Memories of Visual Events Can Be Formed Without Specific Spatial Coordinates

To what extent does specific spatiotopic location accompany the remembered representation of a visual event? Feature integration theory suggests that identifying a multi-feature object requires focusing on its spatial location to integrate those features. Moreover, single unit data from anterior ventral stream neurons that fire preferentially to complex objects indicates that they have retinotopic receptive fields. It can, therefore, be predicted that location information of features of a complex stimulus is inherent in the memory of a perceived visual stimulus’ representation. To evaluate this prediction, we presented participants with a brief array of characters with instructions to identify and locate the solitary letter among a set of digits. Surprisingly, analysis of trials in which the target identity was accurately reported indicated that in more than 15% of trials (i.e., in Experiments 2b & 2c) participants were almost completely uninformed about the location of the letter that they had just identified. Further analysis showed that there were two main sources of these location errors; misbinding the target to the distractors’ locations and extremely poor spatial representation of the target’s location to an extent that was indistinguishable from guessing. The latter finding indicates that consciously accessible representations of visual events can form despite being untethered to robust and spatially-specific representations, implying that the specific location was either not quite encoded into working memory, or was rapidly forgotten. However, when the target was marked by a single feature (color), there was no evidence of remembering the target identity without remembering its location even with strong masking.


Guessing correction formulas
To ensure that the observed failures to report location of identified targets were not the result of lucky guesses of identity, a correction for possible guessing was applied, and the result was subtracted from the observed proportions of trials to produce estimates of the proportions of trials in which subjects correctly perceived either identity or location.
Note that the essential conclusion is also supported by the uncorrected results. The guessing procedure is conservative with respect to our claim, in that it reduces the proportion of trials in which identity was reported without location.

Guessing correction for identity for all Experiments:
Step 1: The true value of identity accuracy was calculated for all experiments as follows Where Observed_Value is the observed identity accuracy, Est_True_Value is the estimation of the percentage of trials in which the subjects successfully identified the target. Chance_level is the chance of responding correctly on trials in which they had not perceived the target and were guessing at random using a ratio ( ⁄ ). Note that subjects did not know the set of letters used in the experiment.
Step 2: Calculation of what percentage of the observed value results from lucky guesses Awareness without spatiotopic information 2 (2) Note that in Experiments 2a, 2b, 2c and 3, identity guessing was estimated for each subject, instead of using the grand average. This enabled us to run paired-samples ttest between estimated lucky identity and estimated uniform location responses for each subject. The results indicated significant difference between the two, suggesting that not all uniform responses can be attributed to guessing the identity, and there are some trials that people had no idea of where the target was presented.

Guessing correction for location for experiments 1a and 1b:
Since location report errors were discrete in experiments 1a and 1b rather than estimated distributions, a different procedure for corrections was used to correct for guesses for location responses where identity was successfully reported. We present this second as it is more complex and harder to understand.
Step 1: Calculate location accuracy conditional on correct report of identity Step 2: Correct this value for location guesses Step 3: Calculation of identity lucky guesses Awareness without spatiotopic information 3 Step 4: subtract from to estimate corrected location accuracy -

Experiment 1S
The experiment is similar to Experiment 2a, except that participants were presented with 6 stimuli instead of 8. The results have been shown in Figure 1S. Figure 1S. Location report data of Experiment 1S for all correctly identified letters. The polar plot has been log-transformed for the ease of visualizing low-N bins.

Mixture model analysis:
In the mixture model analysis for location reports for correctly identified letters with the identity accuracy of 60.5%, the K value was 15.51 (SD = 14.79

Identity guesses correction:
The Paired-samples t-test indicated non-significant difference between identity guesses (M = 3.72%, SD = 2.44%) and estimated pU (M = 5.35%, SD = 9.47%); t(19) = .88, p = 0.39, BF 10 = 0.52. Although, it might seem that all uniform responses of location came from the identity guesses, given our subsequent Experiments with a greater number of stimuli, we argue that participants could probably remember all the 6 locations, and therefore made educated guesses when they could not remember the target location. Consider for example if participants were presented with just two items in a search array, all location reports would have been on the two locations, even if the target had not been bound to a specific location because participants might have a memory for the spatial configuration independently of the target's location. In this case, the tails of the histogram would consist of 100% Target and NT responses while pU would have been 0%.

Response order analysis:
The following table indicates identity and location responses based on the order of the questions. Comparing the pU% for identity-first and location-first trials suggests that probability of poor location responses were independent of the order of the questions in Experiments 2b and 2c.  Table 1S, ID first are the trials in which the identity question was asked first and Location first are the trials in which the location question was asked first. Note that the pU for location was estimated only for the correct identity trials.