July 2013
Volume 13, Issue 9
Free
Vision Sciences Society Annual Meeting Abstract  |   July 2013
Statistical Inferences Depend on Working Memory Capacity Estimates
Author Affiliations
  • Melissa Trevino
    Dept of Psychology, Univ of Houston
  • Jane Jacob
    Dept of Psychology, Univ of Houston
Journal of Vision July 2013, Vol.13, 1355. doi:10.1167/13.9.1355
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Melissa Trevino, Jane Jacob; Statistical Inferences Depend on Working Memory Capacity Estimates. Journal of Vision 2013;13(9):1355. doi: 10.1167/13.9.1355.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Three versions of the change detection paradigm have been used to measure visual working memory (VWM) capacity, 1) whole-display probe, and two types of single-display probes: 2) probe at prior display location and 3) probe at central fixation. Past VWM capacity estimates, K, have often used Cowan’s or Pashler’s formula regardless of single-item and whole-display probes. However, Rouder et al. (2011) argue that Cowan’s K estimate is appropriate for single-item probe, whereas Pashler’s K estimate is appropriate for whole-display probe. Additionally, Cowan et al. (2012), have recently reported a third K estimate for a single-item probe presented at the center of fixation rather than at its WM array location. Here we investigate how variations of K-estimate formulae might affect inferences drawn from studies of VWM. WM load (4, 6 items), memory feature (color, shape), and type of probe presented (single-item probe centered at fixation, single-item probe centered at its prior location, whole-display probe) were varied. Cowan’s (2001) K estimate, Pashler’s (1988) K estimate and a variant of the appropriate estimate to include Cowan et al.’s (2012) K estimate for a single-item probed at the center of fixation were compared. Results indicate that the type of formula used interacts with main effects of Feature, Probe Type, and WM Load; with the two-way interactive effects of Feature x Probe Type, Feature x Load, and Probe Type x Load; and with the three-way interactive effects of Feature x WM Load x Probe Type. These findings indicate that, in multiple ways, the inferences one might draw from results obtained in VWM studies, while also depending on feature to be remembered, depend significantly on the formula used to compute K.

Meeting abstract presented at VSS 2013

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×