Perceptual Learning Generalization from Sequential Perceptual Training as a Change in Learning Rate

Abstract

With practice, humans tend to improve their performance on most tasks. But do such improvements then generalize to new tasks? Although early work documented primarily task-specific learning outcomes in the domain of perceptual learning [1-3], an emerging body of research has shown that significant learning generalization is possible under some training conditions [4-9]. Interestingly, however, research in this vein has focused nearly exclusively on just one possible manifestation of learning generalization, wherein training on one task produces an immediate boost to performance on the new task. For instance, it is this form of generalization that is most frequently referred to when discussing learning "transfer" [10, 11]. Essentially no work in this domain has focused on a second possible manifestation of generalization, wherein the knowledge or skills acquired via training, despite not being directly applicable to the new task, nonetheless allow the new task to be learned more efficiently [12-15]. Here, in both the visual category learning and visual perceptual learning domains, we demonstrate that sequentially training participants on tasks that share a common high-level task structure can produce faster learning of new tasks, even in cases where there is no immediate benefit to performance on the new tasks. We further show that methods commonly employed in the field may fail to detect or else conflate generalization that manifests as increased learning rate with generalization that manifests as immediate boosts to performance. These results thus lay the foundation for the various routes to learning generalization to be more thoroughly explored.

Trial-dependent psychometric functions accounting for perceptual learning in 2-AFC discrimination tasks

“…Future work with larger datasets will be necessary to identify the best parameterizations of both learning functions and the psychometric functions (and the correct parameterizations may differ across learning domains). Indeed, we note that we ourselves have used a slightly different parameterization in our previous empirical work (Green et al, 2015;Kattner et al, 2017), which used even fewer free parameters, at the cost of additional flexibility. The best tradeoff between these is thus also to be determined.…”

“…This issue also speaks to another benefit of the continuous parametric model-the ability to more fully examine the functional form of learning. Although our approach here was purely descriptive (i.e., the eventual functional form we chose was simply the one that provided the best overall fits to the datasee Supplemental Materials for other parameterizations; also see Kattner et al, 2017;Snell et al, 2015), the general framework can be easily used to test explicit predictions about the best fitting functional form. This is critical as the observed functional form of learning in a task limits the possible mechanistic models that need to be considered.…”

“…Further, because the continuous time parametricmodel fit has the potential to provide estimates of first trial and last trial performance, it could be additionally useful in designs seeking to address questions regarding total learning (i.e., by comparing first trial and last trial estimates rather than first block and last block), rate of learning, as well as questions related to transfer of learning (Kattner et al, 2017). In the latter cases, ''transfer'' could be calculated as the difference in performance on the final trial of the training task and the first trial of the transfer task, or the increase in learning rate in the transfer task when compared with the training task.…”

“…There can also be a change in the rate at which performance improves on the posttest (i.e., performance on the first trial of the posttest matches performance on the last trial of the pretest, but then performance on the post-test rapidly improves). We have referred to these different functional forms of generalization as ''transfer'' and ''learning to learn'' respectively (Kattner et al, 2017). Critically, it is easy to misidentify the functional form if a block approach is used to analyze pretest and posttest data (since such an approach does not given an estimate of immediate performance on the posttest nor an estimate of how performance changes throughout the posttest, but instead considers performance during the posttest to be stationary).…”

“…These models regularly use some form of Hebbian learning principle, wherein the strength of the connection between two nodes is updated after each learning epoch by an amount proportional to the extent to which the two nodes were simultaneously active during the learning epoch. Finally, Bayesian learning models are inherently continuous in nature, as each observed training example increases or decreases the probability that a particular estimate/hypothesis is correct (by an amount that depends on the strength of the evidence Citation: Kattner, F., Cochrane, A., & Green, C. S. (2017). Trial-dependent psychometric functions accounting for perceptual learning in 2-AFC discrimination tasks.…”