That's a very interesting question. Arguably, it is probably the sort of thing that the folks who sold Lektrievers would make in the old days.
  
 I think there are a lot of variables in such a study, but, as you note, intuitively, you have to think that a computer-based system is going to beat a manual system every time.
  
 So let's say that we have two people sitting next to each other. One has a computer that can access a large scale imaging system. The other is sitting at a Lektriever. Both systems contain personnel files. The objective is to record time to access a specific file, not the contents of the file. The person at the Lektriever is well-versed with color-coding and the files are arranged optimally, probably in terminal digit order. Each person has an order to pick 100 files.
  
 My expectation is that, over the course of the 100 picks, the computer system is generally going to be faster at the first page seek for the file. There will be times when the Lektriever is faster for getting to the full file simply because the proper tray will be immediately in front of the operator, but I would suggest that you would want to optimize the pick list for the Lektriever, although over the course of 100 random picks, the seek time would average out. For the imaging system, I would expect that the first image availability would be, in theory, as fast as "sub second" or as slow as 10 to 15 seconds, depending upon the backend imaging system and architecture. Compiling the entire file, depending upon how the images have been captured, may take a while if the documents are spanned across multiple disks. The architecture of the imaging system will have a lot to say about the speed to delivery of the entire file (and the size of the file will matter). With paper,
 the size of the file has little bearing on the delivery time.
  
 The question is really in the next step. If I have 100 files pulled, what am I going to do with those files? If the files are going to the office next door, the electronic method has minimal advantage. If the files are going across country or out of the office and need to be copied, the electronic system wins hands down.
  
 If you were to do another search, say, for 100 employment agreements that are inside of each folder, I would suggest that the imaging system will kick butt. Again, there are some variables and you have to assume some things -- first, you have to assume that the paper files have some internal organization. Second, the imaged documents are identified at the document level. In this test, electronic files identified at the document level will win every time, but there is a document preparation factor to consider when the files are captured. If the test is a somewhat organized paper file versus an unorganized electronic file (i.e. the documents have been captured in the order in which they were filed), the paper retrieval method has some advantage because the person searching can likely flip through the pages and process what they are looking for faster than a computer user can browse all the documents on screen.
  
 So I'm not sure that there is a clear apples to apples test that can be run. In general, the seek time to a single document or file is going to tend to give the advantage to a computer-based system. The seek time to a single document within a larger file of  unindexed documents is going to tend to give the advantage to the paper-based system. However, there is one additional factor. In a paper-based system, once you pull the file, you have to put it back in order for it to be available for subsequent searches. So, in effect, you have to count the refile time as part of your search time in order for the comparison to be accurate. That pretty much hands the speed title to electronic.
  
 Another factor is simply the volume. One person at a computer can access millions of files. A person in a file room or sitting at a Lektriever will have a finite number of files within a ten foot reach. At small volumes, paper and electronic are a wash. At large volumes, electronic is the clear winner.
  
 At the end of the day, the choice of a paper system versus an electronic system is more dependent upon external factors than upon absolute speed of access -- in most cases. There are, however, situations where speed of access and delivery to people outside the file room are absolute requirements and imaging is the only way to make that happen. The best example is a call center where a person may call with an inquiry about a particular document in a file. A paper-based system means a wait time of at least minutes for the document; an imaging system can deliver the document in real time.
 
 Cost is another deciding factor. A warehouse full of paper is relatively low-cost. An electronic system requires infrastructure and a fairly significant investment in capture, particularly if there is a large volume of paper files that already exist and need to be prepped for scanning. But, you have to factor in the labor for retrieval against the labor to prep and capture. An electronic system tends to become more efficient over time and particularly with multiple accesses to files. Since nothing has to be returned to file in an electronic system, every retrieval is a one way trip. Every paper retrieval is a round trip. 
  
 From a strictly academic standpoint, it would be an interesting test to construct and execute. From a real-world perspective, if my requirements involve real time access to the file, vital records protection, delivery across geographic boundaries, and / or simultaneous access by multiple users, electronic access is the only solution. The devil is then in the details of designing the architecture and setting up the capture process in a way that is efficient, yet delivers the required information efficiently.
 
 Patrick Cunningham, CRM

List archives at http://lists.ufl.edu/archives/recmgmt-l.html
Contact [log in to unmask] for assistance