Just finished the IAPHR conference in St Louis. Twas nice to get to present copies of my book to the student presenters.
Not sure my usually combative style went down well….I may have been on a mood high….swung to a low right now so have avoided going out with the others for a final pizza.
Biplor 2 – good article in the Telegraph. That’s enough for today. Will probably be quiet for a few days….normal service will be resumed
soon enough. when I can.
In Birmingham for two days for the end-of-life (European) project concluding meeting plus advisory (steering) group meeting.
Truth be told, wish I weren’t. The Generalised Anxiety Disorder is out in full force at the moment so I am not really in a socialising mood.
Oh well. That’s life. This is my final obligation in terms of academic projects. From here on in you pay for my advice. Saw someone I see not enough of, which was great – I think she will be a useful source of advice on the consultancy front.
Otherwise it is:
- Former PhD advisor (supervisor)
- Former boss
- Former postdoc
- Some others I work indirectly with
Apologies for the downbeat post but I figured that as an “out mental healther” I should give the full story where possible.
Not a lot to report I’m afraid.
I have booked everything for the IAHPR conference in OCtober (St Louis, USA) so anyone going, please say hello.
Not heard of any delays to the book publication, so I hope September is still the launch!
I wrote my first blog posting on my company site, TF Choices Ltd, the other day. It is a slightly whimsical piece about the perils of poor study design based on my experience with a rather popular game app for phones and tablet. I am not naming and shaming the game 😉
Have begun to think about the latter part of this year and IAHPR conference – looking forward to it!
Other things in the pipeline that I will announce soon.
I’ve just got back to the UK from Amsterdam where I attended the first International Academy of Health Preference Research conference.
It was one of the best conferences I’ve ever attended. A collegiate atmosphere where new and established researchers could present work in stated preference research in a friendly non-confrontational environment. The meals out (and alcohol) certainly helped establish new collaborations too!
Thanks so much to Ben Craig for organising this – I hope it becomes the premier forum for health stated preference research. In fact, I’ll be up front. In 10 years time I hope all stated preference health researchers attend two conferences:
(1) The international choice modelling conference (for general choice modelling issues);
(2) IAHPR (for health specific issues).
The business meeting was interesting. There was an attempt to move the next meeting date (despite there being very good reasons why it couldn’t realistically be changed) so it wouldn’t clash with SMDM short course dates. I made my feelings known – that I don’t think SMDM and IAHPR compete – I think (unfortunately) that SMDM is becoming more like ISPOR. IAHPR is a different forum from them both – focusing on methodology and academic rigour, rather than commercial/funding issues. I have to say it – I think SMDM is yet to establish a niche for itself in stated preference research that is different from:
(1) ISPOR (commercial/FDA approval stuff) and
(2) IAHPR (methodology).
I made my feelings known in a survey of attendees of SMDM. Frankly I think their conference in Miami had a lot to be desired. Now I know that there are political reasons to stay “on side” with certain academics who sit on US boards when it comes to ISPOR etc. Frankly, I think that view is misguided. So I’m going to carry on doing what I’m doing.
Not many posts recently I know – I’ve been in Europe doing various things that I’ll make public soon 🙂
Travelling to Amsterdam for the inaugural IAHPR conference on Friday. I’ll be presenting our initial work on response times in validating DCE estimate for the Australian end-of-life study. Then I’ll be travelling back to Sydney via Singapore.
My trip to Miami to present at the SMDM conference in October is booked. Arrive Saturday 18th (flight from London) and depart a bit early – evening of Tuesday 21st. My presentation is on Monday and I think I’ll catch the stated preference session early tuesday afternoon before I rush to the airport. Flying via Charlotte, NC (which I think I’ve used before, but only on internal flights to/from Raleigh/Durham) and staying at a Holiday Inn Express a kilometre from the conference hotel (which was too expensive for me).
I’ll book Amsterdam in due course – hopefully flying from East Midlands Airport with FlyBe again. Both conferences are being done from a master “UK trip” to keep costs (and, I hope, stress levels) down! So I’m in nominally in Europe for a month from mid October.
I’ve had my abstracts on the supplementation of Case 1 Best-Worst Scaling (BWS) data with response time data accepted for both the IAHPR and SMDM conferences – woohoo!
This is extremely important work, done with collaborators at Newcastle University, NSW, Australia. It is the first validation of stated preference methods using physiological data – in this case, how long people take to choose their most and least preferred option. That research on response times uses the linear ballistic accumulator model – a model that has passed rigorous testing all the way from the lab with animals up through psych students, into real life looking at mobile phone choices, and now finally ascertaining how long it takes people to agree/disagree with attitudinal statements about end-of-life care.
This work is particularly interesting, as it appears, at first glance, to give insights into the “fast” and “slow” decision-making styles hypothesised by Daniel Kahneman. These styles will be crucial in helping clinicians and allied health professionals working in eliciting advance care plans – in particular, helping them understand what type of decision style influenced a particular choice made by a patient. We don’t, at this stage, make the normative judgements as to which decision-making style is “most valid”! But it gives the information necessary for society to decide what type should be used to construct an ACP.
The preliminary work showing the choice and attitudinal research results has a revise and resubmit to a journal and some similar work is in press in the best-worst scaling book (forthcoming, CUP).
My presentation yesterday at the Centre for Applied Disability Research conference seemed to go down well. There was one comment that I get on a frequent basis so I thought I’d give a more complete answer here. The question is always a variant on the following:
“Why should we use ICECAP-O or one of those instruments you’re touting when there’s the WHOQoL instrument or instrument x/y/z that has been validated already?”
A tag-on is often to the effect “how can your 5 item instrument beat our 10/15/50 item instrument which is bound to be better for individuals?”
Well the answer is simple – it comes down to a difference in paradigm, in particular the difference between psychometrics and random utility theory. I could be rude at conferences (but don’t) to counter the (occasionally a tad aggressive) attacks I get on ICECAP-O for “not being individual-specific”. My “slightly aggressive” response would be “actually it’s YOUR instrument that isn’t individual specific – psychometrics isn’t about the individual, it uses differences BETWEEN individuals to validate the response categories. Random Utility Theory (RUT) is EXPLICITLY a theory of how the individual makes choices and as such, any instrument based on it is by definition an individual level QoL instrument! For ICECAP-O (or any of the other instruments in the ICECAP family or the CES) I could, in theory, give any respondent THEIR OWN set of scores (a “tariff” to use health economics parlance), if they do the choice experiment. You CANNOT do that with existing instruments, with the exception of some health-based ones that use the time trade-off/standard gamble, IF they’d asked the right set of questions to concentrate on individual level scores.
This individual respondent tariff reflects the trade-offs THAT INDIVIDUAL would make between the items and how bad the various impairments are to that person. You can’t get that from any of these instruments I hear touted as being “superior” since they were validated on the basis of between person differences – by definition they cannot be tested/validated at the individual level. (Not least because there are no scores at that level, certainly not preference based ones that reflect how bad the impairments all are on a common scale).
So ICECAP-O and the other instruments beat them all when it comes to the issue of “the individual level”. We can feed back individual level scores – and indeed we did, for the end-of-life care survey, which you too can do if you click on surveys and go back a page or two. So not only are there 4 to the power 5 (1024) distinct utility values available – it is the PROFILE defined by the set of 5 answers that matters, not the number of questions – but these 1024 scores could be individual specific if we wanted. Indeed Chapter 12 of the forthcoming best-worst scaling (BWS) book (Louviere, Flynn & Marley – Best-Worst Scaling: Theory & Applications, CUP) will present subgroup Australian tariffs for ICECAP-O.
“Testing” ICECAP-O using psychometric based techniques may be invalid – I’m not sure – but one thing I am sure of. Stop throwing mud at us for having an instrument that is “obviously worse” than these existing large-item questionnaires – because I KNOW for a fact you’ve not tested the individual level properties of the scoring of these. At best we can all agree a truce and say there are two differing paradigms in use here and at present there’s been no properly designed study that uses a common denominator on which I could compare them.
The first draft of the entire best-worst scaling book is finished! I’m now making some edits requested by co-authors but it is basically there 🙂
Am also finishing up on some papers I am co-authoring this week and writing a couple of abstracts for a conference.
By the end of the month I should be in a position to sign off work-wise for a couple of months (as I wanted), take holidays in Europe, see family and friends etc.