Skip to main content
. Author manuscript; available in PMC: 2017 Aug 3.
Published in final edited form as: Neuron. 2016 Aug 3;91(3):694–707. doi: 10.1016/j.neuron.2016.07.006

Figure 1. Spatial WM performance recovers following a retro-cue.

Figure 1

(A) On each trial, participants viewed 2 target stimuli (red and blue dots). A subsequent change in the color of the fixation point to red, blue, or purple cued participants to remember the location of the red target, the blue target, or both targets (respectively). On 33% of trials, we cued participants to remember the location of one target over the entire delay interval (fixation became red or blue, Remember 1; R1). On the remaining 67% of trials we cued participants to remember the locations of both targets (Remember 2; R2). This set of trials was further divided in half: on R2-neutral trials, we gave no information about which item was relevant (fixation point became black after an 8 s delay); on R2- valid trials the fixation point became red or blue, indicating which target would be probed at the end of the trial. After the 16 s delay, participants adjusted a horizontal or vertical bar to match the position of the remembered target. Dashed yellow circles illustrate remembered locations.

(B) The 2 targets appeared at positions uniformly drawn from 2 discs (0.6° radius centered 3.5° from fixation; colored circles within dashed annulus). Targets never appeared within the same disc; they appeared ± 60°, ± 120°, or ± 180° polar angle apart on each trial. We randomly rotated the entire target arrangement on each trial.

(C) Memory performance was lower (i.e., higher recall error) during R2-neutral trials than R1 trials in all n = 6 participants. However, a valid cue (R2-valid) improved performance relative to R2-neutral trials, though performance remained lower than R1 trials. Asterisks indicate significance determined by pairwise resampling test, Bonferroni corrected for 3 comparisons. Boxes with horizontal lines indicate 95% confidence intervals (CIs) computed via resampling and mean over resampling iterations, respectively. Each symbol in (C) is a single participant. See Fig. S1 for recall error histograms, and Fig. S2 for univariate fMRI activation for each condition.