Skip to content →

2018/10/18: Finalising analysis and design

What have I been doing all the time?

It has been a while since the last post on this blog. The major reason for this is that I was busy preparing the next experiment (schemaVR3) and planning the next project.

Conerning schemaVR3: Its aim is to corrobrate the evidence that recollection and familiarity are differently involved in the memory of advantage for unexpected and expected locations. Hopefully, I will be able to start collecting data for this experiment on the 29th of October. I selected five suitble sets for schemaVR3 (here) and made sure that spread of those sets is good enough (here). For the pre-registration, I also explored the results of recollection/familiarity further (here) and run a power analysis to determine the sample size (here). The pre-registration link can be found here.

The other thing I have been occupied with is planning a project that examine the effect of hydrocortisone on the observed U-shape between expectancy and spatial memory. I will present the plan at BNA 2019 in Dublin as a pre-registration poster an exiting new format. More will follow as soon as I finished the abstract.

Below I briefly report the relevant aspects of the last meetings with my supervisor and the decisions we made.

Objects to be added as random factor in LME models

After more enquiry, we now agree that objects need to be included as a random factor to deal with the dependencies. I added this factor, which sometimes, changed the results and interpretation of the models but serendipitously never in the way that a significant quadratic relationship become insignificant. In some instance, the reverse acutally happened.

Keeping 3 AFC task in schemaVR3

As the correlation we seek to replicate with schemaVR3 is based on memory performance in the 3AFC task, we decided to keep the task. However, in contrast to schemaVR1 and schemaVR2 we will not base the selection process of the object/locations sets on the expectancy of the foils. We will take the sets we already selected (click here) and will try to find the best foils for those five sets.

What is a schema?

Journal club on Fernández & Morris (2018) During this meeting I led a journal club on the opinion piece by Fernandez & Morris (2018) in Trends in Neuroscience. In that course, we discussed the tag-and-capture hypothesis and behavioural tagging. During this discussion, we developed the idea to test behavioural tagging, which is the memory advantage that arises for content learned just before the experience of novelty. There is one paper showing a similar effect in humans (see Fenker et al., 2008), but I am not aware of any other paper showing this effect. Based on my VR experiments, I can say that most people perceive the VR experience as novel and unprecedented, which should in principle be novel and powerful enough to influence memory for preceding events. An idea therefore is to add this to schemaVR3 or to the subsequent experiments.
Furthermore, we delved into the definition of schema. The discussion was started by the appreciation that there are a lot of different definitions of schema out there (especially true for human vs. animal work), which might not be really compatible. In this discussion mainly Rik argued that a schema is an active knowledge structure that is generalised (i.e. abstract) and comes with predictions for a current situation. The difference between a schema and prior knowledge in general therefore is that the latter is latent and not activated. The representation or knowledge structure needs to be abstract enough to be applicable to a number of different episodes. For every time point in our life, there is a schema activated. Concerning the relationship of semantic memory and schema, it was postulated that schema are a subset (active part) of semantic memory, but not every semantic memory can act as a schema.

Published in open lab diary project 1 research design schema schemaVR3


Leave a Reply

Your email address will not be published. Required fields are marked *