Saturday, 15 March 2014

What can education learn from Quality Management?

Bit of a big post this - it's been evolving in my drafts file for a while now...

I learnt a lot about management and quality during 10 years working in the automotive industry. Due to the complexity of the products this is an industry that has been at the forefront of quality management since the 1950s. Cars are the most complicated consumer product in the world. They are built in vast numbers, require tens of thousands of components to work together when operated by relatively untrained drivers in a massive array of conditions. What's more failures have the potential to be both catastrophic and fatal. Also they are almost all built to a very tight budget, meaning that waste needs to be eliminated from all parts of the business to allow a car company to turn a profit.

The reliability of modern cars is truly gobsmacking, and it is due to the fact that the automotive industry are global leaders in the field of quality management. Versions of the approaches pioneered in the automotive sector are now deployed across manufacturing, and are even being used as models for business management in many non-manufacturing settings. However even the most forward thinking of these alternative applications are usually about 5-10 years behind the latest in the automotive world.

I'm sure having got this far you're now thinking "Kev's lost it, he's gibbering about cars, I thought he was a teacher and this was a blog about education?" Well you might be right! But I really do think there is much to learn about the idea of quality in education.

Don't panic, I'm not about to insist that schools are like production lines, propose time & motion studies or some hideous Fordian uniformity to classrooms; I guess I need to explain where this is coming from in terms of how quality management evolved... (but if you want to cut to the chase then scroll down to the "lessons for education" heading)

An evolution of practice
All quality assurance or management systems currently in place in any industry today have a lineage that traces back to the automotive sector in post war Japan. Before this point it was all about "quality control" - essentially inspection at the end of the production line to check that things were correct. This was ok, but it was only partially effective at catching all potential problems. You physically can't check for everything. As a result some things slipped through the net and caused issues in the field.

From control to assurance
To improve on the partially effective "control" system the emphasis shifted to "quality assurance". The premise was that rather than inspecting at the end of the line where tests were limited and rectifying errors was expensive and took a long time, why not do that inspection earlier in the process? Perhaps even before the parts are fitted? Or even delivered? The automotive firms pushed chunks of their inspection processes back up the production line all the way into their supplier's factory. The idea being that if all the parts arrived certified as "good" then the resulting car must be "assured" to be good. They still inspected a smaller number at the end of the line, but quality had improved as faulty parts were often detected before they were fitted.

However there was still a level of variability inherent in the design. Humans build it for a start, and we make mistakes! Even certified parts have to be supplied to geometric tolerances that cause variations, and other physical, chemical or even biological variations can creep in when you make large numbers of components or large numbers of vehicles. (Biological - really? Well for example loads of car components are made from rubber, which is a natural product. It literally grows on trees! As such, until relatively recently when it has become better understood and controlled, rubber components on cars were subject to variability depending on the season in which the natural rubber was harvested! As another example the quality of some car paint finishes can be affected by the type and quantity of deodorant used by the operators working in the paintshop!)

Error proofing
The next development in quality was to start to manage it from the outset. To do things with the design that prevented errors, or make the performance of the completed vehicle tolerant to variability of its components. Japanese terms like "Poka-yoke" are now commonplace in car design - it means "mistake proofing" and helps to remove human errors on the production line. For example, if an operator has to connect 3 different electrical plugs in the same area of the car each plug should be designed such that it only connects to it's correct socket, leaving no room for human error.

Assurance becomes management
By taking the focus away from inspection/control, wherever it is in the process, and looking in more detail to the systems and processes quality becomes managed rather than assured. This means designing OUT variability, designing IN error proofing, planning for quality from the very start of the process rather than applying it as an inspection at some point.

Continual improvement requires empowerment.
However perhaps the most powerful thing to come out of the Japanese automotive industry was the concept of continuous improvement, and with it the empowerment of everyone in the business to make suggestions to improve the product. This is often referred to by the Japanese term "Kaizen" (literal translation "improvement" or "act of making bad points better"). Toyota used this word to brand improvement activities in its factories and visiting western engineers and managers adopted the word.

At the heart of Kaizen is a philosophy that improvement must be lead from the top, but not directed from the top. Every worker in the factory has their part to play in the quality of the end product, and as such every worker has the right to make suggestions about how to improve it.

Vitally this includes the idea that the person that fits the brakes all day becomes an expert in fitting brakes. As such this person is very well placed to make suggestions about how to minimise errors when fitting brakes. This applies across the whole vehicle and as a result the shop floor assembly workers have a big voice in improving designs and optimising the processes.

Firstly, we're currently broadly applying the "quality control" type of management. We inspect at the end of the process, both by assessing student's progress through final high stakes exams, and by the use of increasingly high stakes observations/assessments for teachers (I say increasingly high stakes due to imminent explicit linkage to pay structures in the UK), and high stakes inspections for schools.

The best schools will use more of a quality assurance model. "Good" practice will be embedded in school and departmental policies to reduce variability in practice between staff. However too often these are policed and enforced through inspection (e.g. observations, work scrutiny, learning walks). To take this on to the next level the structures need to be put in place to make good practice, and therefore success of the students, inevitable.

Developing systems in which good performance becomes inevitable can only come if the people doing the processes are inspecting it themselves. It becomes less about doing it well because you are being watched, and much more about not needing to be watched because you are watching yourself.

Ok that last paragraph sounds like a load of idealism, but if we pick up the Kaizen model in education and truly empower teachers to improve their practice they will feel a much greater ownership of it. By encouraging this ownership we make it much more likely that they will do it.

For example, who is best placed to formulate a marking policy that is workable for a teacher with a full mainscale timetable? It certainly isn't best done as a decision by someone that only teaches a partial timetable. Setting out a basic framework that includes the key characteristics of good marking and then asking teams of all staff to develop a way that this can be done in a manageable way would create a policy much more likely to be adhered to.

Just this week I was told about someone who wants to try #RAG123 marking (not heard of RAG123? It's awesome - see here!). They aren't allowed because it doesn't conform to their school's policy. Under their school's policy this person recently spent over 2 hours marking just 7 books, and they are saying that the quality of their planning is suffering due to the volume of marking they have to do! Nobody on a full timetable could possibly sustain that type of marking alongside teaching, planning and having a life. This is a clear example of a solution being imposed on people without thought to actual delivery.

Similarly it is really important that there is a route and process for all staff to highlight where the school is not working efficiently. For example are there flaws in the school sanctions and rewards system meaning that a group of teachers are struggling to use them effectively? Feedback loops are important in industry, and should be in education too. Vitally though if feedback is sought and given there MUST then be action with support from the top to address the concerns and improve the situation.

Where is the value added?
It is recognised in the automotive industry that the only people actually adding value are those building the cars. They take the components and combine them into something that can be sold at a profit. Everyone else and every other process in the organisation is an overhead that chips away at profits. They may be absolutely needed as part of the long term business, but they still cost money. As such these other processes need to be as efficient as possible, and mustn't interfere with the effectiveness of the production line.

In schools it would be too simplistic to suggest that the only people that add value are the teachers, as it's the total experience at school that's important, and not always just the lessons or exam results. However systems in the school absolutely mustn't make the jobs of those that interact with students harder to do well.

Consider how easy it is to get accurate data about a specific student... Attainment, targets, behaviour, attendance, SEN, FSM, IEP, etc, etc, Is it all in one place or in lots of different places? Is it easy to download a class list of information in a usable format? I know I've worked in schools where each bit of data is stored in a different place, and in different formats.

I've heard regularly that schools do well by prioritising on Learning and Teaching. The question "If it's not improving Learning and Teaching then why are we doing it?" pops up as part of this kind of thing. However how often is that really applied across ALL systems in a school? It may well be applied to guiding CPD, or some directed time activities and meeting agendas, but is the attendance system actually optimised to stop it interfering with learning and teaching? Is the behaviour system an add on administrative activity or an integral part of learning and teaching?

Loads of education practice, particularly on the administrative side, is based on finding a system that basically works, and then iterating as needs change. Sometimes this creates a real monster of a system with add on bits and extra files all over the place. For example all schools I've been in have slightly different ways of managing student data, sometimes this is based on the specific skills or preferences of the staff involved in creating them, or just on how it's been done for years. Sometimes these systems are brilliant, other times they are ineffective. New staff come in and have to learn the foibles of a particular system, and then all the various "work around" methods to get key bits of data in the right format.

It is incredibly rare to find a system that has been completely designed from the ground up to do its job in a way that is completely aligned with the needs of all of the users in the organisation. Process mapping and optimisation of processes are effectively alien terms in education, but they really shouldn't be.

In summary - we need to start actively managing quality
Basically what I'm trying to illustrate is that any push for improving "quality" within a school needs to aim far more at the quality management end, which is the cutting edge of quality practice; as opposed to the quality control end which is a blunt and inefficient instrument.

We need less direct inspection to enforce systems from the outside, and more design of systems to make good performance inevitable. We mustn't invent extra processes to fix problems; instead we should develop systems that simplify the job rather than making it more complicated.

Like choosing to walk across the grass or around the path, people only deviate from policy because there is a shortcut. We need to seek out and use the expertise of the people that will actually work with the policies the most to help redesign them to eliminate the shortcuts! If we make it hard to do the right thing then we can't be surprised if someone does it wrong. The purpose of good leadership and management must be to design and environment where we make it as easy as possible for our staff and students to do it right. That way success becomes inevitable.

As always I'd welcome any feedback and comments... :-)


  1. Quite quick turn around, I love the posts!! AQM

  2. Even certified parts have to be supplied to geometric tolerances that cause variations, and other physical, chemical or even biological variations can creep in when you make large numbers of components or large numbers of vehicles.
    accounting writing service

    1. Yes, but the trick in good engineering is to make sure that the design tolerates that variability and doesn't cause issues. Certification doesn't mean perfect uniformity - it means consistency in quality (or fitness for purpose).