Archive for October, 2013

Making Book 2: Technical Standards

Thursday, October 31st, 2013

Until I read the final page of MARC for Library Use (while preparing the first in this series of memoirish posts), I’d forgotten that KIPI contracted for another book even before the first one was publicly available.

That book was Technical Standards: An Introduction for Librarians. It appeared in 1986—a 299-page 6″ x 9″ hardcover and paperback. (That’s actually about the same length as MARC for Library Use, around 100,000 words in each case: more smaller pages.)


For reasons that still escape me, the first ALA or LITA committee I was formally involved in was TESLA, the Technical Standards for Library Automation Committee, originally part of ISAD (Information Science and Automation Division, which changed its name to LITA). I started attending the meetings in 1976, and served as a committee member from 1978 through 1982 (chairing the committee from 1980 to 1981).

Beyond that, my formal involvement in technical standards was pretty minimal (until somewhat later, when I agreed to be the founding editor for NISO’s new ISQ, Information Standards Quarterly, which I did from 1989 through 1991). For a while, I was RLG’s “alternate representative” to NISO (the National Information Standards Organization, Z39, which did and does prepare technical standards in library and related areas—it’s an accredited ANSI standardization organization)—but that didn’t mean much, because RLG’s primary representative, Wayne Davison, who was very active in NISO. (RLG had a long history of supporting technical standards; it was also a founding member of the Unicode Consortium.)

I had, however, published articles on technical standards—one in Library Trends in 1982 and another, a column arguing against a proposed standard (a Standard Library Patron ID), in LITA Newsletter in 1985—but by then I was also working on this book.


Sandra K. Paul’s foreword to the book begins “Standards aren’t sexy.” True enough, which may be why there wasn’t much in the way of approachable literature explaining technical standards—especially from a librarian’s perspective, as opposed to that of a scientist or engineer. With encouragement from some key people, I tried to improve that situation.

RLG helped by permitting me to use a wonderful anecdote about precision and technical standards—although it wasn’t so wonderful at the time. RLG used to produce millions of catalog cards—remember, this was the early 1980s! In 1982, RLG moved from an IBM line printer to a Xerox 9700 laser printer, using prepunched sheets of card stock that held four cards each and were guillotined after printing. The people ordering the cards assumed a 3″ x 5″ card size. And RLG members/users started complaining: The cards didn’t fit.

They didn’t fit because the standard for library cards (Z85.1, one of few library-related standards not bearing a Z39 number) didn’t specify 3″ x 5″: it specified 75mm x 125mm, and that’s actually been the size of catalog cards for a very long time. 75mm x 125mm works out to 2.95″ x 4.92″—which means that well-made catalog card drawers wouldn’t quite hold the 3″ x 5″ cards. RLG ordered a new set of card stock and reprinted the faulty cards. It was an expensive lesson, but a useful one.

The book covered quite a few aspects of technical standards in general—among other things, noting the number of technical standards you’re likely to benefit from in a typical day—and considered varieties of “standards” other than formal consensus technical standards, a discussion I continue to view with pride. I discussed motives, implementations, problems, dangers, the standardization process and standards organizations, with some emphasis on NISO. I discussed each Z39 (and Z85) standard in existence at the time and a handful of others.

The book did well—certainly not as well as MARC for Library Use, but well enough that KIPI invited a second (revised) edition down the road. More on that in a later post.

This was another book that required a number of trips over to UC Berkeley, mostly to the Engineering Library with its extensive set of published standards. As always, quite a few people helped with the book—reviewing drafts and providing additional insights.

Crawford, Walt. Technical Standards: An Introduction for Librarians. Professional Librarian Series. White Plains, NY: Knowledge Industry Publications, 1986. ISBN 0-86729-192-3. ISBN 0-86729-191-5 (pbk.)

50 Movie Gunslinger Classics, Disc 6

Wednesday, October 30th, 2013

Savage Journey, 1983, color (for TV). Tom McGowan (dir.), Maurice Grandmaison, Richard Moll. 1:36.

Since this set’s already demonstrated that “gunslinger” means “any movie with a gun in it,” I suppose a 95-minute chunk of propaganda for the Mormons is as suitable as anything—and that’s what this is. It leads us from Joseph Smith being tarred-and-feathered some time in the early 19th century through the many tribulations unfairly suffered by the always-good, always-just, never-vengeful, always-united Mormons (and from this movie’s perspective, Joseph Smith and his buddies destroying the printing press at a Nauvoo newspaper that said bad things about him is fully justified and proper), to the promised land in Utah, which somehow becomes “1,500 miles from the nearest food supplies” when locusts attack. (Didn’t know it was 1,500 miles from Utah to any other part of civilization in 1847? Read up on History According to Savage Journey!)

That said, it’s not a terrible picture. Even after it was obvious that it was an entirely one-sided simplification of the history of Mormonism, Smith and Brigham Young, I found it interesting enough to watch all the way through. (It never occurred to me that Richard Moll was Bull on Night Court; he comes across as a slightly wild-eyed prophet as he portrays Joseph Smith.) I’ll give it $1.00.

Savage Guns (orig. title Era Sam Wallash… lo chiamavano ‘Così Sia’ or His Name Was Sam Walbash, But They Call Him Amen). 1971, color. Demofilo Fidani (dir.), Robert Woods, Dino Strano, Benito Pacifico, Amerigo Castrighella, Simonetta Vitelli. 1:28.

I have mixed feelings about this spaghetti Western—and make no mistake, that’s what it is. On one hand, it’s got an interesting score, lots of scenery, action sometimes so “natural” in pace that I used the 2x viewing mode to get through one excruciating “French singer” ballad and one boxing match faster, and cartoon violence. Oh, and it’s sort-of widescreen. My guess is it was filmed in very widescreen mode (based on credits missing parts of the first and last letters), then trimmed—but not to 4×3, rather to 16×9 (widescreen TV) mode. It’s not an enhanced DVD, so you’re losing some resolution, but it zooms nicely to fill an HDTV screen.

And there is a plot of sorts. A gang busts into a saloon, wearing partial masks, confronts the barkeep, forces him to drink tequila pouring out of a barrel they shot into (barrels of tequila in the Old West? why not?), then shoot him and everybody else in the bar, afterwards burning it down. Except that one guy (Wallash or Walbash) was shot in the arm, fell under a table, and managed to escape. The rest of the picture consists of him hunting down and killing a couple of dozen gang members and, eventually, the boss man, Mash Flannigan (or Mash Donovan). (Along the way, we see a flashback with him as a child, in which his father and mother were gunned down in their home—for no apparent reason—by a gang that must have fired 70 or 80 shots to kill two people. It’s The Gang That Couldn’t Stop Firing.)

But the logic of the plot is so bad as to almost defy belief even by spaghetti western standards. Right after the opening scene, the evil honcho tells his gang that this sends a message to assure that nobody will ever rat on him again to the sheriff (which you’d think he would have sent more efficiently if he shot the barkeep but not every witness)—and then, as soon as he learns somebody may have escaped, he says “but if it’s not a bounty hunter, you can be sure he’ll go right to the sheriff.” Ummm… Later, a bunch of the gang surround the Lone Hero and beat him senseless—but don’t kill him. Still later, this clown who’s ridden off with a bullet wound and been robbed of everything at least once seems to have not only unlimited funds (and guns and ammo) but the wherewithal to, overnight, acquire a dummy U.S. Army paywagon with a hand-cranked Gatling gun and two wax dummies dressed in Army uniforms. Oh, and the gang—which, no matter how many are shot—always seems to be as big as it needs to be. But nobody in the gang finds it suspicious that this U.S. Army paywagon has two drivers and no guards riding in front or in back. Never mind the villain’s girlfriend, who the hero’s confident he can instantly turn in his favor, apparently correctly. The whole thing almost appears to have been written randomly. (I didn’t know bar girls got migraines—and called them that—in the Old West. Come to think of it, it can’t have been that old west—in the boxing match, the challenger is introduced as having won medals in 46 states, The 46th state was admitted in November 1907.

One IMDB review calls Demofilo Fidani “the Italian Ed Wood.” I can see why. The song by the French chanteuse is remarkably awful in every way; there’s an introduction of three major killers partway through—but those killers, not part of the regular gang, are never seen again; and… oh, never mind. The musical score is quite good. I find it interesting that neither IMDB nor the reviewers can agree on the hero’s last name.

It’s also not a great print. In the end, I can’t muster enough enthusiasm to give it more than $0.75.

Death Rides a Horse (orig. Da uomo a uomo or From man to man), 1967, color. Giulio Petroni (dir.), Lee Van Cleef, John Phillip Law, Mario Brega, Luigi Pistilli, Anthony Dawson. 1:54.

Reviewed in June 2010 (C&I 10.7). That review:

Remember the blue-eyed blind angel in Barbarella? What if he was a 21-year-old whose family was slaughtered (after his mom and older sister were raped) and house burned down 15 years earlier by a truly evil gang—one of whom saved him from the fire? And he became a crack shot, presumably planning revenge sometime? Now mix in the ever-stoic, ever-slightly-sardonic Lee Van Cleef as an outlaw just emerging from prison after a 15-year sentence, after he’d been sold out by the gang he thought he was part of—and he finds that some of the gang members are now Highly Respected Citizens. Throw in a Morricone score with singing that’s either supposed to be incoherent or is marred by a poor soundtrack—oh, and a Mexican village so suppressed by an outlaw gang that dozens of them won’t rise up against four of the gang left to guard a million-dollar theft.

There you have it: The seeds for a movie that combines vengeance and revenge, generational (and style) conflicts (Ryan, Van Cleef’s character, calls Bill, the younger one “kid”; “Grandpa” is the responding epithet), suppressed memory, lots of trick gunplay and not-so-trick gunbattles, truly bad bad guys and the gray Ryan and more. Law does a fine job as a hate-filled but naïve young sharpshooter; Van Cleef is, well, Van Cleef (after just two movies, I see why spaghetti western aficionados hold him in high regard.) It’s a solid spaghetti western, the print’s generally fine, and even with the muddy score I’ll give it $1.50.

Riders of Destiny,

John Wayne as a singing cowboy? Singin’ Sandy, that is, the notorious gunslinger known across the states—except he’s actually an undercover Federal agent. (And his primary song, done repeatedly in a robust baritone, is about blood and death.)

He encounters a sheriff who’s been shot in the back and saves the sheriff. Cut to…he encounters a scene in which a woman on a horse has the horse shot out from under her by stagecoach drivers…who assume she’s a highwayman (and she did in fact rob them, because they were carrying money meant for her father, and the weekly money loads were somehow getting robbed every. single. week). He saves her.

This all gets into a situation where the evil owner of a land and water company holds water rights to all the water in a valley—except for this woman’s dad’s ranch, which has its own well. The slick villain is trying to buy out the other ranchers for $1 an acre, or will quadruple the price of their water. Meanwhile, his own people are robbing his own stagecoaches and passengers…

Anyway, Singin’ Sandy concocts a quick scheme that saves the day for all concerned and, of course, gets the girl. Wayne is young, the movie’s a classic cheaply-done B programmer, and I guess if you like Wayne at all it’s worth $1.00.


Tuesday, October 29th, 2013

On one of the social networks I sometimes frequent–I think, although it could have been email–somebody asked me:

“If OCLC offered you a consulting or part-time telecommuting gig, would you accept?”

The answer’s simple:


Well, maybe that’s too simple. Assuming that:

  • It was something I felt I was qualified to do
  • The compensation seemed appropriate
  • I didn’t report to either of a couple of specific people in Northern California,

then, yes.

Chances of having such an offer: Probably not very large.

Bad feelings I have toward OCLC itself: Zero.


Making Book 1: MARC for Library Use

Tuesday, October 29th, 2013

This is the first in a (possible) series of memoirish notes on how my books came to be.


MARC for Library Use was not the first book length manuscript I wrote.

That would have been the study of newspaper coverage of the Free Speech Movement, which I wrote (I think) two or three years after FSM itself—thus, in the mid-1960s. On an electric typewriter. Doing nearly all of my research from roll microfilm of daily newspapers. Which, as other oldsters might imagine, left me ever-so-fond of microfilmed newspapers and those lovely manual readers.

What happened to that manuscript? I have no idea. There were two copies—an original and a carbon copy. (Where, exactly, would a penurious student go to get a 400-page manuscript copied in, say, 1966? And how would he afford it?) I submitted the manuscript to the University of California Press. Which rejected it. I would have submitted it elsewhere, I think, but in the meantime loaned it to a “friend” to read. Who disappeared…with the manuscript in his possession. The carbon copy somehow wandered as well.

I think it was a pretty good project. I may be fooling myself. Anyway…


I started developing MARC-based software in 1972, the year I moved from UC Berkeley Doe Library’s circulation department to the Library Systems Office. (USMARC goes back even farther: MARC II originated in 1968.)

When I moved to the Research Libraries Group (RLG) in 1979, I continued to work with MARC—unsurprisingly—and in 1980 became Product Batch Group manager, in charge of the behind-the-scenes work that produced all products from RLG member and user cataloging (except for catalog cards—my group made sure that the intermediate steps worked, but the phenomenally complex and flexible card production software was in another group). Most Product Batch work was in pure MARC—directory, leader and all. I retained one key piece of code from Berkeley: A very compact PL/I subroutine to extract desired fields or subfields from a MARC record with minimal overhead. (Back then, programmers spent a lot of time working on efficiency!) I would note that UC’s statewide library systems group, UCDLA, also borrowed that routine.

I was aware that library vendors, especially smaller-system vendors, had a tendency to call systems “MARC Compatible” that could not, in fact, import and export MARC records on a generalized basis, and that there was a need for better understanding of MARC itself. At one point, my group hired a library school graduate who’d taken a course on MARC and who had his syllabus from the course (since there were no textbooks). I read the syllabus and was horrified: Much of it was wrong or oversimplified (e.g., assuming a certain limitation was part of the format because one interactive system had that limitation).

The field needed a proper book on USMARC. I started talking up the idea with people at the Library of Congress, the people who were the actual experts. (I started serving as a liaison from RLG to the USMARC advisory group in early 1981 into 1987, the last two years as a MARBI committee member from LITA. I became acquainted with Henriette Avram in that role.)

I got nowhere with the effort. And finally said, “I’ll do it myself.” With considerable trepidation.

My vague recollections involve about 12-18 months of research and writing (fortunately, I had my first personal computer by then: A Morrow MD2 with no hard drive but two diskette drives, one for the OS and software, one for data: it had a honking big 128K of RAM and was, I believe, a Z80 CPU). I remember quite a few trips back to Berkeley to work with material in the Library School Library (since shut down, along with the library school). Eventually, I had a manuscript—and got reviews at various stages from several people at RLG, colleagues at OCLC and WLN and Penn State, and some of the actual experts at LC, including Henriette Avram. (Ms. Avram also provided a foreword.)

I submitted the manuscript to the foremost library publisher. They didn’t know what to do with it. They suggested rewriting it as a cataloging manual. They dithered. Eventually, I told them I was offering it elsewhere. Which I did—to Knowledge Industry Publications, Inc. (KIPI). The acquisitions editor there also wasn’t quite sure what to make of it—but thought the topic might be important and took a chance.


In 1984, it appeared—a 222-page 8.5×11″ hardbound and paperback (back then, some library publishers did both versions).

It succeeded better than KIPI had expected and probably better than I had. Of all the books I’ve written (without co-authors), I’m pretty sure it’s the best-selling…and quite possibly the most important. I know that at least one major library automation vendor purchased a copy for each of its salesfolk and told them to read it. I know that within a year or two, companies claiming MARC compatibility had MARC compatibility.

It is a book of which I am proud.

Crawford, Walt. MARC for Library Use: Understanding the USMARC Formats. Professional Librarian Series. White Plains, NY: Knowledge Industry Publications, 1984. ISBN 0-86729-120-6. ISBN 0-86720-119-2 (pbk.)

A hi-def “tragedy” in four short acts

Tuesday, October 29th, 2013

Act 1: 2008

Pioneer introduces the Kuro line of plasma TVs, which offer the best picture quality of any flat-screen HDTV ever made (with the possible exception of the 12″ $2,500 Sony OLED TV, and that’s too small to count). The Kuro TVs use a number of special technologies, including panels that eliminate the air space between the plasma pixels and the glass front, which cost extra but make for a superb picture.

Act 2: 2009

Pioneer exits the TV business. End of the Kuro. It sells several patents to Panasonic, and it’s possible that some engineers moved there as well.

(By the way, the Wikipedia “article” on the Kuro is offensively wrong–saying that Panasonic is the only other significant TV manufacturer involved in plasmas omits those tiny little nobodies Samsung and LG.)

Act 3: 2013

Panasonic introduces the ZT60 series (TC-P65ZT60 and TC-P60ZT60, where the P number shows the diagonal size in inches). It involves several technological advances–including a panel with no air space between the plasma pixels and the glass front.

According to a review in the October 2013 Sound & Vision, done by a reviewer who still owns a top-of-the-line Kuro set and included a side-by-side comparison, the ZT60 is essentially the equal of the Kuro in all key areas of image quality. In other words, it’s the best HDTV (at least at 1080p) you can buy.

Act 4: Later in 2013

Panasonic announces that it’s getting out of the plasma TV business.

And, I guess, we wait for OLED to make it to big-screen and reasonable-price.

(About the only weakness of the ZT60 was that it’s not enormously bright in 3D mode. To which most of us might say, “who cares?” )

How many years ago did downloaded music pass physical purchases?

Tuesday, October 29th, 2013

There’s the question.

Based on way too many articles and posts and other stuff, you’d have to assume it happened quite a few years ago, that CDs and the like have been pitiful ghosts for quite a long time.

That’s partly because digital gurus have become very adept at phony equations–e.g., assuming that a single-track download equals a full-album purchase, then shouting about number of downloads exceeding number of albums (in physical form) sold.

And, of course, because The In Crowd abandoned physical media far ahead of anybody else.

So what’s the answer?

According to Sound & Vision, at least, the answer is…


Last year.

The first year in which revenues for downloaded audio exceeded revenues for physical audio media.

Which, by the way, only means “nobody buys CDs [or LPs] anymore” for that special definition of “nobody”: “none of my friends” or “none of the in crowd” or, in this case, “slightly less than half of all music purchases, therefore essentially vanished.”

Even that magazine used “eclipsed” to mean “exceeded by at least 1%,” which is far more dramatic.

(Other sources say it was 2011, depending on what’s included: To wit, combined revenue for digital downloads and streaming services passed 50% of total music revenue in 2011. But that’s the earliest.)

Huffington Post’s “coverage” of 2012 numbers was typical: The headline says the numbers prove that, among other things, “CDs are Dead.” Because they represented only 40% of total music revenues in the U.S. in 2012–40% is dead. Makes life simple, doesn’t it?

Oh, by the way: Globally, 57% of the music business is still physical (mostly CDs, although LPs are a slowly growing field).

And for album purchases, CDs and LPs still considerably outweighed downloads in 2012–in numbers, 198 million to 118 million.

What? The U.S. saw nearly 200 million CDs and LPs sold in 2012? But physical music is dead!

Apparently some people didn’t get the memo.

How Not to Be the Expert

Monday, October 28th, 2013

Some tips for those who really, truly want to avoid becoming known as The Expert on any single topic (or The Guru, or The Obvious Speaking Choice…) while being professionally active.

These tips come from decades of experience.

1. Don’t specialize

You’ve prepared an important article or book or blog post on a significant topic in your field?

Time to try something else!

Delving deeper into that topic, refreshing your work for newer audiences, or—worst of all—showing how that topic applies elsewhere (even if it’s stretching a point): Don’t do that. You’ll wind up on the speaking circuit, in demand, known for your expertise.

Repeat as necessary: If you do important work in two fields, you may still wind up known as The Expert on one or both.

Better to change topics frequently. Mark Lindner owns the phrase “habitually probing generalist,” but that’s the general idea.

2. Don’t broaden your exposure on a single topic

You’ve done that significant piece—call it a book, just for fun.

The obvious next step would be to do a related article, and maybe propose some speeches on the topic.


This is related to Rule 1, but not quite the same. It’s also related to Rule 3:

3. Don’t propose speeches on your topic(s)

If you show up at CIL/IL, Charleston, state library conferences, ALA, ASIST speaking on your topic, you’re likely to become known as The Expert even if you have other topics.

If somebody really wants to hear from you, let them come to you: Don’t go looking for trouble!

4. Don’t go on the speaking circuit

If you’re following Rule 3, you’re halfway there—but if you’ve made the mistake of doing a couple of good speeches or papers on your topic, you may find a stream of speaking invitations coming in.

If you accept as many as you can plausibly handle (and if your workplace favors professional activity), that can wind up being quite a few…maybe to the point where you’re on the speaking circuit.

Set an annual limit. (I used eight trips or ten speeches a year. That seemed to work effectively. Depending on the situation, you might still consider that being on the speaking circuit—four speeches might be a better limit. Don’t worry: If you’re following the other rules, you won’t have to turn down invitations after two or three years—they’ll shrink on their own.)

5. Don’t act as though you’re The Expert

Rule 5 may be key.

The expert makes sure that her knowledge is available for interviews, etc., and pipes up whenever somebody posts or says something related to his topic.

6. Prefer precision to hyperbole

Avoid “all” when the facts say “most” and “most” when the fact say “some.”

Avoid speculation about certain futures when you really don’t have much basis for such speculation.

Never say “inevitable” unless you’re talking about mortality.

Don’t confuse anecdata with studies, and be aware of the limitations of most studies.

7. Avoid the bleeding edge

Focus on topics that need further exploration and explication, rather than the Hot New Topics.

It’s particularly useful to do something deep and comprehensive at roughly the point that an area is becoming irrelevant or obsolescent. (I would say obsolete, but that’s really tough…and you might become the Expert on curiosities of the past.)

If you are compelled to look at The New, try to make it something people don’t really care much about.

8. Be an introvert

This is a valuable addition to all of the tips above; it will help you to avoid the spotlight.

Where are #9 and #10?

To be a proper listicle, this post needs to have at least 10 items.

A proper expert would always find a couple more things to say, if only by repeating an earlier rule with slight rewording.

But, what the hey…

9. Don’t create or promulgate infographics

What more need be said?

10. Understand your data, and make sure your readers get plenty of it

Numbers! Librarians love numbers! You can never have too many numbers!


These rules have stood me in good stead, as evidenced by the fact that, after 16 (or so) professionally published books, half a dozen (more than that) self-published books and several hundred articles and columns, I am the recognized Expert on…nothing.


Maybe, possibly, Gaia willing and the creeks don’t rise: A multipart discussion of how some books (all on topics about which I am not The Expert) came to be written.

Unless, of course, I decide to read Crime & Punishment instead. (Then again, maybe not…)

Quick update on deadlines and books

Saturday, October 26th, 2013

I had said that, if no copies of $4 to $1 were sold (after October 1) by October 26, I’d drop the current discount on the paperback version.

I’m delighted to say that a copy was sold this week. (I’d be even more delighted if several copies were sold, or if there was finally a site-license sale, but one copy is better than none.)

So the discount will stick around for another month, at least.

I should also note that five copies of Your Library Is… were purchased this week. Since all the sales showed up after I posted something that was not really an ad, but a second descriptive post, at PUBLIB, I’m guessing there’s a direct relationship. And I’m hoping that some of those who buy it find it interesting and worth mentioning, since word of mouth/word of text is definitely the best way to get this into more hands.

[October so far has seen three more sales of The Big Deal and the Damage Done, but no site licenses. That book wasn’t endangered in any case, and a newer (but briefer and more expensive and not self-published) version won’t be out until next summer.]

Still absolutely no sales of The Compleat Give Us a Dollar (either volume) or The inCompleat Give Us a Dollar… which may very well disappear pretty soon.

Rather than individual links for these, since most have multiple versions each with a different URL, I’ll just say: go to and search for the title or Walt Crawford.

Self-publishing reality check: A slight correction

Tuesday, October 22nd, 2013

Turns out I’d missed 33% of sales of Your Library Is… and $4 to $1…

To wit, as of September 19, 2013 (and as of October 21, 2013, for that matter), each of the two books has sold three paperback copies (and no ebook copies), not two.

That’s 50% more sales!

By the way, The Compleat Give Us a Dollar… and The inCompleat Give Us a Dollar… have maintained their sales levels: Zero in each case.


The strongest correlation between circulation and spending

Tuesday, October 15th, 2013

There’s a lot of interesting data in $4 to $1: Public Library Benefits and Budgets, Volume 1, Libraries by Size (currently around $16 paperback and $8 ebook or $32 site license, if you buy it by Friday, October 18, 2013)–data that can help you understand the picture of public libraries across the country as well as how your own library measures up.

One key figure is the correlation between circulation per capita and spending per capita: Libraries that spend more do more.

The correlation is strong (that is, the Pearson’s Coefficient is at least 0.50) for all size categories of public libraries and for the nation as a whole (0.59).

Where is it strongest?

Among 828 libraries serving 40,000 to 99,999 people (legal service area population, not registered borrowers).

As you’ll see on page 150 of the book, that correlation is 072–a very strong correlation.

[Which libraries have the lowest correlation? In fact, only one size category falls below 0.61: The very smallest libraries, those serving fewer than 1,000 potential patrons, with 0.51 correlation. That’s not surprising–with such small patron groups and correspondingly small budgets, you’d expect a wide range of results.]