Ask HN: Are there any studies about book reading retention rates vs. audiobooks?

But if it's something that's new to me, or highly technical, I'll follow up with a course, physical book, or create a study out of it in a different manner.

If I really want to study a subject I will listen to the audiobook, and then re-read it in text form.

There have been studies looking at the relative retention rates of visual vs auditory stimuli where the visual typically comes out ahead:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2667065/ https://scholarworks.iu.edu/journals/index.php/iusburj/artic... http://journals.plos.org/plosone/article?id=10.1371/journal....

However, I would posit that the main reason for this discrepancy is that memory is linked to the level of stimulation, and we typically have developed to place a far higher weighting on our visual systems, not to mention the fact that the visual signal typically has a far higher information content, linking to many existing mental structures.

Once these sort of links are removed, with the stimuli being decontextualized, the two tend to be more similar: https://link.springer.com/article/10.1007/BF00447078

The problem with audiobooks vs reading is far more constrained, and possibly environmentally determined.

One study found that visual vs auditory retention rates for a short block of text were virtually identical: http://psycnet.apa.org/record/1978-21860-001

But another again shows the superiority of the visual system: http://scholar.utc.edu/cgi/viewcontent.cgi?article=1171&cont...

But this was done in a controlled environment where the subjects' full attention was focused on the audio. In such a case, it makes sense that retention rates would be comparable, as the same language centers in the brain are used for processing the input once it is transferred from the appropriate input sense.

The big problem with audiobook retention is that we typically listen to them whilst being involved in other tasks, and there are are a myriad of studies showing that humans are inherently extremely bad multi-taskers, with high task-switching costs even in people who claim to be otherwise.

Aside from just sitting with my eyes closed and listening, I find that retention is maximized with audiobooks when there is a split between System 1 and System 2 activity (ala Thinking, Fast and Slow by Daniel Kahneman). If I am doing something like driving, using System 1, my System 2 can be fully focused on the audiobook, until something darts out in the middle of the road and then I end up having to rewind and re-listen to what I missed.

I've found that I retain a lot more of the details now that I'm listening through the books and I cannot just skim past "the boring parts". I notice the subtle callbacks to details in earlier stories. I notice the little hints or clues dropped by the author (especially in their choice of words). I think I actually enjoy this more because I get a stronger sense of the "craftsmanship" element.

However, this really requires a good reading so that you can distinguish the different voices, emotions, and details.

That said, I find it valuable to learn how a writer thinks more so than the technical details of each point they're trying to make. Right now I'm on a deep dive for an author...listening to ~100 hours of their material. It is somewhat like listening to 100 hours of their lecture. I learn different things while listening to a lecture than I do when reading.

I will stick with my paperback books for retention.

In my case, retention is much lower for listened content compared to read content. So much so that I now only listen to things which I consider entertainment. For anything technical, or that I'd want to retain longer than a few weeks I read with an actual book.

I hesitate to speak for everyone on this, but I will say that a few friends I've spoken to about it have said they see the same lesser retention in audio content.

The only way one of my kids gets through some of the "great literature" he has to read in high school is to listen to it, which I've conceded but on the condition that he follow along in reading the book.

More seriously, some of this, to me, depends on what I’m listening to or reading. Fiction or easy non-fiction audio books do allow some degree of multitasking. However this doesn’t apply to 90% of my reading. Technical material is much faster for me to actually read.

For example, I saw an interesting paper recently[1]. It’s results are important and I’m happy to have come across them, but there is no way I could have absorbed it on an audio format without taking much more time.

I read technical material like this faster because I can skim over the things I already know (like why the result is interesting, applications of the idea, complex proof steps, math background material, etc) and I can slow down for the key results I want to remember. I can go back and re-read sentences that don’t make sense to me or where I’m confused.

For technical books there is so much that I don’t usually have to learn again, I can just skip to the new stuff.

[1] The Power of Two Choices in Randomized Load Balancing, https://dl.acm.org/citation.cfm?id=504343

This, combined with having a workflow for consistently reviewing my highlights[1] has improved my (admittedly self-reported) retention significantly. There's plenty of science that supports how effective spaced repetition is for retention.

[0] https://www.amazon.com/gp/feature.html?ie=UTF8&docId=1000827...

[1] https://readwise.io -- disclaimer: I built this, but I genuinely think it's helpful here.

You can easily convert a couple and run some A/B tests

I mainly use audiobooks now for mostly fiction and narrative strong content like history, biographies/memories, or communication around domains like sales, public speaking, marketing, and relationship books.

I found a lot more interesting things about being more strategical in using the right medium(print vs ebook vs audio) depending on the nature of the content I'm consuming as well as my energy levels[1].

[1] https://juvoni.com/print-ebook-audiobook

http://www.youngscientist.com.au/wp-content/uploads/2015/02/... - "From the results, it was concluded that visual information is recalled the most effectively, followed by audio/visual and then audio information."

From my understanding, listening and reading activate the brain in different ways. Listening also takes a higher effort to turn information into memory. It seems that if you're practiced at information processing through listening (e.g. you've listened to a lot of audiobooks or are a practiced storyteller) then there's less of a difference.

Part of me thinks that reading leads to better comprehension and information recall because it's very easy to re-read a sentence versus rewinding audio. Anecdotally, I know I've been in a scenario where I was reading the end of a paragraph and realized I had stopped paying attention and was quickly able to go back and re-read it.

I do, however, find strong associations between certain book passages and the places where I was when I heard them. Particular road trips are inextricably linked to particular books and vice versa.

For example, if you read a book, chances are you're 100% focused on the book, probably sitting or laying down in a quiet room.

On the other hand while listening to an audio book you might be driving, running errands or cleaning. Most of that is probably on auto-pilot but I have to imagine trying to multi-task whatever you're doing while absorbing content is going to cause a huge hit to retention.

I think it's worth exploring the above because I find that other than first hand experience I learn best by watching videos. It could be someone talking into a camera, or just slides, but in both cases I feel like I retain information better than just reading. However in this case, I'm also 100% focused on the task as if I were reading. Most of those videos could have just audio and it's basically the same thing (comparable to audio books).

I can not imagine listening to technical books though (popular science seems to be fine, I’m listening to Walter Isaacson’s Einstein biography right now), the reader must be a fit (Stephen Fry is gorgeous); and I can only use the times when my head is not engaged otherwise (commuting, groceries, food, housework).

To those unaware: Spotify has a huge audiobook collection now. I wish they would add increasing playback speed as a feature, if readers are too slow my thoughts will wander. I also have an Audible subscription, they have a quite convincing library now.

Source: anecdotal evidence

My experience is that retention depends mainly on two factors: the amount of attention and quality of the reader. I prefer when the author is the one reading because many of them add comments and a special "energy" to the book.

I find the "car mode" on Audible very convenient to add bookmarks without disruption while I'm driving and I use at the gym as well.

Some books that I want to reference later I buy a paper-copy and mark with a highlighter on the same places I bookmarked on the Audible App.

While listening, the faster I play, the more attention I pay. I start at 2X and if I'm listening to the audiobook a second time, I can go up to 3X. 1X and 1.5X are for engaging parts or when I can't understand what is being said.

One thing I don't do is: listen while not able to pay attention. While dring it can be while going through a toll or navigating through unfamiliar streets.

With all said, the feature that I think is the most useful for retention is the Pause button. This way I can reflect on what I just heard and save the information on my brain's long-term storage.

I've found books that I thought were mediocre were much better as audiobooks because the pace forced me to really listen to and appreciate the quality of the prose.

I do agree that more technical material is more awkward in audio form and really benefits from being able to flip back and forth for reference in a real book.