Graeme MacLennan
banner
gsmaclennan.bsky.social
Graeme MacLennan
@gsmaclennan.bsky.social

CTU Director and Statistician

Public Health 32%
Economics 23%

So it's your kids eh?

Reposted by Graeme MacLennan

We are recruiting an information systems manager to support our data management and stewardship efforts. 13-month contract on 53k/year, but I'll be looking to extend the role contingent on renewed funding for our CRF. Small, close-knit team + flexible working arrangements.
' | HRB Clinical Research Facility - University College Cork
New job opportunity with the CRF-UCC: Information Systems Manager The CRF-UCC is #hiring an IS Manager to continue to grow Electronic Data Capture supports offered through the Statistics, Data & Anal...
www.linkedin.com

It's turtles. All the way down, turtle after turtle.
One of my 2026 goals is to be more present on BlueSky! Another is to recruit a great candidate to this 4-yr PhD studentship with us at the Aberdeen Centre for Evaluation -applying behavioural science to the design and delivery of clinical trials. Please share. www.findaphd.com/phds/project...
Improving health by improving trials – applying behavioural science to improve the design and delivery of clinical trials. at University of Aberdeen on FindAPhD.com
PhD Project - Improving health by improving trials – applying behavioural science to improve the design and delivery of clinical trials. at University of Aberdeen , listed on FindAPhD.com
www.findaphd.com

British government, still shit after 250 odd years.

Which means we can write the perfectly correct sentence:

Yous use ewes' yews.

To the tune of Disney Beaty and Beast

Trick as old as time,
Error all too clear,
Measure each group twice,
Claiming truth from lies,
Sig p-values appear.
Royal Statistical Society Publications
In 2007 No. 7 Protect & Perfect Beauty Serum became Boots'fastest ever selling product after a BBC2 Horizon programme reported on a clinical trial of its effectiveness: here, unusually, was a scienti...
rss.onlinelibrary.wiley.com

Boston t-test party?

Reposted by Graeme MacLennan

The recording of this webinar is now available on the @nihr-rss.bsky.social YouTube channel: www.youtube.com/watch?v=4cNz...

If it's not an omnishambles, there's still hope.

Better to be lucky than good (looking).

Lefty Gomac.

Hello from UK. MSc in Applied Stats, no PhD, 27 years at University of Aberdeen, no barrier to progression career-wise. Caveat, all my stats colleagues have PhDs except immediate post-grads. Perhaps they keep me around as a novelty. Happy to bore you with tales from journey without PhD.
We're currently looking for a Trials Methodologist/Senior Trials Methodologist to join our team.

You can find out more about the role and how to apply here:
jobs.icr.ac.uk/vacancies/13...

#ResearchJobs #TrialsMethodology🧪
Trials Methodologist/ Senior Trials Methodologist (Statistics) in Sutton | The Institute of Cancer Research
View details and apply for this Trials Methodologist/ Senior Trials Methodologist (Statistics) vacancy in Sutton. Salary : Commencement on the salary range is subject to comparable skills an...
jobs.icr.ac.uk

Amen to that.

Hard for like for sending me back to the 80s playing Space Invaders 2 at Inverness La Scala cinema at 10p a go. Shit, I am getting old.

My tuppence: Not only is it allowable, it is desirable when appropriate. But I'd replace negative with challenge, acknowledge, and discount anything not useful (respectfully) when required. Otherwise our PPI colleagues are infallible.
We are co-authors on 2 new papers in @bonejointjournal.bsky.social - one on the potential of Bayesian trials in T&O and a Bayesian reanalysis of the WHiTE8 trial.
And … our own Jemma Hudson is the featured author 😊
@docfarrow.bsky.social
@marionkcampbell.bsky.social

Links to papers below 👇

I will email you about this later. Much later.

Reposted by Graeme MacLennan

The number needed to treat: It is time to bow out gracefully - Journal of Clinical Epidemiology www.jclinepi.com/article/S089...
The number needed to treat: It is time to bow out gracefully
The number needed to treat (NNT) has statistical and methodological limitations that make it unhelpful in the context of developing clinical practice guidelines and policy decisions.It is not unusual ...
www.jclinepi.com

Disappointed not to see your correct new title "Unprincipled Statistician".

Disgusted etc.

Reposted by Graeme MacLennan

Job available for a statistician to work on a clinical trials methodology project on improving statistical methods for paediatric cancer ttrials.

Cancer Research UK Clinical Trials Unit, University of Birmingham

Closes 22 September

edzz.fa.em3.oraclecloud.com/hcmUI/Candid...
Biostatistician - Cancer Research UK Clinical Trials Unit - 105965 - Grade 7
This post is for a Biostatistician to be the main statistical scientist for a methodological project on improving statistical methods for the design and analysis of clinical trials in paediatric cance...
edzz.fa.em3.oraclecloud.com

Not even an open coffin

Standard amateur fitba (soccer) sledge might work here: If you want my shirt just ask, you can have it after the game.

Reposted by Graeme MacLennan

I think some of the items included in some trustworthiness tools must also be criticised. e.g. declaring a trial problematic if they failed to meet the target sample size, or declaring a study problematic if it was placebo-controlled but used sealed envelopes (?)

Reposted by Graeme MacLennan

#OTD 1667 John Arbuthnot bapt (d 27 Feb 1735) ‘Physitian in Ordinary’ to Queen Anne, his 1692 translation of Huygens's ‘De Ratiociniis in Ludo Aleae’ was the first work on probability in English. He also developed the first formal significance test (1710) /7🧵

Reposted by Graeme MacLennan

1/2 Snapshots of ACE/CHaRT at the magnificent at Tall Ships 2025⚓. Visitors have fun with our “Lime or Brine” trial! bit.ly/4lKTD1o

That happens. Sometimes.

Is that Latin for "mess about find out"?

excuse the "it's" of course.

you seem to be saying that if we could work out when

θ(t)=θ_0 cos(2π/T)t)

Where: θ(t) = angular displacement at time 𝑡
θ_0 = amplitude
T = period of the pendulum

was at it's maximum, then look at time t we'd always get a good rank...