What are the negative results of purging library records?
November 10, 2012 12:20 PM Subscribe
Libraries purge their records in reaction to Section 215. Why does a shortened life cycle for these records pose a hardship for libraries? In other words, if a library used to keep these records for, say, a year or longer, and in the face of Section 215 now purges weekly or daily, they're losing use of those records for 51 weeks or more--what benefit did the library used to have in keeping those records that extra 51 weeks?
The library I worked at would delete the record of what a patron borrowed as soon as the return is processed. We would know how many times a book had been checked out, but not who borrowed it. Though this is very good for patron privacy, it can create challenges. One example was when we were conducting research into how certain non-book borrowed materials were used, it would have been really helpful to be able to draw a sample from people who actually used these services, but we could only see who has those items checked-out right now. It also prevented us from developing any sort of relevant recommendation system (ie Amazon style, patrons who borrowed X also borrowed Y).There are of course additional privacy concerns that would have to be considered even if we had the data, but the lack of data renders the issue moot. The ILS we used did allow opt-in reading history, but people seemed to use it after they discovered they could not find what they had already read. Our interlibrary loan system did store records until manually purged. I found this to be very useful personally as I would reference it when I forget if I had already seen a book/article when I saw a citation elsewhere.
posted by cspurrier at 1:00 PM on November 10, 2012 [1 favorite]
posted by cspurrier at 1:00 PM on November 10, 2012 [1 favorite]
First of all, only some libraries do this, it's not a thing that all libraries do. I don't know if any library purges weekly, I think they're more likely tp purge when an item is returned. Wasn't sure if you were clear on that or not. As the two commenters above have stated, there are a lot of records that libraries do keep but for the most part they find ways to separate the patron records from what those patrons have checked out. This is good for Sec 215 reasons, can be bad for other reasons. A few examples...
- If you want to keep a record of what you've read. I can never remember which books by a particular mystery/thriller author I have or have not read and sometimes would like to be able to do this. Many library systems have a way that you can specifically and affirmatively opt in to maintaining these sorts of lists but then let patrons know they are less able to protect their privacy.
- Recommendation systems, as stated above. There are technically ways to do this even with anonymized patron data but there are not major efforts in these directions, mostly a lot of vendors hassling libraries about how our approaches to privacy are "outdated"
- Patron habits over time. We know how old most of our patrons are. It would be interesting to notice how their reading habits changed as they aged, both for young patrons as they entered adulthood [were they reading at their age level? above? below? similar to their siblings? no?] and older patrons as their options changed (did they move to large print? ebooks? videos?)
In short, there are a great deal of things that could be determined by looking at library data, some of which is hampered by privacy concerns but a lot more of which is just hampered by lack of staff or motivation. For libraries that operate in a competitive market somehow, this could be important, but for public libraries, this is something that often doesn't rise to the top of the priorities list.
posted by jessamyn at 1:32 PM on November 10, 2012 [3 favorites]
- If you want to keep a record of what you've read. I can never remember which books by a particular mystery/thriller author I have or have not read and sometimes would like to be able to do this. Many library systems have a way that you can specifically and affirmatively opt in to maintaining these sorts of lists but then let patrons know they are less able to protect their privacy.
- Recommendation systems, as stated above. There are technically ways to do this even with anonymized patron data but there are not major efforts in these directions, mostly a lot of vendors hassling libraries about how our approaches to privacy are "outdated"
- Patron habits over time. We know how old most of our patrons are. It would be interesting to notice how their reading habits changed as they aged, both for young patrons as they entered adulthood [were they reading at their age level? above? below? similar to their siblings? no?] and older patrons as their options changed (did they move to large print? ebooks? videos?)
In short, there are a great deal of things that could be determined by looking at library data, some of which is hampered by privacy concerns but a lot more of which is just hampered by lack of staff or motivation. For libraries that operate in a competitive market somehow, this could be important, but for public libraries, this is something that often doesn't rise to the top of the priorities list.
posted by jessamyn at 1:32 PM on November 10, 2012 [3 favorites]
One concrete example from my library is a patron contesting a missing book fee. He asked what book it was because he thought he probably returned it and, of course, we didn't have an answer for him. We ended up allowing him a gimme. That fee helps cover the cost of a new book plus processing staff time. I could see how this could add up in a large system. I should note that this record is kept for a while but is set to purge at a certain time. Certainly much longer than a month.
posted by Foam Pants at 2:18 PM on November 10, 2012
posted by Foam Pants at 2:18 PM on November 10, 2012
Tracking damage can be difficult too, if it didn't get picked up at check in (I'm specifically thinking of a large art book that got vandalised but the damage got picked up AFTER next check out not before) (which is as much of a staff issue as anything else). If there's a pattern, it can be even more difficult (a local library had a persistent vandal defacing books with queer content).
As a patron, and a librarian, borrowing history was really useful for me for my own records and for remote readers advisory (so when I'm trying to recommend/find a book for someone's kid and they can't remember the author - much easier to check the record with permission than fire off the big names and hoping for the best).
When I moved from a 'recording' library to one that had no records apart from the numerical, holds/fines and last borrower, it did change a lot of my customer service habits because I had far less hard data for recommendations (home library service people had spreadsheets for their mob). The next update was supposed to allowing opt-in for recording but it was 'nearly here' the whole two years I worked there. We did have a recommendation system but it was cobbled together via librarything and amazon (I think) and based on the book, not borrowing habits.
posted by geek anachronism at 2:29 PM on November 10, 2012
As a patron, and a librarian, borrowing history was really useful for me for my own records and for remote readers advisory (so when I'm trying to recommend/find a book for someone's kid and they can't remember the author - much easier to check the record with permission than fire off the big names and hoping for the best).
When I moved from a 'recording' library to one that had no records apart from the numerical, holds/fines and last borrower, it did change a lot of my customer service habits because I had far less hard data for recommendations (home library service people had spreadsheets for their mob). The next update was supposed to allowing opt-in for recording but it was 'nearly here' the whole two years I worked there. We did have a recommendation system but it was cobbled together via librarything and amazon (I think) and based on the book, not borrowing habits.
posted by geek anachronism at 2:29 PM on November 10, 2012
Libraries purge their records in reaction to Section 215.
This is...a highly problematic statement in a lot of ways. Let me break it down a bit, so that I can address some of your other questions, too.
-Libraries have long purged patron records (there are exceptions for library records in the Sunshine Laws that way predate the Patriot Act). For a lot of libraries, the Patriot Act was not the impetus to change anything (well, not in this area specifically) because there was nothing to change. So the 'good old days' may not have ever existed for many libraries in the way that you're thinking it did.
-Libraries aren't monoliths. A lot of libraries do keep patron records and didn't change this in reaction to the Patriot Act. I would argue that the view of the Patriot Act as a significant policy changer in libraries is ultimately a little over-stated (rhetoric yes, policy no).
-Libraries can be broken down into all different types. Both public libraries and academic libraries are just as likely to get subpoenaed and yet they will have very different needs and projects for patron records.
But:
-I have been running an experiment on moving DVDs from closed stacks to open stacks. This is a multi-department effort and we've been asked to prove that this is something worth investing in. We've pulled the circulation records on the moved items and can track things by date checked out but it would be much more interesting to track by type of patron and whether it's one person checking out all the DVDs or a variety of people (note that we don't purge circ data weekly but I haven't asked for this info since I don't think it will help our case).
-We could, as mentioned above, use circ info to develop aggregate patron profiles and do a better job with marketing (both in library, things like display racks and poster boards, and out of the library, with social media and other mechanisms) to micro-groups
-Students tracking down that one elusive citation from the stack of books that they just turned in could in fact look at their records and track down the books themselves (rather than sheepishly slouching up to the desk and asking the ref librarians for help).
Not the most useful examples ever, but as stated above, I do think your premise is a little flawed so I'm having trouble working within it.
posted by librarylis at 11:53 PM on November 10, 2012
This is...a highly problematic statement in a lot of ways. Let me break it down a bit, so that I can address some of your other questions, too.
-Libraries have long purged patron records (there are exceptions for library records in the Sunshine Laws that way predate the Patriot Act). For a lot of libraries, the Patriot Act was not the impetus to change anything (well, not in this area specifically) because there was nothing to change. So the 'good old days' may not have ever existed for many libraries in the way that you're thinking it did.
-Libraries aren't monoliths. A lot of libraries do keep patron records and didn't change this in reaction to the Patriot Act. I would argue that the view of the Patriot Act as a significant policy changer in libraries is ultimately a little over-stated (rhetoric yes, policy no).
-Libraries can be broken down into all different types. Both public libraries and academic libraries are just as likely to get subpoenaed and yet they will have very different needs and projects for patron records.
But:
-I have been running an experiment on moving DVDs from closed stacks to open stacks. This is a multi-department effort and we've been asked to prove that this is something worth investing in. We've pulled the circulation records on the moved items and can track things by date checked out but it would be much more interesting to track by type of patron and whether it's one person checking out all the DVDs or a variety of people (note that we don't purge circ data weekly but I haven't asked for this info since I don't think it will help our case).
-We could, as mentioned above, use circ info to develop aggregate patron profiles and do a better job with marketing (both in library, things like display racks and poster boards, and out of the library, with social media and other mechanisms) to micro-groups
-Students tracking down that one elusive citation from the stack of books that they just turned in could in fact look at their records and track down the books themselves (rather than sheepishly slouching up to the desk and asking the ref librarians for help).
Not the most useful examples ever, but as stated above, I do think your premise is a little flawed so I'm having trouble working within it.
posted by librarylis at 11:53 PM on November 10, 2012
This thread is closed to new comments.
But as to the record of who checked out which specific items, one big reason is that, often, patrons want you to have a record of what they checked out, so they can use the library as a sort of record of their own media consumption and refer to it later.
Other reasons to keep data are to establish what kinds of books or movies or whatever go out most, to figure out who is being under served by the library, and, yes, even marketing. Libraries actually market themselves a lot! It's part of "justifying" the budget allotted to them (increasing recognition of services and getting more people to use those services).
posted by itsamermaid at 12:46 PM on November 10, 2012