Paradigm and Hypotheses

This document plots the results of a 60 min version of Irrelevant Cueing, in which subjects performed a simple 1-up, 2-down calibration procedure prior to performing the main cueing task - adjusting the contrast of the target. Following calibration, in which subjects performed the discrimination task in the absence of cues, subjects completed 10 blocks of a simple exogenous cueing task with a twist. Subjects are asked to keep their eyes at the center of the screen and monitor three possible target locations on an imaginary circle around the fixation dot (at 90, 210, 330 degrees) for the occurrence of the target. Prior to the target appearing, a random one of these target locations flashes white. Subjects are informed that this flash is random and meant to be distracting, and therefore should ignore it. Following the occurrence of the cue, the target appears. The target is a gabor patch oriented 30 degrees clockwise or counterclockwise from vertical, and is immediately followed by a visual noise mask at that location. In this task, the target never appears at one of the three locations, rendering it irrelevant once subjects implicitly realize this to the be the case (probably during calibration). At the end of each trial, subjects report which direction the target was turned and report how vividly they experienced the target using the Perceptual Awareness Scale. The scale has 4 points (colloquially outlined here): (1)no experience, (2) I think something was there, (3) I have a good guess, and (4) explicit experience. Subjects report the direction of the target turn using the n (CCW) and m (CW) keys, and the vividness rating using the 1-4 keys at the top of the keyboard.

The purpose of this study is to test whether subjects have “control” over where their attention is exogenously oriented. If exogenous attention is truly reflexive and independent of top-down information, then we should see that d’ is similar following invalid cues both when presented at a possible target location and when presented at the irrelevant location. However, if top-down information can influence the exogenous orienting of attention, then we would expect performance to be better following invalid cues presented at the irrelevant location vs. a possible target location. This is because we would expect fewer attentional resources to be allocated to the cue at the irrelevant location, leaving more resources to process the target location.

There were 3 versions of this experiment, which varied only in what target location was irrelevant. In V1, the top location was irrelevant; in V2, the left location; in V3, the right location.

Note: all error bars are within-subject SEM. Subjects who had a d’ of less than .5 or greater than 3 overall, or more than 10% of trials marked as RT outliers, were excluded.

Subject (N = 141) overall performance on cueing task, before exclusion

Overall d’ for each subject

Calibrated contrast values

Cueing task accuracy (N = 90) after exclusion

Target d’

Size of d’ cueing effect for each subject



Cueing task RT after exclusion

Response Time

Size of RT cueing effect in each subject



Cueing task vividness ratings after exclusion

Average Vividness Response

Vividness Response Distribution

Distribution of vividness ratings, based upon cue validity (valid or invalid)

Statistics

d’ ANOVA

Effect DFn DFd SSn SSd F p p<.05 ges partial_eta_squared
(Intercept) 1 89 1010.583 97.596 921.573 0 * 0.898 0.912
validCue 2 178 2.349 17.523 11.932 0 * 0.020 0.118

d’ t-tests

comparison t_val p_val
valid vs. invalid 4.821697 0.0000058
valid vs. irrelevant 1.792302 0.0764826
irrelevant vs. invalid 3.170781 0.0020857


log10(RT) ANOVA

Effect DFn DFd SSn SSd F p p<.05 ges partial_eta_squared
(Intercept) 1 89 2004.177 10.496 16994.539 0.000 * 0.994 0.995
validCue 2 178 0.039 0.704 4.877 0.009 * 0.003 0.052

RT t-tests

comparison t_val p_val
valid vs. invalid -0.5532306 0.5814926
valid vs. irrelevant 2.1858293 0.0314532
irrelevant vs. invalid -3.2680702 0.0015397


VR ANOVA

Effect DFn DFd SSn SSd F p p<.05 ges partial_eta_squared
(Intercept) 1 89 2475.719 84.664 2602.512 0.00 * 0.966 0.967
validCue 2 178 0.200 3.748 4.747 0.01 * 0.002 0.051

VR t-tests

comparison t_val p_val
valid vs. invalid -0.4003396 0.6898657
valid vs. irrelevant -2.6499009 0.0095279
irrelevant vs. invalid 4.3766746 0.0000327