Quantified book reading: the longer the better
In January of this year, I decided to try to read more books. Last Christmas (2014), usually a time for reassessing the year, I added up all the books I had read in 2014 and was a bit disappointed, it was only around 20. I recalled telling someone once that I read about 40 books a year. I guess that was an exaggeration, or maybe last year was an exception. So, for 2015 I decided I wanted to read more books, documents with hundreds of pages, not short pieces. I think that completing a book gives me a greater sense of accomplishment and is more relaxing to recall than short articles, which seem more unfinished or needy. A book becomes part of me, while essays are harder to own, too fast, like advertising, with lines of thought scattering outwards. A book tends to make a world on its own, the lines of thought going inward.
So far, I’ve read 40 books this year, about 4 books a month, double what I read last year in 12 months. I like creating new habits, and I think two main things were key to establish and maintain this one; 1) choosing to “read” audio books, and 2) having a pretty good idea of the next book I would read after the one I was working on. See other good tidbits of advice at by Austin Kleon. For a similar list in Spanish see El Pais.
I wanted to analyze several things about my book reading, so I made an excel file that included basic information about the book: title, author, when published, birth date of author, sex, audio book or “real” book, total words (according to Amazon.com) and month read. While I was completing the excel file I also wrote down more personal data about my “relationship” with the book; whether I could easily recall the author’s name (1=easily recalled, 2=have an idea but not sure and 3=forgot completely), whether I liked the book (1=bad, 2=so so, 3=liked, 4=liked a lot), and whether I wrote down notes about the book on my computer and/or my log book (hand written). Later on, while looking up each book on the web, I also jotted down whether the author appears on Google (you know, that photo and excerpt about a person’s life on the right hand side, for “famous” people), the book rank, reader scores and total number of comments on Amazon.com and on Goodreads.
Out of the 40 books, only 14 were from the 1900s, one earlier (Marcus Aurelius’ Meditations, year 156 AD) and most from the last 5 years. Only eight of the books were written by women. I realized while doing this exercise that I do not spend a lot of time choosing the books before reading. I have an idea of what I’ll read next usually based on suggestions from the book I’m currently reading, or comments from friends, stuff in the news or suggestions from web sites and blogs. Maybe next year I should try to read at least half by women? Let’s call that book parity.
I wanted to know what kind of books I say like, so maybe I can read more of those. I analyzed the data in the excel file and strangely enough my liking score was not correlated with much. The highest correlation (Pearson correlation=0.43) was with reviewing, that is, the more I liked a book the higher the chance was that I made a summary of it on my computer. But that correlation was not very high. My liking score was inversely correlated with Amazon book rank (-0.59, P=0.002), which makes sense since my liking score gets better as it gets higher and the Amazon score is the opposite. So, my liking a book does correlate with what the public on Amazom.com feels. There was no significant correlation with the scores on Goodreads, but most of those scores always seem to range from 3.5–4, so not very helpful.
I also wanted to know whether listening to audio books had an effect on my liking score or reviewing activities. I found that liking the book did not depend on format, but I did tend to write more about the book in my logbook if it was an audio book (a correlation of 0.45). I think that is because I tend to underline in real books and then write down a summary directly on the computer (read, underline and summarize lines on computer), while with audio books I write down notes on my logbook, and they rarely get onto a computer file.
What about recalling authors? No real relation with anything, not even with their presence on Google. I did a couple ANOVAs to check out statistical significance and found that the longer the book, the greater the presence on Google (P=0.08, not quite significant, but the average was 192 pages for no presence, and 324 pages for presence). Then I thought, ok, do I also like longer books? The result is Figure 1, which basically says that yes, I do like books that are longer, and for those books I also tend to write about them more in my log book. That makes sense. So writers and readers, it seems like we should both be working with longer books.
Finally, my favorite books of the year so far: Meditations (Aurelio, Marcus), El Aleph (Borges, Jorge Luis), Death And Life Of Great American Cities (Jacobs, Jane), The Educated Imagination (Frye, Northrop), The Mezzanine (Baker, Nicholson), How Music Works (Byrne, David), Freedom (Franzen, Jonathan), You Are Not So Smart (Mcraney, David), This Explains Everything (Brockman, John).