[xpe]

Many psychological studies sound convincing but are hard to replicate and/or based on low sample sizes or poor experimental design. Seeing a plausible mechanism underlying a theory isn’t enough. So I’m writing this comment as a placeholder. Who here has checked the source studies?

Roughly speaking what is the total number of people in all studies pertaining to this claimed effect?

[SubiculumCode]

There is an absolutely vast literature studying the effects of context and context changes on many different forms of cognition, going back decades. Your posts really adds little of value, and relies on some over-broad stereotypes of the soundness of psychology research. Do some google scholar searches.

[xpe]

> There is an absolutely vast literature studying the effects of context and context changes on many different forms of cognition, going back decades. Your posts really adds little of value, and relies on some over-broad stereotypes of the soundness of psychology research. Do some google scholar searches.

In four words, the above comment is: harsh, judgmental, not curious.

Below, I hope to (a) offer another point of view about how your comment may be perceived and (b) demonstrate to you that the sample sizes and replicated studies regarding the Doorway Theory are unimpressive.

First, you made this discussion personal, and it wasn't particularly constructive. It is hard to say what goals you might have, but if your goal is curious conversation, I don't think this is the way. Do you? What is your experience/philosophy/science of productive conversation? Based only on this interaction, I wonder if you may not have thought about this very much and/or you aren't putting it into practice and/or you're taking something out on me.

Second, the word "stereotypes" gets casually thrown around. Empirically, how often would you say that using it advances constructive conversation?

Third, using this definition of stereotype: "a widely held but fixed and oversimplified image or idea of a particular type of person or thing." ... why would you think my view of this is "fixed" or unchanging? I don't see a rational reason for seeing that based on what I wrote.

Fourth, for people that think in a Bayesian way -- and I think such a model is fairly useful in modeling how people actually operate -- we all have priors. Please do not accuse me of stereotyping when I'm only sharing a prior. One key question is what we do with them as we gain more information.

Fifth, on what basis would you validly say that my prior regarding my confidence in psychological studies is an "over-broad stereotype"? I've tried, but I don't think you can. You can disagree with my prior -- that's fair game. And we can talk empirically and rationally about how/why we have different priors.

Sixth, your comment shifts away the context I intended (I hope it was clear given the context, but maybe it was not) and criticizes a straw man. You wrote "There is an absolutely vast literature studying the effects of context and context changes on many different forms of cognition, going back decades". Perhaps this literature is largely sound and replicable. But that's not what I was referring to here; I was discussing the Doorway Effect in particular. From what I've seen so far, there were two early studies, both at Notre Dame, consisting of about 90 people in total. This alone certainly isn't enough in my mind to give high credence to the results. At least one subsequent attempt to replicate was mixed. So, please don't tell me to "do some google scholar searches". That is rather presumptive.

Would you like to continue to discussion on a better foot going forward?

[SubiculumCode]

I'll add further an apology. I was being rude, and I am sorry for that. It is easy to start responding in the way I did after my field, populated with many serious, careful, and brilliant researchers, gets regularly dismissed flippantly with such things as: "psychology isn't a real science, psychology isn't replica-table, psychology researchers don't know the scientific method, don't know statistics, are just p-hackers, etc etc etc, when you know that is only true for a subset of scientists, while the best tend to be experts, or collaborate with experts on the various topics needed to conduct serious science. So, please just understand where I am coming from, and do please accept my apology for my rude and presumptive demeanor.

[xpe]

Thank you; I appreciate this. I understand that too many people flippantly dismiss or don't have the ability to properly appreciate the hard work of doing science.

[SubiculumCode]

*"Many psychological studies sound convincing but are hard to replicate and/or based on low sample sizes or poor experimental design."*

Many physics, chemistry, and computer science studies suffer from these as well. Science is hard, and there are a lot of constraints on research. Typically the first finding is made with limited funding, meaning limited ability to do the study right. Also, usually there is no perfect study, just a balancing of different strengths and weaknesses. That is why in psychology and neuroscience research, we talk about converging evidence. 'Converging' here means evidence employing multiple modalities, measures, and procedures. This might mean combining evidence from animal and human models, molecular and neuroimaging, etc. Its messy stuff, and every approach has tradeoffs.

*Seeing a plausible mechanism underlying a theory isn’t enough. So I’m writing this comment as a placeholder. Who here has checked the source studies?*

You are writing a comment as a placeholder for what exactly? So far in your comment, it just seems like an excuse to bash on Psychology research, which is too often a cheap thrill here on HN, and usually from someone who has read little serious psychological research save the most flashy in-the-news findings.

*Roughly speaking what is the total number of people in all studies pertaining to this claimed effect?*

This appears to be a logical question, and within the immediate context, understandable. However, the more informed context would understand that the so-called "Doorway Effect" is just a specific example of a more general and prevalent cognitive phenomenon: The role of spatio-temporal context in memory and cognition.

You ask for specific sample sizes from studies that use tasks that look at the effect passing through doors, and well there may be just an of a few hundred.

But, I being somewhat knowledgeable int he field (PhD in memory research) would instead ask: What is it about doorways that could cause such an effect. Ah yes, it involves a change in the spatiotemporal context. What is the evidence that context changes impose costs to memory function? Well: https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=cont...

So many, for so long.. Its a HUGE literature. Point is, its not about doors, and this is just one finding in a long list of similar findings that created CONVERGING evidence of a phenomenon.

[xpe]

A great, thought-provoking comment. This makes me quite aware that I don't yet have a great theoretical/conceptual foundation for how Bayesian reasoning for one particular experiment links up with experiments that are generalizations of it. In case this sounds too vague, what I mean is this: if a generalized theory has considerable evidence, what bearing does it have one more targeted and narrow experiments? This might be clear or even obvious, but I'm not recollecting written examples of how to run the numbers.