A Novel Test of Pure Irrelevance-Induced Blindness Flashcards Preview

Wien Aufnahmeprüfung M.Sc. Psychologie 2020 > A Novel Test of Pure Irrelevance-Induced Blindness > Flashcards

Flashcards in A Novel Test of Pure Irrelevance-Induced Blindness Deck (27)
Loading flashcards...

What is load theory's claim?

  • that bottom up attention is possible under low perceptual load but not under high


What study's results is at odds with load theory?

  • Eitam et al. (2013)
    • found that under low load (two colors in the display, a ring and a disk) an instruction (top-down) led to better memory performance for the relevant color than for the irrelevant color.
    • control condition: if participants were instructed (top-down) to to attend to both objects participants could memorise both.


What was the conclusion of Eitam's et al. (2013) study?

That if a stimulus was irrelevant that decreased the probablility that one would remember its color, also under low perceptual load.


What issues did the present study take with Eitam et al.'s (2013) study?

  • sub-optimal design
    • lack of more implicit priming measures
    • long interval between stimulus and memory task (more than 500ms)


How did the present study try to improve the study of Eitam et al. (2013)?

  • by improving the retrieval displays but leaving encoding displays as in the original
    • also by reducing time interval
  • and by using an implicit measure


Did the present study replicate findings by Eitam et al. (2013)?

  • partly
    • no evidence of "irrelevance induced blindness"when the ring was assigned relevance
    • but CCMA (continuously cumulative met-analysis) shows that findings do not refute findings completely (if they are added with the study of EItam et al. (2013) they are still significant... but what's the point if they had suboptimal designs?


Explain the main principles of visual attention and processing.

visual attention is guided by principles that inform the selection of visual stimuli:

  • bottom up capture of attention: captures attention by salience
  • top-down control: selection of most relevant timuli


What does load theory have to do with top-down and bottom-up attention?

  • it seeks to explain the interplay between the two
  • it claims that bottom-up attention is possible under conditions of low perceptual load but not high perceptual load


  1. How is Eitam et al's (2013) study at odds with the claims of load theory?
  2. How was their study designed to enable such a claim?

  • because their study had a low perceptual load and still irrelevant but salient visual input was blocked
  1. participants were instructed to focus either on a ring (disc) or out ring (ring)
  2. on the next display (encoding display) the disk was red and the ring was yellowthan (retrieval display)
  3. participants were asked to report the shown colors
  4. there was only one trial (oterwise participants would know that they need to remember both colors and might focus on both)


  • those that were (beforehand) asked to retrieve the ring were better at retrieving the ring (same with the disk)
  • but control condition had to remember both colors and were generally able to
  • hence, a person would have had enough resources to remember both but they still did better at remembering the relant object's color
  • so top-down regulation led to blocking the irrelevant stimulus even if it had enough resources - load theory would say that it would only block it if enough resources have been taken by top-down perception
  • so load theory would have predicted no difference in remembering colors (accuracy)


What were the methodological limitations of Eitam et al. (2013) and how did the present study try to overcome them?

  1. Eitam et al. (2013) only measured explicit memory retrieval accuracy via pressing of buttons
    • but by additionally measuring the eye-movements and fixations implicit memory (priming) could be assessed by evaluating if a person fixates more on the color that they were primed with than the new color on the retrieval display
    • this would give more evidence for possible  bottom -up selection of irrelevant stimulus (more sensitive) revealing resiudal implicit memory
  2. the interstimulus interval was 500ms but functionally it was longer because than the instruction had to be read on the retrieval display and then choose a color
    • in the present study instructions were given auditorily so the eyes were free to already look at the colours wile listening to the instruction. Hence the interstimulus interval was also functionally 500ms.
  3. in Eitam's (2013) study, participants were asked to sequentially retrieve relevant and irrelavnt stimulus'colour... this may have increased time interval from intial stimulus and/or included carry over effects (from relevant object's color to irrelevant's e.g.)
    • in the present study only one color was asked to be retrieved
      • half reported the disk and half the ring.
      • for half or each group the asked-about object was relevant and for the other half irrelevant... hence 25% of participants per group


What were the expectations of the present study?

  • in line with Eitam et al.(2013) that there will be better memory acuracy of the color of the relevant stimulus compared to the irrelevant stimulus
    • load theory would predict the opposite, resources should be enough to attend and retrieve both stimulus' colors explicitly or implicitly


How did the sample of the present study look?

  • 108 university sudents (University of Vienna)
    • 79 female
    • Mage 21.52
  • all had corrected or normal eyesight
  • no ethical approval was needed due to Austrian Universities Act of 2002
  • informed consent



Did the present study conduct a power-analysis and how were missing or invalid responses handled?

  • power analysis was based on effect-sizes from Eitam et al. (2013)
  • because original and replications stdies have a few sources of variation (different sample pools, methods) and orignal studies often have inflated effect-sizes a power of 0.90 was aimed at (G*Power)
    • hence, 49 participants were needed in each conition (total: 98) + a few in case people drop out
  • four participants were excluded in the explicit measures leaving a power of 0.92
  • fifteen participants did not fixate on either stimulus and were excluded from the implicit task leaving a power of 0.87


What were the materials (apparatus) involved?

  • a CRT (Cathode Ray Tube) monitor
    • refresh rate of 85 hz
  • standard QWERTZ keyboard
  • stimuli were presented and responses were collected with Psychophysics (MATLAB
  • recording of fixations EyeLink 1000  (SR research )
    • temporal resolution of 1000hz


Explain the properties of the stimulus.

  • a disk surrounded by a ring
  1. encoding display
  • disk colored red or yellow, ring opposite colour
    • jointly presented for 500ms
    • no instruction to encode stimulus
  1. then a retention interval of 500ms with a fixation cross
  2. retrieval display
    • participants that had to remember disk colos were shown two disks and thos that had to remember a ring color were shown two rings
      • one of the colours was the same as in the encoding display and one was new
    • simultaneously an auditory instruction was asking participants on which side of the screen a disk/ring with the same color as in the encoding display was shown
    • display was shown until response was given


What were congruent and what were incongruent conditions?

  • congruent condition: relevant stimulus during encoding had to be retrieved
  • incongruent condition: irrelevant stimulus had to be retrieved

participants had to report colour of either ignored or focused upon stimulus


Pls. descrie the procedure.

  • particpants ocular dominance was establishedd
    • nine-point calibration and validation
    • eye tracker was centrally below monitor
  1. instruction about what to focus on
  2. after presentation they heard auditory instructions
    • 4s duration of instruction
  3. position of response key corresponded with side of the chosen stimulus' colour


How was the data pre-processed?

  • eye tracking data was processed with R
    • used package was saccades
    • velocity based saccade detection algorhythm
      • every event between two saccades is a fixation
    • Regions of interest were defined (ROI's)- circles with a 2.7*
    • only fixations after 100ms were recorded
  • data-analysis was done with JASP
  • artifacts (e.g.) blinks were removed


Please summarise the results for the explicit memory task!

  • four participants were excluded
  • RT's from 0.79 to 10.79
  • Overall accuracy was 87.5%


  • pecentage correct congruent trials: 94%
  • percentage correct incongruent trials: 81%
    • congruency v. inconrgruency X2= 3.17, p = 0.075 - so almost significant 
  • However, congruency vs. incongruency's relation to accuracy was significant witthiin the disk condition (p = .009) but not in the ring condition (p = .68)


Please summarise results of implicit memory measure.

  • 93 participants fixated on at least one of the two colours


  • pattern of results similar as in the explicit memory measure:
    • no significant difference between fixations on  relevant and irrelevant rings
    • but quantitatively more fixations on relevant disk compared to irrelevant disk
  • overall, difference in congruent vs incongruent conditions on the accuracy of peoples' first fixation (ROI) was insignificant (p = .56)
    • planned proportions test  showed that the proportion of correct fixations in congruent v. incongruent trials was not significant, neither in the disk nor in the ring condition


What is meant by and what are the resuts of "Continuously Cumulating Meta-Analysis (CCMA)?

  • CCMA is a tool to increase precision of estimated effect sizes
    • by combining already existing evidence with the new results
    • therefore counteract publication bias
    • it is an ordinary meta-analysis, reported effects are combined and weighted according to their sample size... an analysis with a larger sample size has more weight
  • this study did not find the same effects as EItam et al. (2013)
    • data of the explicit memory task were computed in Odds Ratio (OR) as well as in studies 1a and 1b by Eitam et al (2013)
    • odds ration in this case was the odds of a correct response (color) in congruent compared to incongruent conditions
      • a positive value was indicative of a positiveeffect of congruency, negative = negative effect, 0 = neutral effect
    • even though the resulting diamond is  smaller than in EItam et al (2013) it is still highly significant
  • a random effect model was computed as it assumes that differences are not only due to sampling variance but also from other sources (e.g. design)


Summarise the key results of the study that aimed to repicate Eitam et al's (2013) study.

  • explicit measure:
    • significantly higher accuracy for relevant compared to irrelevant stimuli only in the disk condition (not in the ring condition)
  • implicit measure (first fixation on repeated v. normal colour):
    • neither disk nor ing condition had significantly different accuracies in relevant v. irrelevant stimuli
  • this study claims to have more sensitive methodology
  • so only a small part of the study is in line with Eitam et a's (2013) ressults


What are plausible explanations for the finding that there was a difference in accuracies in relevant v. irrelevant stimuli in the disk condition?

  • and the disks were presented more centrally than the rings, hence more in line with the gaze
    • possibly due to the fact that the gaze tends to center  on the screen as well as on an object
  • another explanation is that relevant rings in the periphery entail the processing of irrelevant disks because it is within the ring,
    • in other words:proccessing of the ring correspnded to processing of the stimulus as a whole aand processing of the disk did not
      • but if the disks are relevant the gaze might be able to focus on it with exclusion of stimuli in the periphery
  • lastly, attention may be more easily guided to part-whole-color-color conjunction but not to part-part color-color conjunctions


Name the additional arguments that were mentioned against load-theory.

  1. when integrating the present results with those of Eitam et al. (2013) in CCMA results showed irrelevance induced blindness in the low-load condition, which is countering load-theories claims
  2. contingent-capture experiments showed that
    • participants did often not recognise a cue that appeared prior to the targeted stimulus if it hadn't had the same color as the looked for target
    • this cue however, was very salient and it should have been detected with bottom-up processes, it was not due to top-down tasks.
      • even though perceptual load was low (a singleton cue and a single target)
    • also cue-elicited N2pc an electrophysiologic marker was absent


Could there be a distinction between cognitive load and perceptual load (according to this study)?

  • it is possible that even though there was low perceptual load (a ring and a disk) there may have been high cognitive load if participants had to reflect upon the instructions as they heard them after they had seen the stimuli
    • according to load theory high cognitive load invites more processing of irrelevant stimuli, hence, in this study due to higher cognitive load irrelevant stimuli may have been processed more and consequently were higher accuracy rates for relevant and irrelevant stimuli
    • however, why woudl this not have affected the study by Eitam et al. (2013) as well?
    • the argument of high cognitive load does not explain the results of contingent-capture experiments either


What effects may have had the time interval between encoding and retriveing (500ms)?

  • leveling processes may have influenced memory (e.g. active inhibition)
    • features of inittially-attended to stimulus can be quickly forgotten - attribute amnesia
  • during the time of memorising without instructions such processes may have distorted the memory



  • study with (theoretically) more sensitive methodology failed to conceptually replicate Eitam et al's (2013) study
  • evidence for irrlevance induced blindness was found in disk but not in ring condition
  • CCMA still showed overall significant reslust (Eitam et al's + present study)
  • implicit memory accuracy results were not in line with resluts by Eitam et al (2013)
  • maybe more sensitive methods that usee relevant v. irrelavvnat colors "colser to the these stimuli" (e.g. online) rather than memory performance