Abstract
Objectives
Learning and memory abilities decline with age; however collaborative learning paradigms with a familiar partner improve older adults’ performance and reduce these age-related differences. The current study compared participants learning with familiar and unfamiliar partners, and with perceived human and computer partners in a Wizard-of-Oz spoken dialogue system to determine whether collaboration alone is sufficient to improve older adults’ performance.
Participants and methods
Study 1 involved older and younger (n=48) participant pairs arranging abstract tangram shapes in specific orders on a grid in familiar and unfamiliar pairs. Study 2 involved older adults (n=24) performing the same card-matching task with participants believing they were interacting with a human partner, or a computer partner.
Results
In Study 1, older adults initially performed more poorly than younger adults, taking more time and using more words to correctly arrange the tangrams. A learning effect was observed in both groups and by the final trials, older and younger groups’ performances did not significantly differ. In Study 2, initially participants were quicker at completing the task with the computer partner, but by the final trials were significantly faster with a perceived human partner. Participants also changed their answers significantly more with a computer partner.
Conclusions
Within this collaborative learning paradigm, older adults achieve the same level of performance as younger adults over multiple trials. Collaborating with a familiar partner does not improve performance compared with an unfamiliar partner but learning with a perceived human partner was more effective and efficient than with a computer partner.
Learning and memory abilities decline with age; however collaborative learning paradigms with a familiar partner improve older adults’ performance and reduce these age-related differences. The current study compared participants learning with familiar and unfamiliar partners, and with perceived human and computer partners in a Wizard-of-Oz spoken dialogue system to determine whether collaboration alone is sufficient to improve older adults’ performance.
Participants and methods
Study 1 involved older and younger (n=48) participant pairs arranging abstract tangram shapes in specific orders on a grid in familiar and unfamiliar pairs. Study 2 involved older adults (n=24) performing the same card-matching task with participants believing they were interacting with a human partner, or a computer partner.
Results
In Study 1, older adults initially performed more poorly than younger adults, taking more time and using more words to correctly arrange the tangrams. A learning effect was observed in both groups and by the final trials, older and younger groups’ performances did not significantly differ. In Study 2, initially participants were quicker at completing the task with the computer partner, but by the final trials were significantly faster with a perceived human partner. Participants also changed their answers significantly more with a computer partner.
Conclusions
Within this collaborative learning paradigm, older adults achieve the same level of performance as younger adults over multiple trials. Collaborating with a familiar partner does not improve performance compared with an unfamiliar partner but learning with a perceived human partner was more effective and efficient than with a computer partner.
Original language | English |
---|---|
Publication status | Unpublished - 6 Jul 2016 |
Event | International Neuropsychological Society Mid-Year Meeting - London, United Kingdom Duration: 6 Jul 2016 → 8 Jul 2016 |
Conference
Conference | International Neuropsychological Society Mid-Year Meeting |
---|---|
Country/Territory | United Kingdom |
City | London |
Period | 6/07/16 → 8/07/16 |