Project

Back to overview

Laying the groundwork for understanding age-related changes in individual semantic networks and their role in cognitive aging

English title Laying the groundwork for understanding age-related changes in individual semantic networks and their role in cognitive aging
Applicant Wulff Dirk
Number 197315
Funding scheme Project funding (Div. I-III)
Research institution Cognitive and Decision Sciences Fakultät für Psychologie Universität Basel
Institution of higher education University of Basel - BS
Main discipline Psychology
Start/End 01.03.2021 - 29.02.2024
Approved amount 396'972.00
Show all

All Disciplines (2)

Discipline
Psychology
Applied psychology

Keywords (4)

Computational semantics; Cognitive aging; Semantic memory; Network Science

Lay Summary (German)

Lead
Älter werden geht mit Veränderungen in der kognitiven Leistungsfähigkeit einher. Eine Rolle spielt dabei möglicherweise die Entwicklung des Gedächtnisses. Die Grösse und Struktur des Gedächtnisses von Fakten und Bedeutungen unterscheidet sich zwischen Menschen verschiedenen Alters. Das Projekt versucht diese Unterschiede zu charakterisieren und zu beleuchten welchen Beitrag sie zur kognitiven Entwicklung über die Lebensspanne leisten.
Lay summary

Inhalt und Ziele

Inhalt und Ziele
Das Wissen über die Welt und die Bedeutung von Begriffen entwickelt sich über die gesamte Lebensspanne. Das Projekt hat zum Ziel, die sich verändernde Repräsentation dieses Wissens im menschlichen Gedächtnis näher zu erforschen. Dafür werden bei Menschen unterschiedlichen Alters grosse semantische Netzwerke gemessen. Zusätzlich werden verschiedene kognitive Leistungsmessungen bei den gleichen Personen vorgenommen. Die Netzwerke werden auf ihre Struktur untersucht und mit den Leistungsmessungen verglichen. Dabei sollen aktuelle Erklärungsansätze für die sich über das Alter verändernde Leistungsfähigkeit in Anbetracht der neu auf individueller Ebene erstellten Netzwerke evaluiert werden. Zudem werden für das Forschungsprojekt benötigte Methoden und Computerprogramme der breiteren wissenschaftlichen Gemeinschaft zugänglich gemacht.

Wissenschaftlicher und Gesellschaftlicher Kontext

Wissenschaftlicher und Gesellschaftlicher Kontext
Während unsere Gesellschaft immer älter wird, gewinnt das Verständnis der kognitiven Struktur und Leistungsfähigkeit über die gesamte Lebensspanne an Bedeutung. Bisher wenig betrachtete individuelle Unterschiede in der Gedächtnisstruktur werden in diesem Projekt bei Menschen in verschiedenen Lebensabschnitten erforscht. Die Unterschiede werden als mögliche Ursache der mit dem Alter abnehmenden kognitiven Leistungsfähigkeit evaluiert. Die Ergebnisse des Projekts versprechen Einsichten in den gesunden Verlauf des kognitiven Alterns, aus denen sich Implikationen für die Diagnostik altersbedingter Erkrankungen wie Demenz oder Alzheimer ableiten lassen. Des Weitern leistet das Projekt einen wichtigen Beitrag zum methodischen Werkzeugkasten der Gedächtnisforschung, der kognitiven Psychologie und der Linguistik.

Direct link to Lay Summary Last update: 04.03.2021

Responsible applicant and co-applicants

Employees

Name Institute

Project partner

Abstract

Aging seems to have quite a negative effect on cognitive performance. Compared to younger adults, older adults remember fewer words from word lists, show slower perceptual speed, and perform worse in reasoning tasks, to name only three aspect of what is commonly referred to as cognitive decline. Research often attributes this decline to cognitive process deterioration (e.g., Healey & Kahana, 2017; Hills, Mata, Wilke, & Samanez-Larkin, 2013; Salthouse, 1996, 2010), but there is also a more benign perspective on cognitive aging. In comparison with younger adults, older adults continuously learn from the environment and thereby acquire larger and differently structured semantic representations (Dubossarsky, De Deyne, & Hills, 2017; Verhaeghen, 2003; Wulff, Hills, Lachman, Mata, 2016; Wulff, Hills, & Mata, 2018). A controversial thesis suggests that these differences may be responsible for producing the decline observed in older adults’ cognitive performance (Ramscar, Hendrix, Shaoul, & Baayen, 2014, Ramscar, Sun, Hendrix, & Baayen, 2017). Mounting evidence suggests a considerable relationship between the size and structure of semantic representations and cognitive performance (e.g., Griffiths, Steyvers & Firl, 2007; Ramscar, Sun, Hendrix, & Baayen, 2017; Steyvers & Tenenbaum, 2004; see Wulff et al., 2019, for a review). How and whether age-related changes in semantic representations actually contribute to cognitive aging, however, is still unclear due to a lack of dedicated investigations.I propose to fill this gap in the following way. First, I plan-on the individual and trial level-to collect large-scale semantic network data from younger and older adults and use them to predict performance in cognitive tasks that can help distinguish healthy and pathological aging. Second, I will determine which model linking semantic networks and cognitive performance can provide the best account for the collected data. Third, I will conduct an integrative review of the literature highlighting and synthetizing the breadth of representational accounts of cognitive performance across psychology, linguistics, and neuroscience. Finally, I will add computational methods used to analyze semantic networks and their link to cognitive performance to my memnet R package and make it available to the research community, along with a comprehensive tutorial. Overall, my project aims to fill a gap in the current empirical and theoretical knowledge concerning the life-span development of semantic representations and their role in healthy and pathological aging. I propose to fill this gap in two ways. First, I plan to measure large-scale semantic networks of younger and older adults and use those to predict performance in a variety of cognitive tasks relevant to healthy and pathological aging on the individual and trial level. Second, I plan to use models of computational semantics, such as BEAGLE (Jones & Mewhort, 2007) and naïve discriminant learning (Baayen, & Ramscar, 2015), and an age-stratified linguistic corpus, created for this purpose, to simulate healthy representational aging. The goal is to explore how much of cognitive aging can be explained by learning processes alone and which other processes need be assumed to reproduce the observed representations and behavior. To support the proposed and future research into memory representations across the life span, I plan to make three additional contributions. First, I will make available all of the methods used in the empirical and simulation analyses by expanding my memnet R package. Second, I will create a multilingual, general-purpose word list to facilitate cross-study and cross-language comparisons of behavior. Finally, I will create a web platform designed to communicate the results of my studies to researchers and the public and to initiate a citizen-science project that will allow us to study memory representational across longer time scales and to link them to pathological developments. Overall, my project aims to fill a gap in the current empirical and theoretical knowledge concerning the life-span development of semantic representations and their role in healthy and pathological aging.
-