What follows is my personal opinion.

Except where actual quantitative data and criteria are being used (which is not the case in a typical risk assessment of a work activity), numerical scoring is useless, distracting and false (because what happens is that people adjust the scoring to get the outcome they feel is right).

However it is still the norm to do it and auditors and so on expect it, so this tremendous waste of time and energy continues.

I am with Kate that numerical scoring is distracting, it is also dangerous when being interpreted by anyone other than the person applying points because the reader then spends time trying to determine why relative scores have been applied across various tasks.

Still trying to understand how a consultant applied a very high score of 12 (4 x 4 matrix) to the use of an office paper shredder whilst the highest score on a table saw in the factory was 9.

I have had long conversations with tame HSE inspectors about this and they basically think that the numeric scoring approach using a grid to “assess risk” is a complete waste of time and energy, as the numbers inserted are completely subjective and people simply adjust the numbers to get the result they want. Even more useless is the tendency to have multiple colours instead of just RED stop and Green ok. The various shade of yellow that some grids have mean that having gone to the trouble of collecting your information and trying to assess the risk you still don’t know if it is Green for safe or RED for not safe.

The HSE make and enforce the rules. If they think this process is a waste of time, why should anyone else get use it?

Answers on a post card to Blue Peter.

Originally Posted by: achrn

Yes, useless distracting and false, so potentially dangerous.

What's even worse are the ones that insist you have one lot of scores for 'before controls' and then another lot 'after controls'.  And then for the full whammy have spurious rules like 'you can't reduce the severity score'.

Totaly agree with all the coments so far, i teach the IOSH Managing Safely Course so have to use the before and after 5 by 5 scoreing for projects. While I agree you CAN reduce the severity, the one thing that candidates constantly strugle with is undertsanding if thier suggested controls have reduced, consequences, serverity or both. So even though the controls are good and the end result is reasonably practicable risk reduction they end up loseing marks!

Originally Posted by: peter gotch

I'll add my name to the "DON'T"!

The numbers invariably connect to some variant of a colour scheme where RED says STOP and GREEN says everything is spot on.

BUT.....sometimes the number might be such that the colour should be RED but it might not be reasonably practicable to do more.

Completley agree, the example i use is an F1 race around silverstone, no matter what you do  a multiple crash at 200 or so MPH is always a realistic probibility and death is a realistic consequence - so how do you get a green score?

I stopped adding a risk matrix to my fire risk assessments as generally speaking, my reports are aimed at non fire safety/H&S readers and caused confusion and anxiety.  Most customers want to know is: What do I need to do?  Why do I need to do it? How do I do it? Is there another way of doing it? Can you point me in the right direction? When do I have to do it by? So I made sure that when provided with a signifiant finding that required action, I would explain the rationale why (if it wasnt obvious) , provide a British Standard or similar that might apply (fire alrms/EL and FFE etc) and provide a  risk phrase to each finding (Extreme. High, Moderate and Low). Each phrase was linked to a nomnal time scale - Extreme = Immedately. I found that those with no or little H&S understanding would be able to interpret words in a table much easier than a matric - numbered or not!!!

AK, also depends on whose risk the appetite is for!

So, the car driver is about ten times less likely to die than the pedestrian, if we count the macro stats per billion miles or kilometres travelled.

Hence, if the risk assessment is for the driver at work quite easy when playing numbers games to forget that it is not the worker who is at the greatest risk but A N Other who just gets in the way but probably doesn't get counted on the driver's employer's accident statistics.

Then, when playing those numbers games there may be a tendency to presume a higher "severity" rating when the driving is at higher speeds when actually the fatality is much more likely to happen where there is a speed limit of 30 mph or less.

Department for Transport statistics show very clearly that rates of fatal and serious accidents per passenger mile are MUCH lower for "high speed" roads (defined as speed limit of 50 mph or more) than for low speed roads. Mostly as there are less "vulnerable" users (excepting motorcyclists) on the high speed roads.

So, if anything if someone is about to drive a 200 mile round journey and 190 of those miles will be on motorway or dual carriageway the person doing the numbers game risk assessment should choose an lower severity than for a 30 mile round trip all on urban roads.

However, what tends to happen in practice is that the assessor(s) go for the worst case scenario for "severity" UNTIL such time as they want to cheat.

All risk assessments are subjective. If you don't use scores or matrices how do you describe the level of risk?

To be fair, most documents I see claiming to be a risk assessment are actually hazard analyses. They are often good enough for purpose but mislabelled. This applies to the ones that do and don't use scores/matrices.

I have less antipathy towards scoring than most here, but I do agree they can make the process overly complicated and I too would counsel against playing the numbers game with them.

We use a Low/ Medium/ High matrix for both Severity and Likelihood ratings because we don't want our staff to play the numbers. But I'll let you into a secret - under the bonnet they are coded as 1/ 2/ 3 respectively, so a Severity = medium and Likelihood = high event is at 2x3 = 6 on the underlying matrix, which then converts back to a High/ Red risk requiring work to not commence or stop so a safer redesign of the process can be done.

It's how you use the process that counts and I entirely understand why HSE is against matrices given the potential for misuse.

Edited by user 18 January 2024 14:00:29(UTC)  | Reason: Typo

I thought it was a discussion forum!

Going to stick my head well and truly above the parapet (and get well and truly shot for it no doubt...!)

You are a seasoned, experienced health and safety professional. Do you need to use a risk assessment matrix?! No! You could do a risk assessment blindfolded with your hands tied behind your back! Would you recommend somebody else use a matrix?! No! Why?! Because you wouldn't yourself! But you are a H&S pro...

To use an analogy, you are like a Tour de France cyclist when it comes to risk assessments but someone new to the process is still trundling along with stabilisers. Maybe they need to use a matrix to help them begin to understand the process.

It is a tool, and like any tool it can be mis-used. Yes, I have heard people arguing over whether the likelihood should be a 3 or a 4, and yes I have heard people arguing over whether they have 'Reduced' or 'Isolated' but these people are not the H&S pro that you are, and that is where they have to start.

Look at anything. Music. Compare the first songs and the most recent songs of any artist and you see the quality improve.

Art. Compare the first drawings, paintings of an artist to their last works and the quality improves.

Risk assessments are the same - people need to learn the process with risk assessment matrices, just as we all had stabilisers when we were learning to ride a bike. Maybe you and I need to remember they aren't all as seasoned and professional as us yet.

Lastly, risk assessments are best done as a group - that solves some of the differences of opinion I find...

Right. Open fire...

Originally Posted by: andrewcl

Look at anything. Music. Compare the first songs and the most recent songs of any artist and you see the quality improve.

When I hear a piece of music in order to decide whether I like it I first assign a 1-5 score to the rhythm and then a 1-5 score to teh melody, then multiply them together and if the result is less than 9 then (and only then) do I realise that I like it.

Absolutely!  I  apply 1-5 for colour palette and 1-5 for number of brush-strokes.  Obviously you need to do this for what would have been teh score before the death of the artist and again now after the death of the artist, and it's not possible to change the brush-strokes score.

Well, we agree on that.

Originally Posted by: andrewcl

In the same way, the experienced risk assessor can do the process without consciously thinking about each step.

Well that is a major leap of faith and a consideration I must fully disagree with - anyone conducting any task on "auto-pilot" is a danger to themselves and others around them.

As to the "trainees" consider an auditing check list - if the training follows the list without deviation it does not make for a better auditor nor does it provide the business with useful feed back (see Kate's reply in Boots not Leaks https://forum.iosh.co.uk/posts/t133562--Boots-not-Leaks). Focusing on applying scores to what they see distracts from seeing anything other than the blatantly obvious.

No one should be conducting lone risk assessments for the tasks of others - the assessor doesn't know what they dont know and even talking it through with an experienced operator you can be led on a path of perceived best practice through unconcious bias where only a subsequent incident flags the difference to reality.

In my experience people don’t understand what “medium” actually means any more than 14 on a 5x5 scoring system. They only want to know can I do it this way or not, if not what is acceptable.

I have used both. Current job it is low medium and High, but in a previous place I inherited a scoring system, which the local managers and team leaders etc preferred to have. I did use the scoring to help prioritise areas to look at. It’s a tool so it does not matter how you do it or think about it as long as you get to a safe way of working, which is commercially acceptable to everyone.

What a great topic, my take is that any risk assessment is only only as good as the knowledge and experience of the activity that the Assessor has.The suitable & sufficient test is only applied(in most cases) after an incident or accident by which time it's to late.

More importantly, shouldn't there be a correlation about whats taught, what auditors want to see and what the HSE rexpect as the final arbiter of whats suitable and sufficient ?

I feel the scoring is a massive waste of time...we all assess things differently and will always have differrent risk ratings. It should be focused on what we are doing to control the risk.

Risk assessment - https://youtu.be/29b7GFjJBes

self promotion?

Pete, interesting question!

The consensus HERE might be that numeric risk assessment matrices are far from the best thing since sliced bread, but those who have posted on this thread are probably far from being a representative sample of the H&S profession.

It is a VERY long time since I helped out with delivery of Managing Safely when I was there mostly to see how one of our junior staff was getting on with their change of career direction from being a secretary to turning their hand as an OSH professional.

So, I can't remember whether the mandatory training slides then included any preference to risk matrices. May be that was added later.

The real test is almost invariably whether what is DONE rather than ASSESSED meets the requirement to do what is "reasonably practicable" and it is rare that ANY method of risk assessment fully evaluates that - and doesn't usually NEED to as what happens is that those doing risk assessments can look at what the specific legal requirements are + what industry etc guidance there is and follow these to come to conclusions as to appropriate precautions for any particular scenario. 99.9% of the time that will meet the required standard for what is a "suitable and sufficient" assessment.

However the European Commission came in and negotiated a fair few H&S Directives at a time when the UK was fully on board and one of the major influencers in the text in the Directives, which time after time said "The employer must assess the risks of X,Y or Z" and if X,Y or Z wasn't spelt out in the title of the Directive then the Framework Directive said "The employer must assess the risks anyway".

The UK could have "transposed" each of these Directives without including any specific requirement for risk assessment but rather stating that the employer should do what is reasonably practicable (or occasionally some other qualifier).

......and on one occasion the EU took infringement proceedings against the UK for allegedly not fully transposing a Directive when it included those two key words "reasonably practicable" and the UK's defence was successful.

However, the inclusion of multiple requirements for risk assessment in the various Regulations that have transposed EC Directives, the UK created a health and safety industry which has sometimes got out of control, breeding thought leaders who come up with pet ideas which then take hold - one of them being numeric risk matrices (IMHO!).

It's not clear exactly how many delegates attend each IOSH approved course each year but my guess is that Managing Safely is still IOSH's top product.

So, if that remains successful and incorporates a bit on numeric risk assessments, it may be difficult to persuade the Institution that it is time to drop this particular element.

Remember also that many of those approved to deliver MS may not be that experienced as OSH professionals, not least as the going rate for delivery doesn't leave much profit margin for most providers - hence the providers tend NOT to be the experts in H&S consultancy who can make MUCH better margins from other services (EXCEPT to the extent that they might maintain their licences to deliver MS for tactical reasons - it might be a small part of the broader offering).

Edited by user 08 February 2024 17:59:51(UTC)  | Reason: Typo

I deliver IOSH managing safely. God I hate the risk assessment module! 

FAIL

Morning HS Snail

"Apita" has posted three times on these Forums.

The first time on 23 January. A blatant advert for a training company with the post being REPORTED.

The second time on 31 January eliciting this response.

#5 fails the Turing test.

Is there an AI yet that can evaluate against the Turing test as well as a human?  And what would we call that test? 

I know, can we call it the Kate test?  I always wanted to have a test named after me.

So, this is their third attempt which is just gobbledegook. Even less like Plain English that the language that some doing H&S exams are supposed to answer in.

It fails the "Kate test" (AGAIN!)