The authors have declared that no competing interests exist.
Conceived and designed the experiments: CK PF DM DJS. Performed the experiments: CK. Analyzed the data: CK. Wrote the paper: CK PF DM DJS.
People sometimes fail to notice salient unexpected objects when their attention is otherwise occupied, a phenomenon known as inattentional blindness. To explore individual differences in inattentional blindness, we employed both static and dynamic tasks that either presented the unexpected object away from the focus of attention (spatial) or near the focus of attention (central). We hypothesized that noticing in central tasks might be driven by the availability of cognitive resources like working memory, and that noticing in spatial tasks might be driven by the limits on spatial attention like attention breadth. However, none of the cognitive measures predicted noticing in the dynamic central task or in either the static or dynamic spatial task. Only in the central static task did working memory capacity predict noticing, and that relationship was fairly weak. Furthermore, whether or not participants noticed an unexpected object in a static task was only weakly associated with their odds of noticing an unexpected object in a dynamic task. Taken together, our results are largely consistent with the notion that noticing unexpected objects is driven more by stochastic processes common to all people than by stable individual differences in cognitive abilities.
When their attention is otherwise engaged, people sometimes fail to notice a salient and fully visible, but unexpected object or event, a phenomenon known as
Many variations in the nature of the unexpected object influence noticing rates: size [
Although many aspects of the situation affect noticing, it is less clear whether some people are more prone to inattentional blindness than others. Typically, establishing stable individual differences in performance requires repeated tests of the same participants on the same task. Unfortunately, in inattentional blindness tasks, once people know that an object might appear, it will no longer be entirely unexpected, meaning that they might devote some resources to its detection. Consequently, most inattentional blindness studies include only one critical trial for each participant.
Rather than testing the same participants repeatedly, individual-difference studies of inattentional blindness have instead sought correlates of noticing or explored group differences. For example, studies of group differences find greater noticing by people on the autism spectrum [
The inconsistencies might simply reflect sampling noise or they could result from differences in the task used to induce inattentional blindness. Inattentional blindness tasks can be clustered into two categories, those governed by the limits of spatial attention and those driven by more central limits on attention [
Perhaps individual differences in cognitive resources will more reliably predict noticing for inattentional blindness tasks that draw on similar resources. For example, working memory or cognitive capacities may contribute to noticing more for those inattentional blindness tasks that are governed by central limits whereas individual differences in spatial attention may predict noticing for spatially-driven tasks. We explored whether controlling for the type of inattentional blindness task might yield more consistent individual differences in noticing. We hypothesized that individual differences in working memory capacity should contribute to noticing when the task is driven by central factors (e.g. expectations or cognitive load), whereas individual differences in attention breadth should contribute more to noticing when inattentional blindness is induced by diverting attention spatially.
In two studies with large, independent samples we assessed inattentional blindness using a static task (
Study 1 investigated the hypothesis that individual differences in working memory capacity would predict centrally-induced inattentional blindness whereas differences in attention breadth would predict spatially-induced inattentional blindness. Specifically, we predicted a link between working memory capacity and noticing for objects near the focus of attention, because failures to notice in such cases might result from central capacity limits. And, we predicted a link between attention breadth and noticing for unexpected objects appearing away from the focus of attention, because inattentional blindness in such cases might result from spatial limits.
Once participants have learned of the existence of an unexpected object, subsequent trials measure their ability to detect now-expected objects when performing the primary task. We predicted that individual differences in working memory capacity would correlate with noticing on these “divided-attention” trials because people with greater working memory capacity should be able to allocate their attentional resources more flexibly [
It has been argued that different processes might underlie the capability to inhibit known distraction and the tendency to stay unaware of
Methods, hypotheses, data preparation, and analyses for Study 2 were pre-registered before the data collection started. Although Study 1 was not formally pre-registered, we had specified our hypotheses in a prior grant proposal, and our procedures and predictions followed the same plan that we later pre-registered for Study 2. Data from the working memory and attention breadth tasks in Study 1 were presented in a previous paper that examined the relationship between these two constructs [
The reported studies were reviewed and approved by the ethics committee of the German Sport University Cologne. All participants gave written informed consent prior to their inclusion in the study and they were debriefed afterwards.
A total of 123 participants gave written informed consent, reported normal or corrected-to-normal vision, and received 13 € for their participation. Data from two of these participants were excluded from the analysis because they reported that they had anticipated the unexpected object. Four other participants either failed to report the unexpected object or could not identify either its position or its shape when looking for the critical object was their only task (i.e., with full attention). In such cases, the failure to notice an unexpected object becomes ambiguous because participants might not have followed instructions or they might have other (perceptual) problems that limit their ability to see the object even when they are trying to. In keeping with tradition in the inattentional blindness literature (e.g., [
There were no strong associations between inattentional blindness and any of the demographic variables we included. Details regarding these exploratory analyses and the exact statistical values can be found on
In addition to an inattentional blindness task, each participant completed a battery of cognitive tasks designed to measure different attention abilities. The battery included three working memory tasks, two attention breadth tasks, a Flanker task, and the German version of a questionnaire measuring cognitive failures in daily life (the CFQ; [
Participants completed a static inattentional blindness task (IB Cross task; [
Each participant completed ten easy practice trials (smaller arm 4.5°) before the thresholding began. Once the threshold had been determined for an individual, the critical trial occurred immediately and without forewarning. The cross on this critical trial used the threshold value determined for that participant, and a grey square (0.9° x 0.9°; RGB: 128,128,128) appeared along with the cross for the entire 200 ms. The square was always presented on one of the imaginary 45° lines bisecting the quadrants defined by the cross, with the particular quadrant chosen randomly for each participant. In the Near condition, the square appeared 2° from the center of the cross. In the Far condition, it appeared 7° from the center of the cross. After reporting which line they thought was longer, participants were asked if they had seen anything other than the cross that had not been present before. They were then asked how confident they were of their answer (very, somewhat, not at all), where the additional object had appeared (upper right, lower right, lower left, upper left), and which shape it had been (six choices). They were asked to guess if they had not noticed anything. Participants were coded as inattentionally blind if they did not report noticing the unexpected object or claimed to have seen something but could not define either its location or its shape.
After these questions, participants were told that the experiment would continue with more trials of the line-judgment task. Following three “normal” line-judgment trials (at threshold length), the grey square appeared for a second time and at the same distance from fixation as on the critical trial, although the quadrant was again chosen randomly. After reporting which arm was longer, participants answered the same questions that had followed the critical trial. Following the divided-attention trial, participants completed a final trial on which they were told to not perform the cross task (full-attention trial). The location (quadrant) of the additional square was again chosen randomly, but it was positioned at the same distance from the center of the cross as on previous trials. Participants were not asked to perform the line judgment, but were given the same questions about the additional square.
The three working memory measures included the automated version of the operation span task (Aospan; [
The two attention breadth measures included a useful-field-of-view task (UFOV; adapted from [
In our version of the UFOV task, participants tried to detect a peripherally presented circle among square distractors while also judging whether a central figure (< or >) pointed left or right (see
Sequence of events in a trial from the attention breadth measure useful-field-of-view (UFOV). One possible stimulus configuration was randomly picked for this display.
In the BoA task, participants fixated a central cross for 1000 ms and then judged the total number of gray circles appearing among two spatially separated pairs of shapes that appeared for 200 ms (circles and squares that were gray or black; see
Sequence of events from the breadth-of-attention task (BoA). Two possible stimulus configurations were randomly picked for this display.
The Eriksen Flanker task measures the ability to inhibit distractor stimuli [
Participants completed the German version of the Cognitive Failures Questionnaire (CFQ) to assess whether cognitive failures in daily life are associated with failures of awareness of unexpected objects. High scores in this questionnaire indicate a higher tendency to cognitive failures in daily life.
A chin rest (NovaVision, Magdeburg, Germany) was used for the inattentional blindness task, the BoA, and the UFOV. It was positioned 50 cm from a 24-inch display (resolution: 1920 x 1080 pixels, controlled by an Esprimo 710 3.3 GHz Core i3-3220 computer). For all other tasks, participants sat approximately 50 cm from the display. The Aospan was presented using E-Prime 2.0 (Psychology Software Tools, Pittsburgh, PA) and all other tasks were presented using Presentation (Neurobehavioral Systems, Berkeley, CA). Participants used a keyboard or, in case of the Aospan, a mouse to respond.
Participants were tested alone or in pairs. When participants were tested in pairs, they were separated by dividers that prevented them from seeing each other. Instructions were delivered on the screen prior to each task, but participants were also encouraged to ask questions before starting. The inattentional blindness task was always completed first, followed by the other cognitive tasks (three working memory tasks, two attention breadth tasks, Flanker task) in a randomized order for each participant. Whenever two participants completed the study in the same session, they were given the same randomized task order to minimize interruptions and distraction, and the experimenter waited for both participants to complete each task before starting the next task. A general questionnaire (retrieving demographics, anticipation of the unexpected object, and familiarity with inattentional blindness) was administered after the inattentional blindness task and the CFQ was administered after the completion of three cognitive tasks. The entire testing session took approximately two hours.
Due to technical errors, data were missing for some participants in some tasks (six subjects in the 2-Back-Identity task, four subjects in the 2-Back-Spatial task, three subjects in the BoA task, and three subjects in the Flanker task). One participant did not fully complete the CFQ. The correlational analyses included all participants for whom we had data for both measures. All of the reported significance levels are based on two-tailed tests.
As noted earlier, we analyzed the relationships between and among the working memory and attention breadth measures in a separate publication (see [
Given that we did not have separate hypotheses for the relationship between inattentional blindness and the individual working memory or attention breadth measures, we formed composite measures of both cognitive constructs to use in predicting noticing rates. This seemed reasonable as the working memory measures were moderately to highly intercorrelated with each other (2-Back-Identity and 2-Back-Spatial:
ALL | NEAR | FAR | ||||
---|---|---|---|---|---|---|
Notice | Notice | Notice | Notice | Notice | Notice | |
(critical) | (divAtt) | (critical) | (divAtt) | (critical) | (divAtt) | |
Working Memory | .08 | -.01 | .24 | .07 | -.08 | -.06 |
[-.10, .26] | [-.19, .17] | [-.02, .47] | [-.19, .32] | [-.34, .19] | [-.32, .21] | |
N = 116 | N = 116 | N = 60 | N = 60 | N = 56 | N = 56 | |
2-Back-Identity | .05 | -.10 | .23 | -.03 | -.13 | -.12 |
[-.14, .24] | [-.28, .09] | [-.04, .47] | [-.29, .24] | [-.38, .14] | [-.38, .15] | |
N = 110 | N = 110 | N = 56 | N = 56 | N = 54 | N = 54 | |
2-Back-Spatial | .04 | .09 | .04 | .18 | .06 | .03 |
[-.15, .22] | [-.10, .27] | [-.22, .30] | [-.09, .42] | [-.21, .32] | [-.24, .29] | |
N = 112 | N = 112 | N = 57 | N = 57 | N = 55 | N = 55 | |
Aospan | .07 | -.06 | -.03 | -.15 | -.08 | |
[-.11, .25] | [-.24, .12] | [-.28, .23] | [-.40, .12] | [-.34, .19] | ||
N = 116 | N = 116 | N = 60 | N = 60 | N = 56 | N = 56 | |
Attention Breadth | .14 | -.01 | .06 | .00 | .01 | |
[-.04, .31] | [-.19, .17] | [-.20, .31] | [-.26, .26] | [-.25, .27] | ||
N = 116 | N = 116 | N = 60 | N = 60 | N = 56 | N = 56 | |
BoA | .10 | -.04 | .02 | -.10 | -.02 | |
[-.09, .28] | [-.22, .15] | [-.24, .27] | [-.36, .18] | [-.29, .25] | ||
N = 113 | N = 113 | N = 60 | N = 60 | N = 53 | N = 53 | |
UFOV | .14 | .03 | .09 | .05 | .03 | |
[-.04, .31] | [-.15, .21] | [-.17, .34] | [-.22, .31] | [-.24, .29] | ||
N = 116 | N = 116 | N = 60 | N = 60 | N = 56 | N = 56 | |
Flanker | .08 | -.09 | .02 | -.19 | .13 | -.04 |
[-.11, .26] | [-.27, .10] | [-.24, .28] | [-.43, .07] | [-.14, .38] | [-.30, .23] | |
N = 113 | N = 113 | N = 59 | N = 59 | N = 54 | N = 54 | |
CFQ | -.22 | .15 | -.23 | .20 | ||
[-.45, .04] | [-.11, .39] | [-.47, .04] | [-.07, .44] | |||
N = 115 | N = 115 | N = 60 | N = 60 | N = 55 | N = 55 |
To create the composite measures we first standardized the scores for each task across individuals and then averaged each participant’s z-scores across measures for each construct to provide a single working memory z-score and a single attention breadth z-score for each participant. If a participant was missing a score on one of the measures, we formed the composite measure from the remaining measures.
Study 1 | Study 2 | ||||||
---|---|---|---|---|---|---|---|
Mean | Mean | ||||||
2-Back-Identity | 110 | 17.39 | 4.78 | 197 | 18.28 | 4.59 | 0.78 |
2-Back-Spatial | 112 | 16.04 | 5.52 | 197 | 17.05 | 5.19 | 0.49 |
Aospan | 116 | 37.06 | 17.85 | 198 | 40.18 | 16.27 | 0.69 |
BoA | 113 | 286.92 | 59.86 | 198 | 284.69 | 57.10 | 0.77 |
UFOV | 116 | 0.77 | 0.18 | 198 | 0.77 | 0.15 | 0.87 |
Flanker | 113 | 9.81 | 4.39 | 197 | 10.22 | 5.04 | 0.52 |
CFQ | 115 | 42.37 | 11.42 | 198 | 45.55 | 11.98 | 0.80 |
Navon | - | - | - | 198 | 9.96 | 13.09 | 0.59 |
Navon-Switchspeed | - | - | - | 198 | 21.87 | 12.21 | 0.33 |
We hypothesized that working memory measures might predict noticing in the Near condition but not the Far condition, and that attention breadth measures would predict noticing in the Far condition but not the Near condition. Inconsistent with our hypothesis, none of the measures was associated with noticing in the Far condition. Consistent with our predictions, Aospan was correlated with noticing in the Near condition. However, the other working memory measures were not, and the composite working memory measure was not significant either. Moreover, both attention breadth measures and the composite attention breadth measure were also associated with noticing in the Near condition. Note, though, that none of these correlations would have been statistically significant following correction for multiple tests, so they should be treated as suggestive and they require replication with a larger sample. Scatter plots of the relationships between inattentional blindness and the working memory and attention breadth measures are depicted separately for the Near and the Far condition in
Scatter plots of the relationships between inattentional blindness (0 = miss, 1 = notice) and the working memory and attention breadth measures in Study 1. The plots were prepared separately for the Near and the Far condition. The y-axes depict the test scores for each measure as described in the method section. Each circle represents a single participant. The blue lines depict the linear regression lines for each relationship.
To determine whether the cognitive measures in combination could effectively predict noticing of the unexpected object on the critical trial, we conducted two logistic regression analyses, one for the Near condition and one for the Far condition (see
NEAR | ||||||
Variables | Wald | Exp( |
Exp( |
Exp( |
||
Constant | -0.07 (0.28) | 0.07 | 0.93 | |||
Working Memory | 0.42 (0.41) | 1.06 | 1.53 | 0.68 | 3.43 | |
Attention Breadth | 0.66 (0.38) | 2.99 | 1.94 | 0.92 | 4.09 | |
Constant | 2.13 (1.56) | 1.86 | 8.41 | |||
Working Memory | 0.45 (0.42) | 1.16 | 1.57 | 0.69 | 3.57 | |
Attention Breadth | 0.68 (0.40) | 2.91 | 1.97 | 0.90 | 4.31 | |
CFQ | -0.05 (0.03) | 2.82 | 0.95 | 0.90 | 1.01 | |
Flanker | -0.01 (0.07) | 0.02 | 0.99 | 0.86 | 1.14 | |
FAR | ||||||
Variables | Wald | Exp( |
Exp( |
Exp( |
||
Constant |
-0.93 (0.32) | 8.74 | 0.39 | |||
Working Memory | -0.24 (0.46) | 0.27 | 0.79 | 0.32 | 1.93 | |
Attention Breadth | 0.04 (0.46) | 0.01 | 1.04 | 0.42 | 2.55 | |
Constant | 0.26 (1.31) | 0.04 | 1.29 | |||
Working Memory | -0.37 (0.48) | 0.62 | 0.69 | 0.27 | 1.75 | |
Attention Breadth | 0.01 (0.47) | 0.00 | 1.01 | 0.40 | 2.53 | |
CFQ | -0.05 (0.03) | 3.15 | 0.95 | 0.90 | 1.01 | |
Flanker | 0.08 (0.07) | 1.17 | 1.08 | 0.94 | 1.25 | |
*
Consistent with the correlational analysis, for the Far condition, the working memory and attention breadth composites explained almost none of the variance in noticing (approximately 1%), and including the CFQ and Flanker measures as predictors did not improve the model, either. Again, this finding does not support our hypothesis that attention breadth would predict noticing in the Far condition. One possible explanation for the lack of a relationship might be that the maximum spatial separation at which people can perform two tasks, as it is measured by the UFOV and the BoA, is not related to the direction and spread of attention within the IB task. The cross task does not require a wide spread of attention and people might not distribute their attention more broadly here even if they could.
For the Near condition, the composite working memory and attention breadth measures jointly accounted for about 12% of the variance. However, neither measure improved the model significantly on its own, and adding the CFQ and Flanker task did not significantly improve the model. This result mirrors that of the correlational analysis, suggesting only a relatively weak relationship between centrally-induced inattentional blindness and individual differences in working memory and attention breadth.
The lack of a link between the Flanker task and noticing in any of these analyses suggests that the ability to actively inhibit a known distractor is not related to missing unexpected items when focusing attention. This finding is consistent with the lack of a relationship between inattentional blindness and performance on the Stroop task [
We had also hypothesized that working memory capacity would predict performance on the divided-attention trial because working memory capacity typically has been linked to greater attentional control and to more flexible and goal-oriented allocation of attention [
Study 2 was designed to replicate and extend the findings of Study 1 with a larger sample size, a second inattentional blindness task, and additional cognitive measures. Specifically, we added a dynamic inattentional blindness task to explore the possibility that working memory and attention breadth might better predict performance when participants must sustain focused attention continuously for a longer time. In both tasks, we included a Near and Far variant, allowing us to replicate the conditions of Study 1 and to compare performance between comparable conditions in a static and dynamic inattentional blindness task.
By including a second task, we also could explore whether inattentional blindness in one task predicts inattentional blindness in another task. If performance in these two inattentional blindness tasks is uncorrelated, it is less likely that the tendency to notice or miss an unexpected object is a stable aspect of an individual. Instead, the lack of a correlation would imply that noticing on any given inattentional blindness task is a stochastic process, with some probability that any individual will happen to notice. If true, individual differences in other cognitive abilities also probably would not predict inattentional blindness across tasks and situations. To our knowledge, no previous study has examined whether noticing or missing unexpected objects is a stable individual-difference trait. Typically, once a participant knows that an unexpected object might appear, any subsequent trial of the same task no longer tests inattentional blindness. Previous studies have shown that people can miss unexpected objects even if they are familiar with the construct [
In addition to the second inattentional blindness task, Study 2 included two variants of the Navon task [
As in Study 1, we explored whether performance on the Flanker task would predict noticing. Although we found no evidence that individual differences in inhibition influenced noticing in the static inattention blindness task in Study 1, the dynamic task used in Study 2 requires participants to ignore some items. In that context, inhibitory control might enhance the ability to focus just on the attended shapes, thereby increasing inattentional blindness.
Study 2 was pre-registered. All hypotheses, procedures, data preparation, and analyses were specified in advance, and data and materials are available at the Open Science Framework (
A total of 200 participants gave written informed consent, reported normal or corrected-to-normal vision, and received 13 € and a chocolate bar for their participation. Data from two participants were excluded from the analysis because they could not understand the language in the instructions sufficiently well and the session had to be aborted. All participants successfully reported the unexpected object on the full-attention trial of at least one of the inattentional blindness tasks and no participant reported having expected the additional object on
There were no strong associations between inattentional blindness and any of the demographic variables we included. Details of these and other exploratory analyses can be found at
Except as noted, all materials and procedures were identical to those of Study 1. Study 2 added a dynamic inattentional blindness task (IB Motion task) and two variants of the Navon task (described below). The chin rest was used for both inattentional blindness tasks, the BoA, and the UVOF. The IB Motion task was always completed first and the static inattentional blindness task (IB Cross task) was presented last, with the remaining cognitive tasks interspersed between them. The two Navon tasks were always presented in succession, with the standard task first (from the participant’s perspective, they were essentially the same task). Otherwise, the order of the cognitive tasks was randomized for each participant. As in Study 1, whenever two participants took part in the same session, they completed the tasks in the same order. Each participant was assigned to the same condition for both inattentional blindness tasks (Near or Far). The general questionnaire was administered immediately after the IB Motion task and the CFQ was administered after the third cognitive task.
The IB Motion task was adapted from Most et al. [
Each participant first completed three practice trials in which the objects moved at 3.8° per second followed by three more in which they moved at 8.6° per second. After these practice trials, they completed a set of trials in which the speed of the objects was adaptively adjusted (3–1 method; [
In the Navon task, participants viewed centrally presented letters (5.15°) that were composed of a set of identical smaller letters (0.69°), with each horizontal and vertical segment of the larger letter consisting of five smaller letters. Participants were asked to judge whether there was an H or L in the display and to press the corresponding key on the keyboard. On each trial, either the large letter was an H or L or it was composed of all Hs or all Ls. The other stimulus on each trial was an F or T. The stimuli followed a 1500 ms fixation cross and remained visible until the participant had responded. Participants completed 16 practice trials (each stimulus combination twice) followed by 64 experimental trials (each combination 8 times), with trial types randomly intermixed. The primary measure of global/local attentional style was the percent increase in response time to local stimuli over the response time to global stimuli for correct trials: [(local-global)/global]*100.
The stimuli and procedure for the Navon-Switchspeed task were identical to those for the Navon task. The inter-stimulus interval was reduced to 16 ms, and participants completed 161 trials. For 80 of these, the target location was congruent with the preceding trial (a global target followed a global target or a local target followed a local target) and for 80 it was incongruent (a global target followed a local target or a local target followed a global target). The first trial was neither congruent nor incongruent and was not considered in the analysis. Congruent and incongruent trials were randomly intermixed, and the stimulus on each trial was randomly chosen from the set of letter combinations that fit the condition (i.e., from the four global or the four local stimuli). Only correct trials were analyzed and the primary measure (switch speed) was the percent increase in response time for incongruent trials relative to congruent trials: [(incongruent-congruent)/congruent]*100.
For the IB Motion task, three subjects reported having expected an additional object and six either failed to notice the object in the full-attention trial or failed to identify at least two of its features. After excluding data from these nine participants, the analyses included 96 participants in the Near condition and 93 participants in the Far condition for that task. Participants were coded as inattentionally blind if they did not report the unexpected object or claimed to have seen something but could not define at least two of the three features. For the IB Cross task, 13 participants reported expecting an additional object and three either failed to report the object in the full-attention trial or could not identify at least its position or its shape. After excluding data from these 16 participants, analyses of the IB Cross task included 95 participants in the Near condition and 87 participants in the Far condition. Participants were coded as inattentionally blind if they did not report the unexpected object or claimed to have seen something but could not define at least either its location or its shape.
Due to technical errors, data were missing for some participants in some of the cognitive tasks (one subject in the 2-Back-Identity task, one subject in the 2-Back-Spatial task, and one subject in the Flanker task). For correlational analyses, we included all participants who had data for both measures. As in Study 1, we formed a composite working memory measure from the three working memory tasks and a composite attention breadth measure from the two attention breadth tasks (see
For the critical trial in the IB Motion task, participants in the Near condition noticed the unexpected object more often (65.6%) than those in the Far condition (36.6%),
Study 2 provides one of the first opportunities to explore whether people who notice an unexpected object in one task are more likely to notice one in another task. If inattentional blindness is a stable personality trait across situations and paradigms, noticing on our two inattentional blindness tasks should be correlated. If, however, noticing is a stochastic process rather than a stable individual difference, noticing on one task may be unrelated to noticing on another. Data from two participants who inadvertently were assigned to different conditions in the two inattentional blindness tasks were excluded from this analysis, leaving a total of 172 participants. The association between noticing on one task and noticing on the other was small (
As in Study 1, we hypothesized that working memory capacity would predict noticing in the Near condition of each inattentional blindness task and that attention breadth would predict noticing in the Far condition of each task. Consistent with our predictions and with the results of Study 1, noticing in the Near condition of the IB Cross task was associated with the composite working memory measure, the 2-Back-Spatial task, and Aospan (see
ALL | NEAR | FAR | ||||
---|---|---|---|---|---|---|
Notice | Notice | Notice | Notice | Notice | Notice | |
(critical) | (divAtt) | (critical) | (divAtt) | (critical) | (divAtt) | |
Working Memory | .05 | .03 | .15 | .08 | ||
[-.10, .19] | [-.17, .23] | [-.06, .35] | [-.13, .29] | |||
N = 182 | N = 182 | N = 95 | N = 95 | N = 87 | N = 87 | |
2-Back-Identity | .04 | .01 | .01 | -.01 | .11 | .03 |
[-.11, .18] | [-.14, .16] | [-.19, .21] | [-.21, .19] | [-.10, .31] | [-.18, .24] | |
N = 182 | N = 182 | N = 95 | N = 95 | N = 87 | N = 87 | |
2-Back-Spatial | .05 | .00 | .08 | .10 | ||
[-.10, .19] | [-.20, .20] | [-.13, .29] | [-.11, .30] | |||
N = 181 | N = 181 | N = 94 | N = 94 | N = 87 | N = 87 | |
Aospan | .05 | .06 | .14 | .04 | ||
[-.10, .19] | [-.14, .26] | [-.08, .34] | [-.17, .25] | |||
N = 182 | N = 182 | N = 95 | N = 95 | N = 87 | N = 87 | |
Attention Breadth | .12 | .13 | .09 | .05 | .12 | .19 |
[-.03, .26] | [-.02, .27] | [-.11, .29] | [-.15, .25] | [-.09, .32] | [-.02, .39] | |
N = 182 | N = 182 | N = 95 | N = 95 | N = 87 | N = 87 | |
BoA | .10 | .07 | .03 | -.06 | .13 | .18 |
[-.05, .24] | [-.08, .21] | [-.17, .23] | [-.26, .14] | [-.08, .33] | [-.03, .38] | |
N = 182 | N = 182 | N = 95 | N = 95 | N = 87 | N = 87 | |
UFOV | .10 | .11 | .14 | .08 | .16 | |
[-.05, .24] | [-.09, .31] | [-.06, .33] | [-.13, .29] | [-.05, .36] | ||
N = 182 | N = 182 | N = 95 | N = 95 | N = 87 | N = 87 | |
Flanker | .13 | .04 | .17 | .06 | .08 | .01 |
[-.02, .27] | [-.11, .19] | [-.03, .36] | [-.14, .26] | [-.13, .29] | [-.20, .22] | |
N = 181 | N = 181 | N = 94 | N = 94 | N = 87 | N = 87 | |
Navon | -.08 | .10 | -.09 | .05 | -.10 | .13 |
[-.22, .07] | [-.05, .24] | [-.29, .11] | [-.15, .25] | [-.30, .11] | [-.08, .33] | |
N = 182 | N = 182 | N = 95 | N = 95 | N = 87 | N = 87 | |
Navon-Switchspeed | -.03 | -.09 | -.11 | -.14 | .04 | -.04 |
[-.18, .12] | [-.23, .06] | [-.31, .09] | [-.33, .06] | [-.17, .25] | [-.25, .17] | |
N = 182 | N = 182 | N = 95 | N = 95 | N = 87 | N = 87 | |
CFQ | -.10 | -.07 | -.05 | -.08 | -.17 | -.06 |
[-.24, .05] | [-.21, .08] | [-.25, .15] | [-.28, .12] | [-.37, .04] | [-.27, .15] | |
N = 182 | N = 182 | N = 95 | N = 95 | N = 87 | N = 87 |
ALL | NEAR | FAR | ||||
---|---|---|---|---|---|---|
Notice | Notice | Notice | Notice | Notice | Notice | |
(critical) | (divAtt) | (critical) | (divAtt) | (critical) | (divAtt) | |
Working Memory | -.10 | -.06 | -.03 | -.16 | -.16 | .05 |
[-.23, .04] | [-.20, .08] | [-.23, .17] | [-.35, .05] | [-.35, .05] | [-.16, .25] | |
N = 189 | N = 189 | N = 96 | N = 96 | N = 93 | N = 93 | |
2-Back-Identity | -.07 | -.07 | -.04 | -.10 | -.09 | -.04 |
[-.21, .07] | [-.21, .07] | [-.24, .16] | [-.30, .10] | [-.29, .12] | [-.24, .17] | |
N = 188 | N = 188 | N = 96 | N = 96 | N = 92 | N = 92 | |
2-Back-Spatial | -.10 | -.04 | -.08 | -.19 | -.07 | .13 |
[-.24, .04] | [-.18, .10] | [-.28, .12] | [-.38, .01] | [-.27, .14] | [-.08, .33] | |
N = 188 | N = 188 | N = 95 | N = 95 | N = 93 | N = 93 | |
Aospan | -.06 | -.02 | .05 | -.06 | -.18 | .02 |
[-.20, .08] | [-.16, .12] | [-.15, .25] | [-.26, .14] | [-.37, .03] | [-.18, .22] | |
N = 189 | N = 189 | N = 96 | N = 96 | N = 93 | N = 93 | |
Attention Breadth | -.11 | -.03 | -.02 | .03 | -.11 | |
[-.25, .03] | [-.17, .11] | [-.22, .18] | [-.17, .23] | [-.31, .10] | ||
N = 189 | N = 189 | N = 96 | N = 96 | N = 93 | N = 93 | |
BoA | -.04 | -.03 | -.03 | -.02 | -.12 | -.08 |
[-.18, .10] | [-.17, .11] | [-.23, .17] | [-.22, .18] | [-.32, .09] | [-.28, .13] | |
N = 189 | N = 189 | N = 96 | N = 96 | N = 93 | N = 93 | |
UFOV | -.02 | -.01 | .06 | -.11 | ||
[-.16, .12] | [-.21, .19] | [-.14, .26] | [-.31, .10] | |||
N = 189 | N = 189 | N = 96 | N = 96 | N = 93 | N = 93 | |
Flanker | .07 | .06 | .04 | .10 | .10 | .01 |
[-.07, .21] | [-.08, .20] | [-.16, .24] | [-.10, .30] | [-.11, .30] | [-.19, .21] | |
N = 188 | N = 188 | N = 95 | N = 95 | N = 93 | N = 93 | |
Navon | -.02 | .05 | .02 | .02 | -.05 | .08 |
[-.16, .12] | [-.09, .19] | [-.18, .22] | [-.18, .22] | [-.25, .16] | [-.13, .28] | |
N = 189 | N = 189 | N = 96 | N = 96 | N = 93 | N = 93 | |
Navon-Switchspeed | .07 | .11 | -.03 | .03 | .15 | .15 |
[-.07, .21] | [-.03, .25] | [-.17, .11] | [-.27, .23] | [-.06, .34] | [-.06, .34] | |
N = 189 | N = 189 | N = 96 | N = 96 | N = 93 | N = 93 | |
CFQ | .08 | -.04 | .16 | .20 | .13 | |
[-.06, .22] | [-.24, .16] | [-.04, .35] | [-.00, .39] | [-.08, .33] | ||
N = 189 | N = 189 | N = 96 | N = 96 | N = 93 | N = 93 |
Inconsistent with our prediction that people with greater attention breadth would be more likely to notice unexpected objects in a spatially-driven inattentional blindness task, noticing in the Far condition of the IB Cross task was largely unrelated to measures of attention breadth. And, contrary to our prediction, noticing in the Far condition of the IB Motion task was
Scatter plots of the relationships between inattentional blindness (0 = miss, 1 = notice) and the working memory and attention breadth measures in Study 2. The plots were prepared separately for IB Cross and IB Motion and for the Near and the Far condition. The y-axes depict the test scores for each measure as described in the method section. Each circle represents a single participant. The blue lines depict the linear regression lines for each relationship.
As in Study 1, we conducted binary logistic regressions separately for the Near and Far conditions of each inattentional blindness task, predicting noticing from the composite working memory measure, the composite attention breadth measure, the Navon task, and the CFQ. The working memory and attention breadth measures were entered first, with the other measures entered in a second block. For the IB Motion task (see
NEAR | ||||||
Variables | Wald | Exp( |
Exp( |
Exp( |
||
Constant |
0.65 (0.22) | 8.99 | 1.91 | |||
Working Memory | -0.08 (0.30) | 0.08 | 0.92 | 0.52 | 1.64 | |
Attention Breadth | -0.04 (0.29) | 0.02 | 0.96 | 0.55 | 1.68 | |
Constant | 0.96 (0.91) | 1.12 | 2.61 | |||
Working Memory | -.07 (0.31) | 0.05 | 0.93 | 0.51 | 1.71 | |
Attention Breadth | -.04 (0.29) | 0.02 | 0.96 | 0.55 | 1.68 | |
Navon | 0.00 (0.02) | 0.01 | 1.00 | 0.97 | 1.04 | |
CFQ | -0.01 (0.02) | 0.15 | 0.99 | 0.96 | 1.03 | |
FAR | ||||||
Variables | Wald | Exp( |
Exp( |
Exp( |
||
Constant |
-0.62 (0.23) | 7.49 | 0.54 | |||
Working Memory | -0.17 (0.35) | 0.23 | 0.85 | 0.43 | 1.68 | |
Attention Breadth | -0.52 (0.28) | 3.49 | 0.60 | 0.35 | 1.03 | |
Constant |
-1.99 (0.89) | 5.05 | 0.14 | |||
Working Memory | -0.12 (0.36) | 0.10 | 0.89 | 0.44 | 1.81 | |
Attention Breadth | -0.52 (0.28) | 3.38 | 0.60 | 0.34 | 1.04 | |
Navon | -0.02 (0.02) | 0.77 | 0.99 | 0.95 | 1.02 | |
CFQ | 0.03 (0.02) | 3.10 | 1.03 | 1.00 | 1.07 | |
*
For the IB Cross task (see
NEAR | ||||||
Variables | Wald | Exp( |
Exp( |
Exp( |
||
Constant |
0.56 (0.22) | 6.34 | 1.76 | |||
Working Memory |
0.72 (0.31) | 5.40 | 2.05 | 1.12 | 3.75 | |
Attention Breadth | 0.10 (0.29) | 0.13 | 1.11 | 0.63 | 1.95 | |
Constant | 1.20 (0.96) | 1.59 | 3.33 | |||
Working Memory |
0.72 (0.32) | 4.98 | 2.05 | 1.09 | 3.85 | |
Attention Breadth | 0.11 (0.29) | 0.15 | 1.12 | 0.63 | 1.97 | |
Navon | -0.00 (0.02) | 0.03 | 1.00 | 0.96 | 1.03 | |
CFQ | -0.01 (0.02) | 0.46 | 0.99 | 0.95 | 1.03 | |
FAR | ||||||
Variables | Wald | Exp( |
Exp( |
Exp( |
||
Constant |
-0.68 (0.23) | 8.38 | 0.51 | |||
Working Memory | 0.36 (0.37) | 0.94 | 1.43 | 0.70 | 2.94 | |
Attention Breadth | 0.19 (0.30) | 0.39 | 1.21 | 0.67 | 2.18 | |
Constant | 0.62 (0.96) | 0.41 | 1.85 | |||
Working Memory | 0.35 (0.38) | 0.87 | 1.42 | 0.68 | 2.98 | |
Attention breadth | 0.13 (0.31) | 0.17 | 1.14 | 0.62 | 2.08 | |
Navon | -0.01 (0.02) | 0.50 | 0.99 | 0.95 | 1.02 | |
CFQ | -0.03 (0.02) | 1.52 | 0.97 | 0.93 | 1.02 | |
*
Given that we did not make separate predictions for the influence of the Navon-Switchspeed task or the Flanker task for the Near and Far conditions, we combined across these conditions in each task and performed two separate regressions predicting noticing in each inattentional blindness task from these measures (
IB Cross | ||||||
Variables | Wald | Exp( |
Exp( |
Exp( |
||
Constant | -0.53 (0.45) | 1.43 | 0.59 | |||
Flanker | 0.05 (0.03) | 3.14 | 1.06 | 0.99 | 1.12 | |
Navon-Switchspeed | -0.00 (0.01) | 0.04 | 1.00 | 0.97 | 1.02 | |
IB Motion | ||||||
Variables | Wald | Exp( |
Exp( |
Exp( |
||
Constant | -0.55 (0.44) | 1.56 | 0.58 | |||
Flanker | 0.32 (0.30) | 1.13 | 1.03 | 0.97 | 1.09 | |
Navon-Switchspeed | 0.12 (0.01) | 1.03 | 1.01 | 0.99 | 1.04 | |
Given the significant and relatively large overall effect of condition (Near, Far) we additionally calculated a regression model that included the Flanker task and the Navon-Switchspeed task as part of the main regression model separately for Near and Far. These additional analyses can be found at
As noted in Study 1, the measures of attention breadth might not reflect the role of spatial attention in performing either the line-judgment task or the dynamic tracking task. Rather, both the BoA and UFOV measure the
As in Study 1, individual differences in working memory performance were not associated with noticing the additional object in the divided-attention trial for either condition of either inattentional blindness task. To the extent that working memory capacity is associated with the ability to divide attention across multiple task demands, that ability does not enhance the ability to perform the primary task while also trying to detect a critical object. Although the strength of this relationship is limited by the relatively high noticing rates in the divided-attention trials, the data show no clear trend favoring such a link.
The primary goal of these studies was to explore whether individual differences in attention and working memory predict inattentional blindness. More specifically, we tested whether inattentional blindness induced by attention to the wrong location would be predicted by differences in attention breadth and whether inattentional blindness induced by general resource limitations would be predicted by working memory differences (see [
None of our cognitive measures reliably predicted spatially-induced inattentional blindness, regardless of whether the inattentional blindness task was static (
Taken together, the link between individual differences in working memory and noticing of unexpected objects is fairly small and does not generalize to inattentional blindness tasks and situations for which working memory might be expected to play a role. This pattern raises the possibility that such differences are less important for more ongoing, dynamic attention-demanding situations like those we might experience in the real world.
Most prior studies of individual differences in noticing have tried to predict noticing of a single unexpected object based on personality or cognitive ability measures. Study 2 was among the first to compare noticing by the same individuals in two inattentional blindness tasks. To our knowledge, only two other studies examined inattentional blindness within the same individuals across multiple instances. One used a small sample (N = 36) to test whether familiarity with and prior experience in an inattentional blindness study would affect noticing [
The challenge in measuring inattentional blindness twice is that, for the second task, participants know that an additional object might appear. Thus, the additional object in the second task is not entirely unexpected, meaning that it might not measure inattentional blindness. For several reasons, we believe that our second task did measure inattentional blindness and that our participants did not anticipate the unexpected object in the second task: (1) The two tasks were embedded in a larger battery of cognitive tasks; (2) The tasks looked different and the primary task in each placed different demands on participants; (3) Few participants reported having expected anything unexpected to appear on the second task; and (4) Noticing rates in the static task were comparable across studies, suggesting that expectations did not alter noticing rates substantially.
Apparently, the tendency to notice unexpected objects is not a stable and robust individual difference that applies in a general way across a variety of tasks; noticing of an unexpected object in one task only weakly predicted noticing in the other. Nonetheless, even if proneness to inattentional blindness is not a stable trait that applies across a range of situations and tasks, individual differences may reliably predict noticing for a given task. For example, distinct individual differences factors might affect noticing in static and dynamic tasks.
Given that it is not possible to measure inattentional blindness repeatedly using the same task (but see [
Our goal was to measure individual differences in the ability to notice unexpected objects, not the ability to perform the primary task. Equating primary-task performance is perhaps the best way to isolate individual differences in noticing from individual differences in primary-task performance. Although other studies suggest that individual difference in primary-task performance do not predict noticing [
Although we found little relationship between noticing and other cognitive tasks, our ability to detect such correlations might have been limited by the reliability of those measures. Several of our measures (2-Back-Identity, Flanker, and Navon-Switchspeed) had low test-retest reliabilities (based on our small sample of retested participants), meaning that their usefulness as predictive measures might be relatively low. Still, even those measures that showed high reliability, including the main measures of attention breadth and working memory, were largely unrelated to noticing.
Assessing the reliability of individual inattentional blindness tasks is problematic due to the single-trial nature of the phenomenon [
Another approach to exploring the reliability of an inattention blindness task might be to vary expectations over the course of a study; when the properties of a critical object vary over the course of a study, wrong expectations can produce repeated inattention blindness in the same task [
Our ability to predict noticing from individual-differences measures might have been hampered by a restricted range of performance on the cognitive tasks. Even though all of our measures did show substantial between-subject variability, meaning we had sufficient variability to detect associations among them, our sample consisted of university students, and ideally, these findings should be replicated with a non-student population. Performance on many of the cognitive tasks we used [
Individual differences in cognitive abilities such as attention breadth and working memory do not reliably predict noticing of unexpected objects. Moreover, working memory and attention breadth did not separately predict noticing in centrally- and spatially-induced inattentional blindness. Although working memory capacity was weakly associated with noticing in the central condition of a static inattentional blindness task, it did not predict noticing in the comparable condition of a dynamic task, suggesting that working memory does not predict noticing of unexpected objects in general. And, to the extent that it does predict noticing, the association is relatively weak. The minimal correlation between noticing in our two inattentional blindness tasks suggests that the ability to notice unexpected objects in general is not a stable individual-difference trait. Consequently, inattentional blindness appears to be driven more by situational and task factors, or even by chance, than by individual-differences variables.
The authors thank Carla Greving, Constanze Rau and Karlotta Schlösser for their assistance in data collection.