Getting the ingredients right for quality analysis
I’m a big advocate for quality management, but it’s not always obvious what this means in the context of analysis. But quality management can really help ensure that quality is baked on from the start to the end of the analytical life cycle, improving transparency and understanding at each step.
The term quality management reflects all the activities that are in place to ensure that the outcomes being produced are fit for purpose. This includes understanding the quality of the inputs and the required quality for the analytical outputs, but it also takes into account the impact on quality from the wider environment in which the work is taking place, including the tools available and the culture around quality improvement.
Quality management in simple terms
Imagine you’re making a cake. You need some inputs (flour, eggs etc), a process (the recipe) which (fingers crossed!) leads you to an output you are expecting (a cake). You can be fairly confident that if you measure the ingredients well, add them in the right order and bake it in the oven then the outcome will be as expected. But making a cake that is fit for purpose is more than just the inputs, processes and outputs. You also need:
1. To understand who will be eating it
Who are you making the cake for? Is it a birthday cake for a small child who loves Bluey, a wedding cake for a very fussy bridezilla, or intricately decorated cupcakes for a Bake Off competition? Maybe there are dietary requirements to consider? What can you include and what must you exclude from the recipe?
2. The right tools and environment
It’s probably preferable to make a cake in the kitchen compared to the bathroom, for example. You’ll also want your favourite cake making spoon, the big mixing bowl, the spring-form cake tin that always gets results etc – so the right tools to do the job.
3. To know what to do when it doesn’t go to plan…
And if things don’t quite go to plan, you might need to take some corrective action. Maybe the oven has got too hot and you need to reduce the temperature? Maybe you haven’t got the right inputs? (no butter? Find the margarine!) Maybe you haven’t got the right process, so defer to a more reliable recipe? (do you have a favourite recipe that’s covered in globs of aged cake mix?!) Or maybe your methodology isn’t working, or you don’t follow the process in the right order? Any of those actions can have an enormous impact on outcomes, and the quality of your cake.
4. And work out what you will change for next time…
Having made your cake, you want to move on to bigger and better things (the showstopper!), so you look at what didn’t go so well (the oven was a bit too hot?) and what did go well (lemon icing, yes!), and refine and adapt for next time.
This understanding of the connection between input and process is vital to understanding and communicating the quality of the analytical outputs. But the quality control actions, culture of improvement, and having the right tools in place, are also fundamental parts of managing quality effectively and being able to evaluate whether the quality of the outcomes meets the agreed requirements.
What this means for analysis
As analysts, putting effective quality management in place from the start means that quality needs are understood, the impact of decisions made during the analysis can be anticipated and risks to quality identified early, and the resulting quality of the analysis can be communicated to better inform its use.
The Analysis Function provides a wealth of guidance and resources you’ll need to develop your baking (sorry, analysis) skills, and is a like the Good Food website of information for government analysts with both ‘tried and tested’ and innovative content.
So, only if you consider who you’re making it for, whether you have the right ingredients (and equipment!), what the correct process is, and only if you check on progress, taking remedial action where required throughout the process, do you get a fantastic (indeed edible) cake. Sorry, I of course mean high quality analysis.
Cuppa and a muffin anyone?
The Data Quality Hub