On diagnosis, parents, teachers and clinicians

A couple of summer holidays ago I wrote a book. I enjoyed writing it a great deal, it was a strangely cathartic process, and it was published last May.

You can find details of it here, and buy it if you like.

Over the course of that summer I found myself asking a question – should I include the bit about specific labels, or would doing so blind readers (labels can be a bit dazzling), and get in the way of them finding out about the child?  Or, should I leave it in, as information about the sort of diagnoses that teachers are likely to come across is useful in that it demystifies them? I debated with a number of my teacher friends, and in the end I compromised. I left them in, along with a largeish section on the problem with labels, the different kinds of meaning they hold for different people (adult and child) and how a teacher needs to be aware of this and bear it all in mind when teaching.

The other thing I thought was important to put into the book, bearing in mind that being the expert in the classroom can give a person the impression that they are the expert in everything (or at least, that’s how it can be perceived from the outside – there is also a section on saying sorry and how to manage mistakes), was clear and unequivocal guidance on what a teacher is able to diagnose – or not.

Here is a quick taster:

Autism Spectrum Disorder : a teacher cannot diagnose

Attachment Disorder : a teacher cannot diagnose

ADHD/ADD : a teacher cannot diagnose

Down’s syndrome : a teacher cannot diagnose

I’m sure you get the picture. These things are diagnosable, not by teachers, but by clinicians, that is, medical doctors and psychologists (who may also be doctors).

You can find a reflection on a medical diagnosis and what it looks like in practice here.

What this means is that the problems a child is facing in the classroom are problems that they face everywhere – down the shops, in the swimming pool, in the home, all the time. Aside from Attachment Disorder – and even then, when a child is adopted, this is not the case – these diagnoses, or labels, have nothing to do with parenting, style or anything else.

But to get back to my list, next up:

Dyslexia – can be diagnosed by specialist teachers (the specialist bit is important –  you need to complete further, demanding qualifications in order to be a specialist dyslexia teacher with the ability to diagnose), parents/schools (does a school ever do this?) have to pay around £500 for a full assessment of dyslexia

And then we come to yesterday. Yesterday there was a report published (you can read it here), based on a survey commissioned by an assessment company, GL Assessment, that claimed (in a nutshell) that teachers think that labels of SEN are obtained by the pushiest of parents (and by default, the wealthiest and most middle class), and that this means that some children, who need and deserve support (presumably the children of the poor, or JAMs), aren’t getting it. Sounds like truth, doesn’t it? The squeaky wheel and all that. Except when you contrast it with the statements above, and the fact that we have (for now) a National Health Service.

Are we teachers really saying that we don’t believe in clinical diagnoses? (see press release here)

Now personally, I don’t think anything of the sort, but I do think, because I am a teacher and I know how difficult it is to balance a class and to understand the sort of SEN that doesn’t come with a diagnosis (the sort that is most common in classrooms), that questions about teacher perceptions of labels of SEND need to be very carefully framed. Firstly, because SEND does not mean some sort of group of children where each one is the same, facing the same kind of difficulties: homogeonised. And secondly, because, if you are not careful, what you actually get in answer to your question is something completely different.

If you look at the survey (link here) I think what you find are two things:

  1. A lack of understanding of the reality of SEND by the questioner.
  2. A neat exposition of teacher attitudes towards a certain set of parents (mothers, let’s face it) who are exhibiting anxiety about the educational progress (or not) of their child/children.

I’m not going to go into the reporting of the results of the survey (although the Guardian – what were you thinking??), I’ll  save that for another day, except for one thing – a press release is an important document.  Read an excellent exposition of the dangers here.  You can see two contrasting reports on it here, from TES and here, from the Guardian, and draw your own conclusions.

Needless to say, if I was writing it up, I’d have written it very differently, because guess which SEND assessments (among others) GL Assessment sell? Dyslexia.

Advertisements

19 thoughts on “On diagnosis, parents, teachers and clinicians

  1. A thing that concerns me, that you haven’t mentioned, is the matter of whether this ‘study’ was subject to any sort of peer review. So, as a result, we don’t know how methodologically sound the research was.

    I just looked into the report itself and I’m actually appalled.

    The questions used were utterly stupid:

    “Do you think that there is currently a
    misdiagnosis of special educational
    needs amongst school children?”
    Y-57% N-26% DK-18%

    But where was the “I’m not qualified to answer this question” option?

    “Do you think that pressure from parents
    has led to some children at your school
    being categorised as having special
    educational needs unnecessarily?”
    Y-54% N36% DK-10%

    Again, where was the “I’m not qualified to answer this question” option?

    Here’s a couple of doozies, and I really think we should highlight this stuff here:

    “I worry some genuine special needs
    children don’t get as much help as
    they need because resources are being
    diverted to children that don’t really
    need the help: How strongly do
    you agree or disagree?”

    This is a measurement of attitude. You would therefore expect a Likert scale to measure this. As a psychologist myself, I’d say that a scale of 0 to 4 (5-point Likert) would be minimally appropriate:
    0- strongly disagree
    1- disagree
    2- neutral
    3- agree
    4- strongly agree

    Strength of attitude is specified in the item, isn’t it?

    So a scale that has only Agree, Disagree, Neither and DK can hardly be accused of measuring the strength of an attitude. And ALL the key attitude questions suffer from this lack of an appropriate scale!

    Now, you mention, as one of the clear points that comes from this report, the following:

    “A lack of understanding of the reality of SEND by the questioner.”

    I’m not just inclined to agree with you: I’m lying on the floor in agreement, as it were. But there’s another, even more serious issue (and that one is bad enough!):

    1- This was a simple survey, and surveys are hard to mess up if you know what you’re doing – but they’re easy to butcher if you’re incompetent.
    2- This survey was overtly measuring the strengths of attitudes towards certain target objects, and the appropriate scale to use for this is a Likert scale (rather than the totally unreliable scale that was used).
    3- There was no actual mention of any sort of checking of these attitudes against even a sample of students identified as having SENDs let alone against such a group with a control group of non-SEND students.

    For a couple of years I was teaching elementary research methods and data analysis in the psychology of education as a visiting lecturer at my alma mater. These three faults in and of themselves completely void the reliability of the results they cite and the validity of any claims they are making.

    Barney Angliss’ bit in the report is … I have the utmost respect for Barney. I’ve interacted a lot with him and found him to be a very knowledgable chap, with excellent points to make. Let’s just say that the only thing he failed on was to disappoint. And that’s something we’d all want to fail on. Have you noticed that only his section of the publication is actually referenced?!

    Quality!

    The section written by Poppy Ionides is pretty sound. Becci Hawes’ section is reasonable indeed. I have only one problem with Tom Guy’s bit and that is the urge to avoid labels. But that is because many LEA educational psychologists might join in with that. For teachers who are not diagnosticians – yes, of course … don’t use labels that you don’t know the bases for. Leave that to the psychologists and medical practitioners. There is always a value in a label, despite what sociologists try to tell us. A correctly-used label is an indicator of whole set of issues that a student may be facing, and a good teacher will have identified those in order that the psychologist can make a reliable diagnosis and tie it to the most appropriate support strategy for that student. Used wrongly, as happens a lot in Finland (where I live and occasionally get to work!), the result can be disastrous. A diagnosis must always be expained.

    I’m not even sure what this report’s purpose was … except maybe to capitalise on people’s concerns regarding special educational needs and difficulties as a marketing ploy. The only good thing to come out of the report is that we get some very useful information and ideas from people who weren’t involved in the writing of the actual report or the poorly-designed survey on which it is based.

    Excellent blog post.

    1. Thank you – and YES! I’m no scientist, and I have only taken very baby steps into research through questionnaires, but even *I* could see that it had flaws.

      And the bits that were the headline grabbers? Not the knowledgeable parts written by experts, oh, no siree. The pretty pie charts.

      I reckon this is far more about selling a diagnostic screening test and pandering to the schools who might buy it, and *that*, my friend, is where the real story lies.

      That said, the more take-downs like yours, the better. I am simply not qualified, or knowledgeable enough to do it! 👍🏼

      1. Thank you! *blush*

        Glad I could be of used. Eighteen years in Finland as a foreigner with a disability and the authorities using denial of service to force me to leave … being able to add something to an excellent post on such an important issue did a lot to improve the day for me.

  2. A really useful and thought provoking blog. Whilst I do believe that access to diagnosis is important for some, I also believe that clinical diagnosis is subjective and flawed. There are ways of ensuring you get a diagnosis, and there are also ways in which the system can fob a parent off. If you are well informed about processes and options, you have a better chance of securing a diagnosis. In an education system with very limited resources, those with diagnoses based on a better informed approach will more easily gain access to services and support.

    There will be some children for whom the approach to clinicians is not as well considered, if there is any approach at all, and these children may not get access to services and support in school which will be allocated based on identified level of need – a lack of diagnosis may exclude them from some support.

    There will be a limited number of hours of support available and so a school ends up having to decide who is most in need, rather than identifying the support required by all and giving the right support to individuals. For me the whole system is deeply flawed and under-resourced and that is not due to teachers failing to understand clinical diagnosis (not in all cases anyway). We are a very long way from achieving equality of access, and with support in school being hugely underfunded I don’t see this getting fixed anytime soon.

      1. I can explain this to some extent. No two people see the same event in the same way. Perspectives – both mental and physical – will ensure that over, let’s say, twenty observations of a particular person in a particular setting there will be at least one observation upon which at least two simultaneous observers will differ with respect to how the observed behaviour should be coded. There’s a statistic that comes from this, called the inter-rater reliability coefficient, with the label k (for kappa): it is the quotient of two things – the difference between observed proportionate agreement and the probability of random agreement (in which we see the rate of chance agreements between two raters expressed as a decimal fraction of 1.0) and the difference between 1.0 and the probability of random agreement. This particular kappa statistic is Cohen’s kappa, and it is a useful measure of reliability of observations between two observers. Although no universal guidelines exist on interpretation of this statistic, it is usually accepted that: 0 < k < 0.40 shows poor reliability of the results, 0.40 ≤ k ≤ 0.75 means fair to good reliability, and 0.75 < k < 1.0 (a k = 1.0 state would raise suspicions, since that would be perfect reliability, and that so rare and occurence as to be deemed extremely improbable).

        If that made you dizzy, then good … that's the effect I wanted. Even the process of finding out how well two people agree with each other on what is what is actually really complicated. Most people, especially practitioners under hefty workloads, would become heavily burdened if they tried to do this for every case that encountered. So, they tend to go with a more 'intuitive' approach involving making a very subjective judgement about the significance of whatever it is they're trying to count. The problem here is that this will reduce the actual reliability of the results of the observation and therefore the validity of the diagnosis (it being a claim made on the basis of the results of the observation process) will suffer.

        We haven't even gotten into the Bayesian model of probability, which involves a probability based on subjective belief, yet! That one really is hard, but it leads us to a much more secure way of calculating the probability that what we think is presenting in someone is what is actually presenting. Phone apps such as Diagnose use this: this is because we have a system there that allows us to factor in prior probabilities in order to estimate posterior probabilities. These apps – and indeed any diagnostic work-up that is performed using this Bayesian model, will be somewhat more reliable than the inter-rater reliability statistic and much more reliable than would the intuitive approach.

        For this reason, clinical diagnoses are inherently subjective. Not sure if this is what headucator meant, but this is what it is from a practitioner persṕective.

      2. Love this – It’s been a while since I’ve been exposed to proper Psychology! It would be great if parents all had access to the same advice and guidance in the first instance, before trying to navigate the system, but even then it would be applied differently, and interpreted differently.

      3. On reflection I should probably direct the subjective and flawed comment towards ‘access to clinical diagnosis’. I have worked with parents who have been obstructed at various points in the process. I have also worked with parents who have had the right advice and support and been able to navigate the system.

        A lack of resources everywhere means that there is lots of gate keeping going on.

        This means that there can also be a time lag for families who perhaps don’t have all the right advice, or the same skills to gain access.

        I think access is the biggest issue.

        I have seen unusual decisions where no diagnosis has been given first time around, and then with additional advice, a diagnosis materialises second time around.

      1. There’s the thing … because most practitioners are using a very ‘intuitive’ approach … the process of gaining access to proper diagnostic services is made difficult.

    1. Going back to what I was saying about the subjectivity issue in diagnosis:
      https://www.ncbi.nlm.nih.gov/pubmed/28233080
      —–
      Previous research shows inconsistency in clinician-assigned diagnoses of Autism Spectrum Disorder (ASD). We conducted an exploratory study that examined the concordance of diagnoses between a multidisciplinary assessment team and a range of independent clinicians throughout Australia. Nine video-taped Autism Diagnostic Observation Schedule (ADOS) assessments were collected from two Australian sites. Twenty-seven Australian health professionals each observed two video-recordings and rated the degree to which the individual met the DSM-5 criteria for ASD. There was 100% agreement on the diagnostic classification for only 3 of the 9 video clips (33%), with the remaining 6 clips (66%) reaching poor reliability. In addition, only 24% of the participating clinicians achieved ‘good’ or ‘excellent’ levels of agreement (Cohen’s kappa > 0.6) with the original ASD assessment. These findings have implications for clinical guidelines for ASD assessments.
      —–

      This is why it’s hard for autistic people to get diagnoses at all, let alone get them taken seriously. I may have more to say later.

  3. A great response Nancy. One of my big issues was the lack of context around what is a complicated and sensitive subject for many SEN parents, and teachers, who struggle to get the right support for a child’s needs.

    The second is the lack of quantification – using terms such as “some parents” is in my view misleading. This could range anywhere from 2 to 2,000 – and therefore risks making this look like a much bigger issue that it really is.

    There were only 800 teachers out of 512,000 teachers – can we be sure that their views are representative of all teachers?

    If each of the 54% of the 800 teachers that agreed with SEN misdiagnosis due to parental pressure, felt this way about just 2 parents the implications are very different to the level of issue being reported.

    This works out to 846 parents. Assuming avg. of 2 children with needs per parents that is 1,692 children. In 2016 there were 1,228,785* children with SEN needs, not including the many children who have not been able to get a diagnosis.

    Not such a big headline if you start to look at the maths of how many children out of the total population this applies to.

    Granted that the number of children that teachers feel this way about may be higher, but without the numbers being captured in the report – the actual number of children that this potentially applies to is anyones guess.

    * sourced from https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/539158/SFR29_2016_Main_Text.pdf

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s