recall differed significantly by age group. As expected, young adults recalled significantly more words in the memory tasks than their older counterparts, and middle-aged adults outperformed adults in the oldest cohort. Moreover, older adults’ performance deficits were more pronounced in the tasks requiring that they make an association between a pair of unrelated words. Across all cases, transient mood states were significantly related to memory scores; however, individuals in the oldest cohort were particularly vulnerable to mood fluctuations. This cohort experienced a significantly greater decline in positive affect and a significant greater increase in negative affect while undergoing memory testing, highlighting their greater vulnerability to stressors inherent in a memory testing situation. Although scores on the measure of attention were near ceiling, indicative of participants’ level of effort, motivation, and engagement with the memory tasks, the measure of attention discriminated between older and younger adults’ results, and was a key predictor of memory performance. Noteworthy is that attention scores significantly contributed to performance variability in younger and older adults but not in middle-aged adults. Across all cases, age, education, and attention were the key contributing factors to variability in memory scores. Although four lifestyle factors: (1) subjective sleep appraisal, (2) body mass index (BMI), (3) physical activity, and (4) caffeine intake were significantly associated with performance in the memory tasks, once the effects of these key variables were removed, lifestyle factor did not uniquely contribute to performance variability. Moreover, no association was found between hormone replacement therapy (HRT) and episodic memory performance across the broader sample. However, in a small subgroup of older women (n = 15, M age = 66 years), long-term users of this treatment, HRT had a significant effect on memory performance and was indicative of better recall on the memory tasks. The effect of subjective memory appraisal, MSE included, on objective performance outcomes was examined. The results showed that memory self-evaluations were not a significant contributing factor to episodic memory performance, confirming that memory self-appraisal is a poor predictor of actual memory performance, and thus does not pose a challenge to the measurement of age-related changes to memory abilities. Although there were commonalities, factors influencing memory performance differed by age cohort. For example, in young adults, positive mood, a perception of sleeping well, subjective health, and attention were significantly related to performance on the memory tasks. However, once the effect of attention was removed in the regression analysis, no other variable was predictive of episodic memory functioning in this cohort. In contrast, the single significant predictor of memory performance in middle-aged individuals was education, and neither attention, nor positive mood, or physical activity had a significant effect on this cohort’s performance. Similarly, having more years of formal education benefited older adults’ episodic memory functioning. However, high scores on global cognitive functioning and on the tasks measuring attention were equally important to episodic memory recall in this age group. In sum, the significant contribution of age to memory variability attested to the utility of the memory measures in detecting age-related changes to episodic memory functioning, which were independent of deficits in attention or level of education. Moreover, the effect of several factors (e.g., transient mood, lifestyle) on memory scores was explained by an effect on attention, and this has clear implication for the proper evaluation of long-term changes to memory functioning. Limitations of the study and suggestion for future research are discussed.

Ce travail a pour but d'identifier les déterminants du Sentiment d'Auto-efficacité Mnésique (SAM). Dans le chapitre 1, nous définissons le SAM et décrivons ses déterminants proposés dans la littérature. Les déterminants que nous proposons s'organisent en deux axes. Il a été montré que le SAM décline avec l'âge. Une explication proposée à ce déclin est les croyances culturellement partagées à propos du déclin de la mémoire avec l'âge qui amèneraient les personnes âgées à évaluer qu'elles ont une mauvaise mémoire, indépendamment de leur performances mnésiques réelles. Cependant, les mécanismes explicatifs de cet effet ont été peu étudiés. Nous supposons que l'attribution causale de ses propres manquements normaux de mémoire à l'âge devrait expliquer le déclin du SAM avec l'âge. Dans le chapitre 2, nous montrons que cette attribution à l'âge médiatise partiellement le lien entre âge et SAM. Dans le chapitre 3, nous présentons une étude longitudinale menée afin a) d'examiner si l'attribution à l'âge et l'identification au groupe des personnes âgées diminuent le SAM ainsi que le changement de SAM avec le temps et b) d'identifier les déterminants de l'attribution à l'âge, en faisant l'hypothèse que les croyances générales d'un déclin de la mémoire avec l'âge et l'identification au groupe des personnes âgées influencent l'attribution à l'âge ainsi que le changement d'attribution à l'âge à travers le temps. Les analyses transversales appuient partiellement nos hypothèses mais nous échouons à observer ces effets avec une méthodologie longitudinale. Le SAM est à la fois spécifique à un domaine particulier (e.g., mémoire des chiffres, mémoire des visages) et généralisable à plusieurs domaines. La littérature sur le SAM nous en apprend peu sur les conditions permettant la généralisation d'un SAM spécifique à un domaine à un autre domaine. Nous nous attendons à ce que la généralisation du SAM dans un domaine à un autre domaine dépende d'une théorie naïve particulière, la conception d'unité de la mémoire (conception unitaire vs. multiple), avec l'hypothèse que la généralisation du SAM apparaisse d'autant plus lorsque les participants ont une conception unitaire de la mémoire plutôt que multiple. Dans le chapitre 4, nous présentons quatre études testant cette hypothèse. Globalement, nous échouons à observer les effets attendus, probablement à cause des domaines choisis. A notre connaissance, aucune étude ne s'est intéressée à identifier les domaines de mémoire utilisés pour catégoriser spontanément les situations de mémoire. Dans le chapitre 5, nous présentons deux études visant à identifier ces domaines. Dans le chapitre 6, nous discutons des résultats au regard de la littérature sur le SAM. / This work aims at identifying the determinants of Memory Self-Efficacy (MSE). In chapter 1, we define MSE and describe determinants of MSE proposed in the literature. The determinants we propose are organized according to two axes. MSE has been shown to decline with aging. One suggested explanation for this decline is that culturally shared beliefs about memory decline lead older adults to evaluate their memory as being poor, regardless of their actual memory performance. However, explanatory mechanisms of this effect have rarely been studied. We assume that older adults' tendency to attribute their everyday memory failures to their age should explain age-related decline in MSE. In chapter 2, we show that the tendency to attribute one's memory failures to one's age partially mediates the link between age and MSE. In chapter 3, we present a longitudinal study conducted in order to a) examine whether this age attribution and old age group identification decrease MSE and change over time in MSE, and b) identify age attribution determinants, with the hypothesis that general beliefs about memory decline and identification with the older adults influence age attribution, as well as change over time in age attribution. Transversal analyses partially support our hypotheses but we fail to observe the same effects using a longitudinal method. MSE is both specific to some domains (e.g., memory for numbers, memory for faces) and general across several domains. Literature about MSE does not tell much about the conditions allowing the generalization of MSE from one domain to another domain. We assume that the generalization of MSE from one domain to another depends on a particular lay theory, the unity conception of memory (unitary vs. multiple conception), with the hypothesis that MSE generalization would appear more when participants conceive memory as unitary rather than multiple. In chapter 4, we present four studies testing this hypothesis. Globally, we fail to observe the expected effects, likely due to the chosen domains. To our knowledge, no study has been conducted to identify memory domains used to spontaneously categorize memory situations. In chapter 5, we present two studies designed to identify these domains. In chapter 6, we discuss our results regarding MSE literature.

No description available.