Failures in Figures – Sloppy Location Tracking Part 4

The reasons and what can be done about it II

wantedIn part 3 of our series we took a look at what part of our human nature causes the errors we’ve discovered in part 1. But they are not the only reason. Procedures and technology play a big part in this.

1. Number of steps towards location change

Whenever I compare notes with colleagues in all honesty, we come up with almost the same observation: however exact and detail-oriented we tend to be, about 10% of the time we just get it wrong. Most of the time we confuse numbers, but there is the whole range of errors we discovered in part 1. That almost none of those mistakes make it into our data bases is due to cross-checkig procedures we impose on ourselves. For example, after we have written down a number, we compare it again with the number on the object or we work together with a colleague.
If there is a high risk of confusing numbers whenever we write them down, it is totally logical that the more often we have to do this in the process, the more likely we get it wrong. It’s just as logical that the number of possible mistakes increases the more people are involved in the process and the more time passes between the original location change and having it secured in the data base.
The worst location change process I ever encountered looked the following:

  1. The collection manager noted the location change on a piece of paper at the offsite storage.
  2. The collection manager or one of his/her assistants composed an email with the location change and sent it to the documentation team when he/she got to a computer with web access.
  3. A team member of the documentation team made the location change in the data base.

Obviously, there are three times the accession number and location is noted which means where you can get it wrong. Along with the “normal” confusion of numbers, there is the possibility to read a handwritten number wrong. In addition to those writing or typing mistakes there is the additional possibility that the location is entered wrong, because the final change in the data base is done by someone who isn’t familiar with the location numbering system in the storage. While the collections manager would probably realize that he/she can’t have put a grammophone needle on a heavy duty rack, this detail escapes the documentalist in his/her office.

What can be done about it:

  • Limit the steps it takes to make the actual location change to a minimum. Ideally you have data base access in all storage areas that allows immediate location change.
  • Anyone who actually changes locations of objects has to have the possibility to do location changes in the data base.
  • If more than one staff member is involved, make sure there is a feedback loop when the location has been changed, so the original changer of object location has a chance to check for accuracy.
  • Technology like barcoding, if proper implemented and working, reduces confusion of numbers to zero (Read the examples of the National Galleries of Scotland and the TECHNOSEUM.

2. Complicated numbering systems

This is not really a surprise, but seldom someone thinks about it: if your location numbering system and your accession number is confusingly complicated, you increase the likeliness of mistakes to happen. Accession numbers that follow the logic “year of accession/number of object that year/parts number” are far more easy to remember and to write correctly than an accession number that tries to convey a multitude of information like “number of department/number for material/year of accessioning/number that indicates if loan, education collection or permanent collection/number of object/parts number” 1. Unlike computers, humans just aren’t good at remembering numbers and even if it’s just a brief moment between looking at the number and writing it down, there is a memory process involved. One Mr. Miller came to the conclusion that the human brain can hold 7 items at the same time in 1956 2. I don’t doubt that there are really brainy colleagues who can remember much more, but I found this rather accurate.
elephantsOne reason why the three-part numbering system is easier to remember than other numbering systems is that one part of it, the year of accession, is telling people something. As a human being you don’t read 1977 as 1-9-7-7, you read it as the year Elvis died or your daughter was born. That’s probably why I seldom discovered that someone got the year of an accession number wrong and when it happens, it’s mostly because there were numerous other accession numbers with that year to take down, so when one item comes from another year, your brain just copied and pasted what it had read several times before (remember what I said abut concentration in part 3!) or that the digits in the year are notorious to be confused when written on paper like 5 and 6 or, in some terrible handwriting 8, 9 and/or 0. So, one could say that the first part of the three parts system is just one bit to remember, not 4. The next part contains 3 or 4 digits, which is possible to remember, as well as parts numbers, as long as they are not too many. A 1988.1243.001 is easier to remember than a 1988.1243.193, simply because the first one meaas that you have to remember 6 bits (1 year + 4 digits + 1 digit) while in the second example it’s 8 bits (1 year + 4 digits + 3 digits).

What can be done about it:

  • Choose a numbering system the human brain can remember.
  • Keep it simple. Don’t expect the accession number to include ALL the information. It is totally sufficient if the accession number makes it possible to distinguish one object from another, similar one. All other infomation can be drawn from the data base or an accompanying inventory card.
  • Avoid whenever possible the necessity of memorizing numbers.
  • Mark your location codes clear and readable on every location unit. Just because the system seems obvious and logical to you it isn’t for the next colleague who has to deal with it. Yes, I’m looking at you, collections managers! If you choose not to label each shelf of a shelving unit, don’t expect anybody to know which is “shelf a” and which is “shelf e”.

3. Responsibilities

With every person responsible for location changes the likeliness of mistakes to happen increases. And the more involved, the harder to track where something went wrong. Also, with each person on the team that is involved with object handling, the variety of mistakes increases. That might be a little hard to comprehend, so let’s make an example: Our valiant collections manager X is very detail-oriented but has a serious quirk: she tends to confuse left and right. As locations on the shelf boards are separated by “left”, “center” and “right” she sometimes confuses those, too. The curator Y is often taken away by the sheer beauty of his objects and tends to forget from which shelf he has taken an artifact. As he is at the same time convinced that he exactly knows what he is doing, he often puts the object back on some other shelf. Conservator Z is marvelous with treating artifacts but terrible with numbers. When writing three accession numbers on a box she sure gets at least one number wrong. Each of those quirks looked upon seperately are easy to mend: depending on who has handled the object last, you just know that you have either to search on the opposite side of the shelf (handled by X), the shelfs in the vicinity of the original location (handled by Y) or play around with certain possibilities of number confusions (handled by Z). As soon as you don’t know who handled the object last, you have to take all possible mistakes into consideration, leading to more invested working time in discovering the object.

What can be done about it:

  • Limit the number of people responsible (and allowed) to handle objects and make location changes.
  • Track every location change not only with date and reason, but also with who actually did it.
  • In larger institutions: Make sure you assign and communicate responsibilities clearly. For example: who is making the location change, the giving or the receiving party? If the collections manager sends an object to the conservation lab, the curator or to the photographer, the collections manager is responsible for the location change. If a conservator sends an object to the photographer, the conservator is responsible for the location change. If the photographer sends an object back to storage, the photographer is resposible for the location change.

This was the last part on location tracking in our series “Failures in Figures”. I’m sure that there are other points to consider which I forgot to mention. I’m looking forward to your comments, additions and ideas! I’m also looking for new suggestions on what we should examine in “Failures in Figures”.

Best wishes
Angela

  1. Don’t laugh at that, I actually worked for one institution that had a very similar accession numbering system, including a “.1.” right in the middle that none of the people currently on staff could explain, it just had always been there and they never encountered any other digit there.
  2. Miller, George A., The Magical Number 7, Plus or Minus Two: Some Limits on Our Capacity for Processing Information, Psychological Review, 1956, p. 81-97
Facebooktwittergoogle_plusredditpinterestlinkedintumblrmail

This post is also available in: German

One comment

  1. Rebecca DuBey says:

    What a wonderful study; I wish all museum people could read this to get an idea of how easily mistakes are made, and why those of us in charge of artifacts demand we be allowed to follow our protocols, methodologies, and Best Practices. Every time I break one of those, I pay for it later in the time (and stress) it takes to straighten things out. thank you!

Leave a Reply

Your email address will not be published. Required fields are marked *