Information bottlenecks at Big U

I was asked to write a few notes for management at the Big University (Big U) where I have an office that I use when not extreme telecommuting from one of my favourite islands. It occurred to me to rather place my comments here, where they are not so easily lost.

E-mail archiving (not)

Where better to start, than the pothole on the information superhighway that I struck just today. In a somewhat tersely worded e-mail from the person in charged of archiving at the library, I was requested to provide a copy of an "outstanding" document from October 2007. Fortunately, I clearly recalled a pleasant interchange of e-mail at the time and quickly found an e-mail message dated 1 October 2007 in which the same person thanked me for sending the documents which she was now requesting - 18 months later.

It is, of course, possible that this colleague simply did not think to check her e-mail archives, but I suspect that she does not have an e-mail archive because Big U provides staff with roughly 10% of the storage space (per user) that is allocated per student and has been known in the past to purge staff e-mail accounts of messages older than a few months. This is outrageously stupid, especially given how inexpensive storage is, and Big U should seriously consider archiving all internal e-mail forever.

Processing payments

Students apparently find it quite hard to have their payments acknowledged at the Big U. Someone at a Big Four auditor told me this week that a trainee in Pretoria had to fax a proof-of-payment to the office of the Registrar 65 (sixty-five) times before the payment was registered. I heard from another student that she was instructed to join a mile-long queue in order to pay R35 [sic] in order to join some tutorial group. Why it takes any time at all to register a payment is really beyond me. Any online shopping site or cellphone company can credit one's account instantly. At the university, students really only have a single account balance so it should not be more difficult to handle than a prepaid cellphone account. Why can Big U not make it easier for people to pay it?

Multiple submissions of "identical" documents

Multiple submissions are required of many critical documents. In spite of protestations to the contrary, Big U is still just a correspondence university which sometimes conducts this correspondence by e-mail. Lecturers submit tutorial letters separately (1) to the online environment (myBigU) and to (2) the department Scheduling. The latter requires one blue docket per language version of the letter, and on each blue docket information is repeated several times. They may later ask the lecturer to provide fresh "original copies" (3) of the tutorial letters for "peer evaluation" or some other purpose. The online versions (myBigU) are periodically purged. Nobody really checks whether any of (1)-(3) are the same. Why not let Scheduling simply source the documents from the online system, myBigU - where such documents can be approved electronically by the Chairs of departments - or at least rethink the procedure?

No currency information about documents

Watercooler gossip has it that the Advertising Standards Authority has been breathing down the neck of Big U because it has been charging students fees different from those advertised on its website. It is certainly easy to find some very outdated information on the Big U website, including by using the Google search box which Big U is so proud to display. There is a very simple solution to this - just place a "packed on" and "best by" date on every document. The same applies to internal policies and forms. How often does one complete a form, only to have it returned with a instruction to use a later version?

Information in an MS-Word DOC file cannot be considered "captured electronically"

I call this one, with no disrespect intended, Secretary's Delight. An old-fashioned paper form is lovingy retyped, with loads of megapixel scanned bitmap graphics to a document which is almost impossible to complete using the original word-processing software. Unfortunately, the request is often to return said form "electronically", so we slave on and somehow manage to massage the MS-Word tables into something that can conceivably be printed and e-mail it back. At this point, nothing has really been captured electronically. The documents being mailed around are not amenable to machine processing and no database records have been created. Until such time as the capacity exists to properly handly information in an electronic fashion, I suggest that such forms be reduced to plain text files with the required fields numbers 1.1, 1.2 etc. and comments and instructions given in lines starting with one of the usual symbols for commented computer code, e.g. the percentage sign %. A small amount of education in the use of plain text editors will be required, but Big U could dump thousands of licenses for MS-Office and improve the accessibility and archiving of their data at the same time.

No document workflow or tracking

Although many forms are available on the intranet, the workflow for the vast majority of these is not defined and/or unpublished. I am now in month 5 of my first (monthly) claim under the BigU broadband subsidy policy, which will finally pay about R125 per month. It is still not clear whether my claim has been accepted or not although - after much trial and error - I now believe that the correct procedure has been followed. Since this is a claim which should be made by hundreds, if not thousands, of staff members, would it not have been worth adding a single and authoritative paragraph to the form to detail the procedure?

Determining the progress of a critical application or process is a Herculean task. With the exception of very few processes (the submission of the "blue docket" version of a tutorial letter being one) information about the progress applications, submissions, etc. is very hard to obtain. It should not be terribly difficult to contrive a system whereby each official document is assigned a unique number and whereby each individual handling the document can register - on a publicly accessible website - the hand-over of the document or request down the line. The recipient of a document or request might, naturally, have the right to dispute a notification of hand-over but a system incorporating this feature would do wonders for university-wide tracking and record keeping.

Information that is readily available is not produced where required

Information that has been registered officially and that is easily available, is often not made available as and where it could be very helpful. A simple but non-trivial example is the coversheet that accompanies student assignments send in the internal mail to lecturers. Normally, the students' initials and surnames appear on the coversheet but no information that could be useful to the lecturer in case s/he wanted to contact the student instantly - such as a telephone number or e-mail address. It's in the database and should simply be printed.

The issuing of cheques is another opaque process. By the time that a staff member receives a cheque, s/he no longer knows what exactly the cheques was for. If every offical document had a unique number, that number could be printed on the cheque. At the counter where cheques have to be collected, staff apparently have access only to the recipients' telephone extensions and sometimes spend weeks trying to call staff members to let them know that cheques are ready. Why not just e-mail them? E-mail addresses could be printed on cheques, or if each document had a unique number, staff could simply use the unique number to look up the contact details of the originator of the document.

Presence information is not available online

Big U's electronic staff directory is not linked to the system for administration of leave, so it would typically require quite a bit of research to determine that someone is - for example - on leave for the month, or even on sabbatical for a year. It would be horribly simple to add a single record to the staff directory, where "away" messages could be posted by staff members themselves. The information in this staff directory should be made available to students as well, so that lecturers can leave messages like

At a conference in Namibia, available by e-mail only

for instance.

Password and privacy policy

A ridiculous password policy pertains at Big U. Apparently, "auditors" require that passwords be changed frequently and that accounts be locked out when a password has been entered incorrectly three times. This is, how one can gain complete control of the account of user X:

1. Look up the username of X - this is simply the part of their e-mail address before "@bigu.ac.za".
2. Go to any website that requires authentication, e.g. staff.bigu.ac.za and use the username of X and arbitrary text for the password - three times, so that the account gets locked out.
3. At this point you e-mail the ICT helpdesk - from any old account - and ask them to reset the password for X. It could happen that they ask you to provide the staff number for X, in which case you should have determined it in advance. If you don't have the staff number for X, just try again later - there is no restriction on the number of password reset requests.
   

If you are successful in obtaining a valid password for X's account, you probably have all night to use it since X will typically ask for a reset the next morning and will not suspect much to be wrong, because passwords frequently expire. Big U should reconsider their entire password and username system. The administration of the student online site myBigU requires a tedious process of associating usernames with courses. Why not allow any lecturer that is logged on to edit any student online course site? Just track the edits in order to be able to follow up an unlikely case of malféasance.

Open access to study material

Big U has problems with its registration process that have been widely reported in the press. One of the consequences is that students have difficulty gaining access to their study material (guides and tutorial letters) while waiting for their registration to be finalised. If Big U were simply to grant read-only access to all its tutorial letters and study guides (on an open FTP archive, for example), this problem would be solved instantly. However, Big U - in spite of the fact that its management is completely dominated by individuals professing a bleeding-heart ideology - is apparently mortified by the idea that *some* people might be able to read our study material without having paid. With MIT and other universities posting their study material for free on the Internet, is this likely anyway?

This posting is released into the public domain and may be copied and used in any way whatsoever, provided that any reference to the author is removed in case the text is modified.