Tuesday, May 7, 2013

Connect and ICT in schools

Creation of ICT is fraught with problems.  When government seeks to create applications they tend to be ill focused and narrow in their application.  In education ICT has continuously failed to live up to expectations.  IMHO the risk in application development is best left with private enterprise in commercial solutions and the government should take a "best practice" approach, being reasonably ruthless in what they ask for when spending each dollar.

Connect is the next generation of the education department portal, seeking to bring data together for use by students and teachers.  It is a mixed bag of applications seeking to replicate what is already done by commercial and department applications that for the most part are freely available.   I hate to think of the amount of money spent on it.

I think the reason for not using public money is multifold.

1.  Replication of software that already exists freely is illogical
2.  A created solution will lag features available in larger communities where scale can be applied
3.  A small solution does not have access to the required depth of developers and designers
4.  Small projects tend to be lead by programmers and technical capabilities rather than by need of users
5.  Early adopters get burned.  Established applications proven to scale to millions are less likely to fail.
6.  Infrastructure is expensive in closed systems and rarely scales well

The Connect solution is experiencing issues that we viewed today at a presentation promoting its use.

1.  Connect infrastructure is already struggling.  Other solutions use commercial clouds and leverage their infrastructure.  Users of connect are already being asked to restrict their storage or are required to use external servers (such as YouTube).  This is an issue similar to the current email problem where a small storage limit precludes teachers using email as a silo for information.  It seems counterintuitive to say we need a managed private closed solution if we are locating resources (and sending students) to public locations.  We watched Connect struggle under load when trying to log teachers into the system, something that would quickly derail a class if used.  Timeouts were frequent.  Given that this was run in a closed system (within a school) controlled by the department on a day when students were not present does not bode well for the design of the solution.

2.  Connect allows communication directly between students and monitors this through sending multiple emails to teachers.  This is problematic as it creates spam in teachers email and creates a significant monitoring overhead (something overcome by the approach of other solutions).

3. The interface is suboptimal and has not evolved from 90's thinking (even in the presentation style).  The Connect interface is generations behind solutions such as schoology and edmodo (which is expected as Connect is drawing on a much smaller user base of 100 users vs millions of users).  Connect needs to move beyond thinking of itself as a portal and think about being a user's experience (eg. an SLN).  Students do not need to see "Connect" advertising on the landing page, they need to see what they need to know.  If a landing page serves little purpose, put something on it that has purpose, preferably in colours linked to the school (as belonging is a key component in driving usage).  If Connect is seeking to be another SLN and use this to drive usage, then this needs to be the "centre" of thinking, not being a content aggregator.

4.  It does not do anything better than what already exists and appears to seek to copy or relocate features of existing applications (cynically I would suggest to bloat functionality).  There is not a single feature that I could see that was a significant benefit over other already existing solutions (other than they had been aggregated in one place moving administration overhead to teachers).  Many features were being oversold (single login, presentation, RTP integration, notice board functionality, submission of work online, access to the ill organised and predominantly useless resource bank, password changes) and features that are lost were undersold (familiar interfaces, access to wider communities, access across school sectors, marking student work online (a feature that is very unlikely to be implemented but that already exists freely in full workflow form), quizzes, online testing, polling, usage statistics, integration with other learning communities, tablet app integration).  Little of the actual student benefit (social interaction, making work available 24/7, increase in information flow, connectivity to the classroom, collaborative peer support, BYOD device support) or the requirements of successful implementation was discussed (high levels of teacher interaction, high levels of organisation/preparation, ICT support requirements within the school, deadlines for implementation, staged rollouts, student education).

5. Proposed usage is not timely and demonstrates a lack of understanding of student usage. If a student is stuck at home they can call a friend, facebook them, email them, re-read notes, ask a parent, ask a tutor or wait until class the following day.  What they are unlikely to do is log into a portal, navigate to a community, hope that someone sees the message and then wait for a reply.  The response time is too slow for this type of approach.

6. Something is good if it is readily adopted and seen as useful.  I sat and listened to an early adopter claiming that Connect was a good solution.  Once we dug a little further, his actual student adoption with Connect was inconsequential yet he claimed he had great success with equivalent commercial applications (then why change?).  It made me laugh as it reminded me of the letter I received from the department saying I was a frequent user of OTLS (complete BS) and thus was chosen to use Connect.  Saying something definite does not make it true.  What the presenter was really saying was that the commercial solutions was better but if we had to use Connect it was ok and the interface was not unusable like OTLS.  We need to use Connect as it overcomes legal issues relating to student data.  To this I would say a) success is only success if usage can be shown and that compromising by using connect when better solutions exist because of closed system requirements (that cannot remain closed due to infrastructure requirements and costs) is poor reasoning. If we allow the open internet in schools (and we do - check any library at lunch) with only limited filtering,  a commercial solution is viable.

7. Parent involvement is overstated.  If I said to a parent, I will have grades emailed to you they might say great and read the email when I push it to them at a regular interval.  If I say, here is a portal, log in (requiring finding a password) if you wish to check how your student is doing, it is unlikely to be used often (what was my password again?). There is not enough information here for parents to review to make regular use viable.  The same can be said for homework review - this would take a cultural change that is unlikely in the near future.

8.  It is promising more than it is delivering.  It promised better access to SAIS (not available), access to RTP (not available), ability to store resources (but is limited in storage), access to classes (but not all classes were accessible on the day), email attachments to students directly (not available).  The "tell us your wants" and we'll make it happen statement is scarily present implying that direction for the application is poorly defined and lacking direction (flexibility in development screams scope creep and cost blowout in public sector application development).

Many of these reasons disappear once we pass through stages of early adoption.  I just question whether the need for another portal or SLN exists.  I have often been wrong before, but I believe that better solutions only occur if we challenge what is being done.  The challenge for Connect is to become an indispensable tool. It still has some way to go.