Category Archives: Information Resources

Last week, in response to some data about paper book holdings in Nebraska libraries, the Dude asked the question: So what? Today will be an attempt to provide some context. If we look at the title to last week’s NET line graph, Bill Kelly added the following label to the top of the graph: SHELF LIFE: Fewer Library Books, and of course the assumption is that libraries have fewer paper books, presumably in response to less demand, and more eBooks. And over the time period in the NET graph (2010-2014) that’s true. But as was mentioned last week, if the data is expanded back to 2000, there was an actual increase of 1.19% in paper book holdings in Nebraska libraries during those 14 years, but the trend over the past few years is in fact lower amounts of physical items and increased amounts of electronic items in libraries.

The eBook conundrum is a difficult one, and analyzing data for library holdings of eBooks is hard to illustrate. Why? Well, a few reasons, including the fact that eBooks aren’t technically purchased by libraries, they are leased. And those lease terms vary by publisher (both the duration and the price). Some publishers still don’t make electronic copies of their books available to libraries at all. Perhaps that is the deeper story to something like SHELF LIFE: Fewer Library Books; namely how the eBook lending platforms for libraries are much different than the traditional right of first sale traditional paper model. Notwithstanding all of this, it might be more productive to look at trends about the actual use of the items in the library collection. And even that might be misleading; a library could contain a bunch of crap that no one wants to check out, but for one reason or another has never been weeded. When a new go-getter takes over the library, perhaps all this lower demand stuff gets weeded and replaced with stuff that actually will be used. The end result to that would be lower numbers of paper books and VHS tapes (assuming those are the ones that would be weeded), and increased circulation of other things like light sabers and power tools. The chart, above right, demonstrates the total circulation of items in Nebraska over the last 14 years (incidentally, while the one above has the Y-axis at zero, it doesn’t look remarkably different when the Y-axis is started at 5 million).

The reality is that overall in the state, circulation of electronic items went up about 80% from 2012-2014, and circulation of physical items declined 6.42% during the same time period. Much of this is due to the fact that more libraries got involved in consortia (such as Nebraska OverDrive) that allowed more communities (and thus overall more people) to access materials. Keep in mind that, notwithstanding the recent decline of physical items held by Nebraska libraries, over 12 million physical items did circulate in 2014 (more than 15 times the number of electronic items used). Therefore, when looking at these numbers, many times the debate is framed as an either/or proposition when it comes to the physical and electronic in libraries (such as part 5 of the NET series – Will a Nebraska Community Tech Center Force Us To Consider Libraries Without Books?) and ultimately that’s probably not the question to ask (at least the data doesn’t support it).

In actuality, libraries are offering new things and expanding services in addition to offering the traditional ones. As the chart second from the top illustrates, program attendance has gone up during the 2000-2014 time period. Additionally, the number of programs offered has risen at a rate fairly consistent with program attendance, and annual library visits have remained fairly steady for a number of years at around 8 million per year. The real questions (and the NET series tackled some of these) are how libraries are adapting to meet the needs of the community, by either expanding the offering of things in the collection, programs, and spaces (which admittedly is broad ranging, and could mean things like meeting rooms, community learning spaces, or collaborative makerspaces). And that is why the real question should be framed by an “and” instead of an “or”. Shaka.

Today’s post will focus on line charts (or graphs, if you prefer). The Dude thought about this in response to the graph (second from the top, on the right) that NET posted on its website to supplement the recent 5 part LIBRARIES: The Next Chapter series. At first, the Dude took note of the Y-axis of the NET chart (it’s the one with the books and readers in the background), remembering some rule that says to always start these charts (and bar charts for that matter) with 0. So today’s post will take a look at the line chart and axes, and next week’s post will most likely take a closer look at this data and something called context. Note that the data from these charts is from the annual public library survey.

In 1954, Darrell Huff wrote a book called How to Lie with Statistics. One of the main takeaways of Huff’s book is the notion that bar and line charts should always start the Y-axis with 0. Anything other than 0 is a fabrication or at least a misrepresentation, according to Huff. The Dude created the first chart, top right, which illustrates the same data that appears on the NET chart. The only difference is the Y-axis starting at zero on the top chart, and 5.5 million on the bottom chart. Both charts illustrate an overall decline of 5.67% in paper book holdings from 2010-2014, or a decline of 6.56% from 2010-2013, and an increase of .95% from 2013-2014. Taking into consideration the differences in both of these charts, which one is a better illustration of the data? Well, maybe neither one of them.

The Dude is a huge fan of data statistician Nathan Yau. If you haven’t seen some of his data visualizations on FlowingData, check them out. So for this general question, namely, should the Y-axis always start at zero, the Dude consulted Nathan Yau. Ironically, Nathan’s FlowingData website had a post on the main page about this very topic, and if you have a chance, the Vox video he links to in his article is well worth the two and a half minutes. The gist is that the Y-axis need not always be zero, but the most important thing, as the Vox video illustrates, is context.

Originally, the Dude admits that he thought that the NET chart misrepresented the decrease in paper books held by Nebraska libraries (by illustrating a much larger decrease than actually occurred), but isn’t the increase illustrated from 2013-2014 just as misleading, since it tends to show a sharp and somewhat dramatic increase from 2013-2014 when in actuality the figure is less than 1%? Also, does the first graph at the top fail, in that it doesn’t show any change at all, or very little? If we expand the data back to 2000, we get a different illustration of the trend (see third and fourth right, with two different Y-axes). The real question here, generally speaking, after looking at the data, is the greatest of all philosophical questions: So what? That’ll hopefully be the topic for next week. Shaka.