Periodic Reporting for period 1 - NeuraCOLM (The Neural Correlates of Object Location Memory)
Période du rapport: 2022-08-01 au 2024-11-30
Understanding how we remember the location of objects is fundamental to everyday life. While this ability, known as object location memory (OLM), is crucial, the underlying mechanisms, particularly the influence of object knowledge, remain largely unknown. This project aimed to expand our knowledge of OLM by investigating both individual differences in spatial cognition and the neural correlates involved, specifically considering the role of contextual factors such as knowledge about objects (e.g. ownership), and environmental cues. This research addresses internal spatial representations, reflecting the flexible and individualised nature of our cognitive processes. We conducted 8 experiments, acquiring data from 269 participants.
In the first line of studies, the project looked at individual differences, exploring variation in spatial memory across gender and age groups. A first exploratory study was designed as a stepping stone into examining the development of spatial and temporal memory. Participants were presented with five objects in a contextual cluster (e.g. objects one finds in a museum, in a garden, etc.). The objects were presented one-by-one at a given location on the screen, at a given order. The contextual information was manipulated in 3 levels: no context, 3 landmarks on the screen, or a full image in which objects were placed at ‘possible’ locations (ie,. objects would be on a shelf, the floor, etc). After object presentation, participants indicated first which objects they saw (testing object memory), then in which order they saw them (temporal memory), and finally where in the scene they saw the objects (testing object-location binding). We found a significant effect of age, but no effect of gender. One reason for this might be a floor effect, as most age groups performed at (near) ceiling level when remembering the 5 objects and their order of presentation. Follow-up studies are planned with an increased level of difficulty, to increase the variation in results and allow a more fine-grained distinction between participant groups. This line of work confirmed the initial proof of concept and paved the way for further confirmatory studies, which are currently being planned.
A second line of research looked at the effect of object knowledge on spatial memory in a screen-based paradigm. Combining previous research on the effects of object knowledge on object location memory with research on boundary extension. Previous studies have shown that ownership, or the mere suggestion of ownership, biases memory for both objects and their location. Specifically, when an object is owned, it is remembered to be closer by compared to objects that are not owned. Conceptually this can be seen as a ‘zooming in’ on owned objects compared to not-owned objects. This was compared to research on boundary extension, in which people extend on the boundaries of scenes in their memory, such that they ‘remember’ a larger scene upon memory than was originally present. Conceptually, this can be seen as a ‘zooming out’. In this line of research we combined these two, and presented participants with images of objects in scenes, in which ownership (owned or not) to objects was assigned. In the first study, we found a significant reduction of boundary extension in scenes with ‘owned’ objects. While results showed a boundary extension of approximately 5% in the ‘not-owned’ condition, this was only 2% in the ‘owned’ condition. To follow-up on this effect we planned a study with two tweaks. First, we hypothesised that the effect might be that participants assumed ‘owned’ trials were more important, so we manipulated the ‘importance’ of trials. By indicating trials might come back for another question later in the experiment, we crossed the manipulation of ownership with importance. A second level was planned to look at cultural variation, comparing Western participants with East-Asian participants, which have been shown to process and remember scenes more holistically. However, when replicating the initial findings with the Western group (importantly, the stimuli where updated as well), results did not show significant differences. The planned replication with East Asian students was therefore postponed, in order to explore whether the failed replication falsified the earlier findings or whether the absense of effects were caused by the changed experimental design. A follow-up study is in preparation.
A final novel, behavioural study design tested memory for object location in real 3D space, by contrasting object locations within and out of reach on the sagittal and the lateral axis. A TV-screen was put flat on a table, in order to manipulate object locations specific to individual participants reach – before the study participicants performed a non-related task in which they would place objects as far as they could reach on the screen. During this calibration were blindfolded as to not give them visual feedback of how far they could reach. Each individual participants’ reach was used as a baseline, to present objects within, or out of reach, and test whether reachability judgments as well as object location memory are biased towards the dominant hand. The results are currently being analised and expected to lead to a stand-alone publication.
Next to studies targeting individual differences designed and executed within the project, the researcher also pursued collaborations in additional related projects investigating individual differences in spatial communication, spatial language in bilinguals, spatial recognition in second language learning, and spatial language development during infancy. These led to 4 separate publications.
The second overarching line of studies looked at the neural correlates of object location memory. A mock study was conducted to confirm that the effects previously found in behavioural tasks with participants sitting at a table would replicate in an MRI-setup, in which participants would be lying down. An additional mock-scanner study was run to test the reliability of the designed apparatus, which reached an error-rate of 1 in 700 trials. After these two mock-studies confirmed proof of concept, we ran two fMRI studies, investigating the neural correlates of spatial memory, including topographic mapping of the SPOC area and neural correlates involved in perceiving objects at varying three-dimensional distances, as well as those associated with object location memory across these distances. Participants were lying in the MRI-scanner, with their heads propped up to approximately 40. This allowed them direct vision of objects presented on the apparatus. Data collection of two studies (both N = 30) has been completed and data analysis for these studies has commenced. While the anticipated behavioural effect of object knowledge on OLM did not emerge as expected, the collected neuroimaging data is currently being analysed and is still expected to provide valuable insights on the neural correlates for spatial memory and a building block for future research.
A second line of research looked at the effect of object knowledge on spatial memory in a screen-based paradigm. Combining previous research on the effects of object knowledge on object location memory with research on boundary extension. Previous studies have shown that ownership, or the mere suggestion of ownership, biases memory for both objects and their location. Specifically, when an object is owned, it is remembered to be closer by compared to objects that are not owned. Conceptually this can be seen as a ‘zooming in’ on owned objects compared to not-owned objects. This was compared to research on boundary extension, in which people extend on the boundaries of scenes in their memory, such that they ‘remember’ a larger scene upon memory than was originally present. Conceptually, this can be seen as a ‘zooming out’. In this line of research we combined these two, and presented participants with images of objects in scenes, in which ownership (owned or not) to objects was assigned. In the first study, we found a significant reduction of boundary extension in scenes with ‘owned’ objects. While results showed a boundary extension of approximately 5% in the ‘not-owned’ condition, this was only 2% in the ‘owned’ condition. To follow-up on this effect we planned a study with two tweaks. First, we hypothesised that the effect might be that participants assumed ‘owned’ trials were more important, so we manipulated the ‘importance’ of trials. By indicating trials might come back for another question later in the experiment, we crossed the manipulation of ownership with importance. A second level was planned to look at cultural variation, comparing Western participants with East-Asian participants, which have been shown to process and remember scenes more holistically. However, when replicating the initial findings with the Western group (importantly, the stimuli where updated as well), results did not show significant differences. The planned replication with East Asian students was therefore postponed, in order to explore whether the failed replication falsified the earlier findings or whether the absense of effects were caused by the changed experimental design. A follow-up study is in preparation.
A final novel, behavioural study design tested memory for object location in real 3D space, by contrasting object locations within and out of reach on the sagittal and the lateral axis. A TV-screen was put flat on a table, in order to manipulate object locations specific to individual participants reach – before the study participicants performed a non-related task in which they would place objects as far as they could reach on the screen. During this calibration were blindfolded as to not give them visual feedback of how far they could reach. Each individual participants’ reach was used as a baseline, to present objects within, or out of reach, and test whether reachability judgments as well as object location memory are biased towards the dominant hand. The results are currently being analised and expected to lead to a stand-alone publication.
Next to studies targeting individual differences designed and executed within the project, the researcher also pursued collaborations in additional related projects investigating individual differences in spatial communication, spatial language in bilinguals, spatial recognition in second language learning, and spatial language development during infancy. These led to 4 separate publications.
The second overarching line of studies looked at the neural correlates of object location memory. A mock study was conducted to confirm that the effects previously found in behavioural tasks with participants sitting at a table would replicate in an MRI-setup, in which participants would be lying down. An additional mock-scanner study was run to test the reliability of the designed apparatus, which reached an error-rate of 1 in 700 trials. After these two mock-studies confirmed proof of concept, we ran two fMRI studies, investigating the neural correlates of spatial memory, including topographic mapping of the SPOC area and neural correlates involved in perceiving objects at varying three-dimensional distances, as well as those associated with object location memory across these distances. Participants were lying in the MRI-scanner, with their heads propped up to approximately 40. This allowed them direct vision of objects presented on the apparatus. Data collection of two studies (both N = 30) has been completed and data analysis for these studies has commenced. While the anticipated behavioural effect of object knowledge on OLM did not emerge as expected, the collected neuroimaging data is currently being analysed and is still expected to provide valuable insights on the neural correlates for spatial memory and a building block for future research.
For these studies, a novel fMRI-compatible apparatus was designed to present real objects in 3D space within the scanner environment (see Figure 1). The apparatus can be placed over participants, like a tray table, and allows object presentation from close to the participant’s eyes to 160cm further (for each participant, the extend of reach was used as a baseline around which object presentations were centered). The apparatus allows the automatic selection of 1 of 5 objects, their presentation anywhere on the apparatus. Furthermore, participants can respond using a button-box, with which they can move the train holding the objects, to a precision of .2mm, in order to indicate the location of the remembered objects. This apparatus allows for the concurrent acquisition of behavioural and neural activation data, facilitating a more in-depth understanding of the neural underpinnings of OLM than previously possible within vision via a mirror or virtual reality setups.
During the reporting period, the project has resulted in the publication of four related manuscripts including contributions to Nature Human Behaviour, Bilingualism: Language and Cognition, Discourse Processes, and Journal of Child Language. Two manuscripts related to WP1 are currently in the review process (at Nature Human Behaviour and Cognitive Science), and one more manuscript is currently being written. Furthermore, 2 manuscripts are in preparation related to WP2. The three novel lines of behavioural research developed during the reporting time require further research, to tease apart the findings and expand on results.
A project website was created: harmengudde.com/neuracolm and the research was presented at four international conferences: the EWIC at Cambridge, UK; the NVP conference in the Netherlands; the Cognitive Science Conference in Rotterdam, and the ISCS conference in Rome. Furthermore, The researcher actively engaged with the general public through several outreach events, including the ZieZo beurs (for people with visual impairments), the Betweter festival (science festival), and the Weekend van de Wetenschap (science festival for kids). These events involved presenting research on spatial cognition and running research projects with the attendees.
This project has made significant contributions to our understanding of the complexities of spatial memory, highlighting individual variability and advancing neuroimaging methodologies in this field. The findings and the newly developed apparatus provide a strong foundation for future research into both typical and atypical memory functions, including memory disorders and aging populations.
During the reporting period, the project has resulted in the publication of four related manuscripts including contributions to Nature Human Behaviour, Bilingualism: Language and Cognition, Discourse Processes, and Journal of Child Language. Two manuscripts related to WP1 are currently in the review process (at Nature Human Behaviour and Cognitive Science), and one more manuscript is currently being written. Furthermore, 2 manuscripts are in preparation related to WP2. The three novel lines of behavioural research developed during the reporting time require further research, to tease apart the findings and expand on results.
A project website was created: harmengudde.com/neuracolm and the research was presented at four international conferences: the EWIC at Cambridge, UK; the NVP conference in the Netherlands; the Cognitive Science Conference in Rotterdam, and the ISCS conference in Rome. Furthermore, The researcher actively engaged with the general public through several outreach events, including the ZieZo beurs (for people with visual impairments), the Betweter festival (science festival), and the Weekend van de Wetenschap (science festival for kids). These events involved presenting research on spatial cognition and running research projects with the attendees.
This project has made significant contributions to our understanding of the complexities of spatial memory, highlighting individual variability and advancing neuroimaging methodologies in this field. The findings and the newly developed apparatus provide a strong foundation for future research into both typical and atypical memory functions, including memory disorders and aging populations.