Exploring the Behaviour, Reliability, and Usability of a new Smartphone Version of Paired Associates Learning
Alexander J Kaula, Nick Taptiklis, Francesca K Cormack- Psychiatry and Mental health
- Cellular and Molecular Neuroscience
- Geriatrics and Gerontology
- Neurology (clinical)
- Developmental Neuroscience
- Health Policy
- Epidemiology
Abstract
Background
PAL is a well‐known and trusted task with good sensitivity to deficiencies in memory capabilities. There has been considerable interest in a smartphone version of PAL, and Cambridge Cognition has now completed work on smartphone PAL. Parallel work has shown that this new version of PAL is of equivalent difficulty to the existing classic PAL, and this study addresses characteristics of the smartphone version alone, in a well‐powered within‐subjects study designed to assess test‐retest reliability, bias, and the usefulness of a familiarisation session, as well as exploring usability.
Method
We obtained three PAL scores over three consecutive days in 86 (31 m, 55 f) participants aged 55+ (M = 57.6, SD = 5.65) via the Prolific online platform. Each day participants used their own devices to complete the smartphone PAL task. We analysed adjusted error scores up to eight‐box problems, with participants showing extreme score fluctuations (M +/‐ 3xSD) excluded on the basis that the task was not taken under comparable conditions. We analysed test‐retest reliability using Pearson correlations of scores over the three days and analysed the effect of a familiarisation session by looking at mean scores over three days.
Result
There was some attrition of users due to technical issues in the prototype version of the task, but adherence was good, and participants reported high levels of task enjoyment and satisfaction. Test‐retest reliability between familiarisation and Day Two was good, r = .63, and very good between Day Two and Day Three, r = .7. Analysis of mean scores showed that the familiarisation session had a significant effect, with task learning between familiarisation and Day Two significantly greater than between Day Two and Day Three. Participants reported good enjoyment of the task, and overwhelmingly described usability characteristics in positive terms.
Conclusion
This study highlights the value of a familiarisation session in reducing learning effects, and, like classic PAL, shows good to very good reliability. Moreover, PAL on a smartphone was well‐liked by these older participants, supporting its use in older populations.