Reporting from the Front Lines
As usual, I need to weed my parts of the nonfiction collection. I could weed based on condition, which would allow me to clear a little space by pulling books off the shelf that clearly shouldn't be there because they're falling apart, or hideously ugly, or both. But what I really need is to make enough space on the shelves for the new books I've been buying over the past (cough) undisclosed period of time since I last really weeded, which means that I have to get serious. For that, I want a report.
I suspect that I care about statistics more than a lot of my colleagues. We do use statistics regularly during the course of our jobs; as reference librarians, we keep a daily record of our interactions with patrons, rather than recording stats once a month or once a year and extrapolating. Even so, I don't think many of my co-workers run reports as regularly as I do or are as familiar with the reporting software. Does this make me a stat-crazed librarian? I'm not sure, but running reports does provide me with a concrete starting place for a daunting project like weeding the 940s. While I'm using the data for weeding, I usually also come across information about the collection that may not be apparent from looking at the shelves. It would be a rare occasion that I generated a list of books in my area of the collection and weeded everything on that list--instead, I try to use statistics to pick off the weakest members of the herd. They should not be used on their own, but with a librarian's accumulated knowledge about his or her part of the collection, the needs of patrons, and the library's collection development policy, statistics can be an incredibly useful tool.
The trick to running reports for the purpose of weeding and collection maintenance is to approach with the goal of getting manageable chunks of data, but also to overestimate what you think you'll need so you don't have to re-run the report because you forgot a statistic that you were interested in tracking. I like generate more data than necessary about the section I'm interested in weeding and then sort and hide and cut until I get exactly what I need to weed effectively. My strongest advice: Do not try to weed the whole collection at once! Your end result should be no more than you can push around on two book trucks, because more than that will take too much time to sift through and may cause a staff rumble as you monopolize the departmental trucks.
Statistics to Export:
The Basics: Title, Author, Call Number, Barcode. I have left barcode off before, and then been confronted by two or more identical books on the shelf. The best way to distinguish between them--let's say one has gone out 20 times, and one has only gone out 9 times--would have been by barcode, but damned if I was going to run that report again . . .
Total Checkouts: This is an important number, especially if it is very high or very low. A low circulation number gives me a valid excuse to weed something. At some point, it really doesn't matter how critically acclaimed a book is if no patron ever checks it out. It's become a shelf-sitter and space-sucker. On the other hand, high circulation counts for books on certain topics will let me know what is of interest to my patrons and possibly lead me to purchase similar items. High circulation numbers for items that are missing or billed will give me a good idea of what I should spend valuable funds to replace in my part of the collection.
Some Statistic That Lets You Know When The Item Was Added to the Collection: This might be an Item Creation field or the publication date, which for older books (or books added to a previous ILS) can be pulled from the imprint information. If an item was just added to the collection, it isn't going to have a ton of circulations, and you don't want to waste your time thinking about whether to weed it. If you have both the date and the number of circulations, you can sort (I use Excel to manipulate my data and am forever sorting things) and create a subset of data such as "everything that has been in this Dewey range since 2003 and has circulated 2 times or less." You now have a manageable chunk of information.
Status: This may seem like a no-brainer, but it really helps to know whether the book is Billed or Missing or Claims Returned or even In Transit before you tear out your hair looking for it on the shelf. If you leave the missing books in your list of books to pull, you can kill two birds with one stone as you miraculously find them on the shelves. If you don't find them . . . well, you've got yourself more records that can probably be weeded, even if there's no book to go with them.
Optional Statistics: Sometimes I like to track the last time the item was checked in and out. This information is most useful when paired with missing/billed lists. We also track internal use here; items left on tables after browsing are taken back to circulation and scanned before being returned to the shelves. I probably wouldn't use this statistic for any section other than graphic novels, which are frequently read by people who don't or can't check them out.
Once you export any or all of these fields for the call number range you are working with, you have the tools to create a customized list of books that provides both more than the number of books you are trying to weed and encompasses the oldest material that has not been circulating. Some of these books you may end up keeping, for a variety of reasons (that's a whole other post), so it's better to overestimate. Then it's time to pick up a cart and start pulling books off to evaluate them for . . . moving on to the next phase of their lives. Going in a different direction. Deaccession.
How do you do it differently? What am I forgetting? Please feel free to weigh in below.
Note: We currently use the Millennium ILS, and that's where the field names I mention here are coming from. We'll be switching to Evergreen in a few months, and I assume my reporting experience may change radically. Woo?
I suspect that I care about statistics more than a lot of my colleagues. We do use statistics regularly during the course of our jobs; as reference librarians, we keep a daily record of our interactions with patrons, rather than recording stats once a month or once a year and extrapolating. Even so, I don't think many of my co-workers run reports as regularly as I do or are as familiar with the reporting software. Does this make me a stat-crazed librarian? I'm not sure, but running reports does provide me with a concrete starting place for a daunting project like weeding the 940s. While I'm using the data for weeding, I usually also come across information about the collection that may not be apparent from looking at the shelves. It would be a rare occasion that I generated a list of books in my area of the collection and weeded everything on that list--instead, I try to use statistics to pick off the weakest members of the herd. They should not be used on their own, but with a librarian's accumulated knowledge about his or her part of the collection, the needs of patrons, and the library's collection development policy, statistics can be an incredibly useful tool.
The trick to running reports for the purpose of weeding and collection maintenance is to approach with the goal of getting manageable chunks of data, but also to overestimate what you think you'll need so you don't have to re-run the report because you forgot a statistic that you were interested in tracking. I like generate more data than necessary about the section I'm interested in weeding and then sort and hide and cut until I get exactly what I need to weed effectively. My strongest advice: Do not try to weed the whole collection at once! Your end result should be no more than you can push around on two book trucks, because more than that will take too much time to sift through and may cause a staff rumble as you monopolize the departmental trucks.
Statistics to Export:
The Basics: Title, Author, Call Number, Barcode. I have left barcode off before, and then been confronted by two or more identical books on the shelf. The best way to distinguish between them--let's say one has gone out 20 times, and one has only gone out 9 times--would have been by barcode, but damned if I was going to run that report again . . .
Total Checkouts: This is an important number, especially if it is very high or very low. A low circulation number gives me a valid excuse to weed something. At some point, it really doesn't matter how critically acclaimed a book is if no patron ever checks it out. It's become a shelf-sitter and space-sucker. On the other hand, high circulation counts for books on certain topics will let me know what is of interest to my patrons and possibly lead me to purchase similar items. High circulation numbers for items that are missing or billed will give me a good idea of what I should spend valuable funds to replace in my part of the collection.
Some Statistic That Lets You Know When The Item Was Added to the Collection: This might be an Item Creation field or the publication date, which for older books (or books added to a previous ILS) can be pulled from the imprint information. If an item was just added to the collection, it isn't going to have a ton of circulations, and you don't want to waste your time thinking about whether to weed it. If you have both the date and the number of circulations, you can sort (I use Excel to manipulate my data and am forever sorting things) and create a subset of data such as "everything that has been in this Dewey range since 2003 and has circulated 2 times or less." You now have a manageable chunk of information.
Status: This may seem like a no-brainer, but it really helps to know whether the book is Billed or Missing or Claims Returned or even In Transit before you tear out your hair looking for it on the shelf. If you leave the missing books in your list of books to pull, you can kill two birds with one stone as you miraculously find them on the shelves. If you don't find them . . . well, you've got yourself more records that can probably be weeded, even if there's no book to go with them.
Optional Statistics: Sometimes I like to track the last time the item was checked in and out. This information is most useful when paired with missing/billed lists. We also track internal use here; items left on tables after browsing are taken back to circulation and scanned before being returned to the shelves. I probably wouldn't use this statistic for any section other than graphic novels, which are frequently read by people who don't or can't check them out.
Once you export any or all of these fields for the call number range you are working with, you have the tools to create a customized list of books that provides both more than the number of books you are trying to weed and encompasses the oldest material that has not been circulating. Some of these books you may end up keeping, for a variety of reasons (that's a whole other post), so it's better to overestimate. Then it's time to pick up a cart and start pulling books off to evaluate them for . . . moving on to the next phase of their lives. Going in a different direction. Deaccession.
How do you do it differently? What am I forgetting? Please feel free to weigh in below.
Note: We currently use the Millennium ILS, and that's where the field names I mention here are coming from. We'll be switching to Evergreen in a few months, and I assume my reporting experience may change radically. Woo?
Comments
One of the other things we look at is edition status. Is there a new edition of that book in the system? Sometimes, things get missed and we'll have a 2001 copy of a travel book floating around on accident. But there are times when people think having SOMETHING from that state/region/country is better than nothing. Even though there may be newer things in the system, it isn't at THEIR branch, so they will hold on to that 2001 book for dear life. Prices aren't the same. Places may not be there anymore. Websites may have changed. New Places may have emerged....but that book is still on the shelf.
Understandable statistics fascinate me and I want to be good at creating and using them. Is there a way for me to learn to do what you describe in this post with a home version of Excel and without having access to the staff side of a library system? Maybe that is too big a question, but this is something I would like to become skilled at but am unsure where to begin teaching myself.