When choosing software, hardware and services to use there are many considerations. We assess the options using these attributes:
Usable
Accessible
Effective
Efficient
Cost effective
Scalable
Reliable
Guaranteed
Sustainable
Persistable – you can save and restore data and the state of the system easily.
Deletable – you can permanently remove data if you need to.
Secure and private
The extent to which each of these matters varies depending on the activity. You and your collaborators need to care about all of these, although you should be able to call on help from specialists (for example UX Designers considering usability and accessibility). IT managers are especially concerned with choosing options that are secure and can be trusted.
The people in charge of running IT systems in universities combine two hard to reconcile attitudes: on the one hand they are advocates for tech, they want people to be able to do amazing things, they want to have a big, exciting, sexy (steady on now!) impact; but at the same time they are deeply risk averse, they know that tech goes wrong, and they don’t want to have to sort out the mess that gets made when it does. But they do have ways to come to balanced decisions, and we can all use those techniques.
This risk aversion has been heightened further as people are more willing to share personal (even intimate) details online, but at the same time the legal and moral consequences of data breaches are increasingly severe.
Security
Every institution now has to take security seriously. Every member of the institution is responsible for avoding risks and following good practice. Software that is recommended by the University is screened for security risks, following a thorough and systematic process. This must be respected.
Trust
When choosing software for projects, or creating our own software, IT managers have to consider if we can trust in the providers of the software and the services required to keep it availale. What guarantees do we have that it will remain available and usable? Consider a software product you use, for example Twitter. Where do you rank it on scale A? I would say we only have a weak guarantee that it will remain available. Recently the twitter API, relied on by many researchers, became a paid for service rather than free. If that happens, what level of dependency does your work have on it? Rate it using scales in B below. If your research has a serious dependency on it, and you can’t afford the new fees, you are in trouble. These are just some of the considerations in the minds of IT managers when choosing software and services. We need to apply these considerations to our projects. Have a sensible, informed, balanced approach. If the impact of losing a service is low, you might be OK to use it even when you have no guarantees. If the impact will be high, seek a solid guarantee or use something else.
A. Guarantee of availability
Solid guarantee: the university is fully in control of and guarantees the availability of the tools and services, provides support and fixes problems;
Good guarantee: you have a contractual arrangement with third parties that guarantees availability for a specified length of time – this still does not absolutely assure availability, but is enough to work with;
Weak guarantee: the resources or system are widely used, with a global user community, and backing from a major company or organisation;
No guarantee: there is no contractual or organisational reason why it might not just disappear.
B. Depths of dependency
No dependency: we could remove the specific resource or facilities and it would have no impact on your projects and operations
Mild dependency: removing access would have an impact on the ability of students to complete the module, requiring some reasonable effort to make up for the loss, but could be done without too much disruption and cost;
Serious dependency: removing access would have an impact on the ability of students to operate, and would require assistance from outside of the team to make up for the loss;
Critical dependency: removing access would prevent further work of access to resources and cannot easily be solved with the available resources within the university.
Leave a Reply