SlideShare a Scribd company logo
1 of 15
What is the point of
point estimates?
Stephen Senn
Consultant Statistician
Edinburgh
(C)
Stephen
Senn
2022
1
“Superficially, point estimation may
seem a simpler problem to discuss
than that of interval estimation”
Cox and Hinckley, p250
Apologies
• This was originally submitted to the
Communicating Statistical Methods session
• But it was not accepted 
• Due to a cancellation from the Cluster Trials
session I am able to present it here
• I have departed from the abstract somewhat
to introduce a cluster trial example
• By doing this I shall annoy two sets of people
• Those who wanted to hear what was in the
abstract
• Those who want to hear about up-to-date
research on cluster trials
(C) Stephen Senn 2022 2
Being befuddled about balance
• A matter that often befuddles amateur (non-statistical) commentaries
on randomised studies is the role of balance
• If factors are balanced, adjusting for them (in the linear case) won’t
change the estimate so what’s the point?
• However, the statistical point of view is more or less the opposite.
• If you have balanced for a factor you ought to have the factor in the
model
• We don’t analyse a matched pairs design like a completely randomised one
• Why?
(C) Stephen Senn 2022 3
As we all know
• The reason is to do with the standard error
• Balancing by a factor eliminates its effect on the estimate
• But it does not, of its own, eliminate its effect on the standard error
• In fact it increases it slightly
• To estimate how well you have done, you need to remove its effect on
the standard error
• This requires putting it in the model
• We all know this but one still regularly encounters arguments that
show we have forgotten
• In my opinion this is because we obsess about point estimates
(C) Stephen Senn 2022 4
An incomplete block design in asthma
Two formulations of formoterol given at different doses
• Details need not concern us
• Some patients received
• ISF24 and MTA6
• Some patients received
• ISF24 but not MTA6
• Some patients received
• MTA6 but not ISF24
• I can calculate both within subject
and between subject estimates of
the effect ISF24-MTA6
• This is not the way I would analyse
these data but I could (and have)
(C) Stephen Senn 2022 5
Data from Senn et
al, 1997
More or Less
More Less
(C) Stephen Senn 2022 6
The Lanarkshire Milk Experiment
An incomplete block cluster allocated trial from 1930
• 67 Schools enrolled
• In some schools children were
allocated 1:1 either to act as
controls or receive raw milk
• In other schools children were
allocated 1:1 either to act as
controls or receive pasteurised milk
• Just over 18,000 schoolchildren
studied
• Four months of treatment
(C) Stephen Senn 2022 7
Analysis
What the authors did
• Analysed data on weight and height
• Pooled all controls
• Not a good idea
• But analysed each of 14 = 2 sex by 7
age groups separately
• Calculated ‘probable errors’
• But it is not clear how
• Concluded that milk was beneficial
but that one could not decide
whether raw or pasteurised was
better
(C) Stephen Senn 2022 8
Prominent commentators
Fisher Student
What I have done
• Developed algebraic expressions for variances of contrasts for six different
types of analysis
• Four that might be reasonable
• Two that the authors carried out
• For each of the six analyses there are two variances
• The true variance
• What one might naively estimate
• Simulated to check my formulae
• Everything looks fine
• This means
• Either it is fine
• Or I made the same mistakes in my simulation as in the theory
(C) Stephen Senn 2022 9
What I have not done
• I have not got hold of the original data
• In fact we don’t even know how many schools of each type there
were
• Therefore I have had to speculate (postulate, guess, fake) what the
variance components might be
• The key issue as all those interested in cluster allocate trials know is
the ratio of the between to the within cluster variance
• I don’t like ICCs
(C) Stephen Senn 2022 10
Milking some
data
Parameter settings are
identical for all six cases
Heights in inches are
considered
The formulae are for a
given sex by age group and
n=10 per group per school
33 schools per milk type
in a & b the school effect is
fixed
in e & f the school effect is
random
in c & d the school effect is
ignored
Sacred cow
The TARGET study
Target study
• One of the largest studies ever run in
osteoarthritis
• 18,000 patients
• Randomisation took place in two sub-
studies of equal size
• Lumiracoxib versus ibuprofen
• Lumiracoxib versus naproxen
• Practical considerations dictated design
• Purpose to investigate CV and GI
tolerability of lumiracoxib
• Sub-study effect explicitly dealt with in
analysis
Lanarkshire Milk Study
• At the time one of the largest nutritional
studies
• 18,000 school children
• Randomisation took place in two sub-
studies of equal size
• No milk versus raw milk
• No milk versus pasteurised milk
• Practical considerations dictated design
• Purpose to investigate effect of milk on
height and weight
• Sub-study effect ignored in analysis
(C) Stephen Senn 2022 12
(c) Stephen Senn 2012
Baseline Demographics
Sub-Study 1 Sub Study 2
Demographic
Characteristic
Lumiracoxib
n = 4376
Ibuprofen
n = 4397
Lumiracoxib
n = 4741
Naproxen
n = 4730
Use of low-dose
aspirin
975 (22.3) 966 (22.0) 1195 (25.1) 1193 (25.2)
History of
vascular disease
393 (9.0) 340 (7.7) 588 (12.4) 559 (11.8)
Cerebro-
vascular disease
69 (1.6) 65 (1.5) 108 (2.3) 107 (2.3)
Dyslipidaemias 1030 (23.5) 1025 (23.3) 799 (16.9) 809 (17.1)
Nitrate use 105 (2.4) 79 (1.8) 181 (3.8) 165 (3.5)
(C) Stephen Senn 2022 13
If this looks familiar it ought to.
Bergen 2012
Baseline Deviances
Model Term
Demographic
Characteristic
Sub-study
(DF=1)
Treatment
given Sub-
study
(DF=2)
Treatment
(DF=2)
Use of low-dose
aspirin
23.57 0.13 13.40
History of
vascular disease
70.14 5.23 47.41
Cerebro-
vascular disease
13.54 0.14 7.75
Dyslipidaemias 117.98 0.17 54.72
Nitrate use 39.83 4.62 29.17
(C) Stephen Senn 2022 14
Some final words
• The very impressive and interesting causal
inference school seems to be promoting
identifiability rather than estimability
• The former considers whether what happens
asymptotically is correct
• It is not always obvious what has to go to infinity
for an asymptote to be reached
• Schools/centres or pupils/patients
• I think we are in danger of loosing a valuable
insight from experimental design theory
• How treatments are varied across the block
structure matters
• The lessons of the Rothamsted School should
be heeded
(C) Stephen Senn 2022 15
stephen@senns.uk
@stephensenn
http://www.senns.uk/Blogs.html
“..the calculation of standard errors is
idle and misleading, if the method of
arrangement adopted fails to guarantee
their validity…”
RA Fisher, The Design of Experiments
section 34
Fisher Yates Nelder

More Related Content

What's hot

Shrinkage in medical prediction: the poor man’s solution for an inadequate sa...
Shrinkage in medical prediction: the poor man’s solution for an inadequate sa...Shrinkage in medical prediction: the poor man’s solution for an inadequate sa...
Shrinkage in medical prediction: the poor man’s solution for an inadequate sa...Maarten van Smeden
 
Minimally important differences v2
Minimally important differences v2Minimally important differences v2
Minimally important differences v2Stephen Senn
 
The replication crisis: are P-values the problem and are Bayes factors the so...
The replication crisis: are P-values the problem and are Bayes factors the so...The replication crisis: are P-values the problem and are Bayes factors the so...
The replication crisis: are P-values the problem and are Bayes factors the so...StephenSenn2
 
QUANTIFYING THE IMPACT OF DIFFERENT APPROACHES FOR HANDLING CONTINUOUS PREDIC...
QUANTIFYING THE IMPACT OF DIFFERENT APPROACHES FOR HANDLING CONTINUOUS PREDIC...QUANTIFYING THE IMPACT OF DIFFERENT APPROACHES FOR HANDLING CONTINUOUS PREDIC...
QUANTIFYING THE IMPACT OF DIFFERENT APPROACHES FOR HANDLING CONTINUOUS PREDIC...GaryCollins74
 
Clinical trials are about comparability not generalisability V2.pptx
Clinical trials are about comparability not generalisability V2.pptxClinical trials are about comparability not generalisability V2.pptx
Clinical trials are about comparability not generalisability V2.pptxStephenSenn3
 
NNTs, responder analysis & overlap measures
NNTs, responder analysis & overlap measuresNNTs, responder analysis & overlap measures
NNTs, responder analysis & overlap measuresStephen Senn
 
Prognosis-based medicine: merits and pitfalls of forecasting patient health
Prognosis-based medicine: merits and pitfalls of forecasting patient healthPrognosis-based medicine: merits and pitfalls of forecasting patient health
Prognosis-based medicine: merits and pitfalls of forecasting patient healthMaarten van Smeden
 
Development and evaluation of prediction models: pitfalls and solutions
Development and evaluation of prediction models: pitfalls and solutionsDevelopment and evaluation of prediction models: pitfalls and solutions
Development and evaluation of prediction models: pitfalls and solutionsMaarten van Smeden
 
Clinical prediction models: development, validation and beyond
Clinical prediction models:development, validation and beyondClinical prediction models:development, validation and beyond
Clinical prediction models: development, validation and beyondMaarten van Smeden
 
Correcting for missing data, measurement error and confounding
Correcting for missing data, measurement error and confoundingCorrecting for missing data, measurement error and confounding
Correcting for missing data, measurement error and confoundingMaarten van Smeden
 
Trials on trial: equivalence, ethics and evidence
Trials on trial: equivalence, ethics and evidenceTrials on trial: equivalence, ethics and evidence
Trials on trial: equivalence, ethics and evidenceCochrane.Collaboration
 
The basics of prediction modeling
The basics of prediction modeling The basics of prediction modeling
The basics of prediction modeling Maarten van Smeden
 
Introduction to prediction modelling - Berlin 2018 - Part I
Introduction to prediction modelling - Berlin 2018 - Part IIntroduction to prediction modelling - Berlin 2018 - Part I
Introduction to prediction modelling - Berlin 2018 - Part IMaarten van Smeden
 
Statistics and ML 21Oct22 sel.pptx
Statistics and ML 21Oct22 sel.pptxStatistics and ML 21Oct22 sel.pptx
Statistics and ML 21Oct22 sel.pptxEwout Steyerberg
 
Regression shrinkage: better answers to causal questions
Regression shrinkage: better answers to causal questionsRegression shrinkage: better answers to causal questions
Regression shrinkage: better answers to causal questionsMaarten van Smeden
 
Guideline for high-quality diagnostic and prognostic applications of AI in he...
Guideline for high-quality diagnostic and prognostic applications of AI in he...Guideline for high-quality diagnostic and prognostic applications of AI in he...
Guideline for high-quality diagnostic and prognostic applications of AI in he...Maarten van Smeden
 
Personalised medicine a sceptical view
Personalised medicine a sceptical viewPersonalised medicine a sceptical view
Personalised medicine a sceptical viewStephen Senn
 
Is it causal, is it prediction or is it neither?
Is it causal, is it prediction or is it neither?Is it causal, is it prediction or is it neither?
Is it causal, is it prediction or is it neither?Maarten van Smeden
 

What's hot (20)

Shrinkage in medical prediction: the poor man’s solution for an inadequate sa...
Shrinkage in medical prediction: the poor man’s solution for an inadequate sa...Shrinkage in medical prediction: the poor man’s solution for an inadequate sa...
Shrinkage in medical prediction: the poor man’s solution for an inadequate sa...
 
Minimally important differences v2
Minimally important differences v2Minimally important differences v2
Minimally important differences v2
 
Predictimands
PredictimandsPredictimands
Predictimands
 
The replication crisis: are P-values the problem and are Bayes factors the so...
The replication crisis: are P-values the problem and are Bayes factors the so...The replication crisis: are P-values the problem and are Bayes factors the so...
The replication crisis: are P-values the problem and are Bayes factors the so...
 
QUANTIFYING THE IMPACT OF DIFFERENT APPROACHES FOR HANDLING CONTINUOUS PREDIC...
QUANTIFYING THE IMPACT OF DIFFERENT APPROACHES FOR HANDLING CONTINUOUS PREDIC...QUANTIFYING THE IMPACT OF DIFFERENT APPROACHES FOR HANDLING CONTINUOUS PREDIC...
QUANTIFYING THE IMPACT OF DIFFERENT APPROACHES FOR HANDLING CONTINUOUS PREDIC...
 
Clinical trials are about comparability not generalisability V2.pptx
Clinical trials are about comparability not generalisability V2.pptxClinical trials are about comparability not generalisability V2.pptx
Clinical trials are about comparability not generalisability V2.pptx
 
NNTs, responder analysis & overlap measures
NNTs, responder analysis & overlap measuresNNTs, responder analysis & overlap measures
NNTs, responder analysis & overlap measures
 
Prognosis-based medicine: merits and pitfalls of forecasting patient health
Prognosis-based medicine: merits and pitfalls of forecasting patient healthPrognosis-based medicine: merits and pitfalls of forecasting patient health
Prognosis-based medicine: merits and pitfalls of forecasting patient health
 
Development and evaluation of prediction models: pitfalls and solutions
Development and evaluation of prediction models: pitfalls and solutionsDevelopment and evaluation of prediction models: pitfalls and solutions
Development and evaluation of prediction models: pitfalls and solutions
 
Clinical prediction models
Clinical prediction modelsClinical prediction models
Clinical prediction models
 
Clinical prediction models: development, validation and beyond
Clinical prediction models:development, validation and beyondClinical prediction models:development, validation and beyond
Clinical prediction models: development, validation and beyond
 
Correcting for missing data, measurement error and confounding
Correcting for missing data, measurement error and confoundingCorrecting for missing data, measurement error and confounding
Correcting for missing data, measurement error and confounding
 
Trials on trial: equivalence, ethics and evidence
Trials on trial: equivalence, ethics and evidenceTrials on trial: equivalence, ethics and evidence
Trials on trial: equivalence, ethics and evidence
 
The basics of prediction modeling
The basics of prediction modeling The basics of prediction modeling
The basics of prediction modeling
 
Introduction to prediction modelling - Berlin 2018 - Part I
Introduction to prediction modelling - Berlin 2018 - Part IIntroduction to prediction modelling - Berlin 2018 - Part I
Introduction to prediction modelling - Berlin 2018 - Part I
 
Statistics and ML 21Oct22 sel.pptx
Statistics and ML 21Oct22 sel.pptxStatistics and ML 21Oct22 sel.pptx
Statistics and ML 21Oct22 sel.pptx
 
Regression shrinkage: better answers to causal questions
Regression shrinkage: better answers to causal questionsRegression shrinkage: better answers to causal questions
Regression shrinkage: better answers to causal questions
 
Guideline for high-quality diagnostic and prognostic applications of AI in he...
Guideline for high-quality diagnostic and prognostic applications of AI in he...Guideline for high-quality diagnostic and prognostic applications of AI in he...
Guideline for high-quality diagnostic and prognostic applications of AI in he...
 
Personalised medicine a sceptical view
Personalised medicine a sceptical viewPersonalised medicine a sceptical view
Personalised medicine a sceptical view
 
Is it causal, is it prediction or is it neither?
Is it causal, is it prediction or is it neither?Is it causal, is it prediction or is it neither?
Is it causal, is it prediction or is it neither?
 

Similar to What is the point of point estimates

What is your question
What is your questionWhat is your question
What is your questionStephenSenn2
 
What is your question
What is your questionWhat is your question
What is your questionStephen Senn
 
The challenge of small data
The challenge of small dataThe challenge of small data
The challenge of small dataStephen Senn
 
Thinking statistically v3
Thinking statistically v3Thinking statistically v3
Thinking statistically v3Stephen Senn
 
Clinical trials: quo vadis in the age of covid?
Clinical trials: quo vadis in the age of covid?Clinical trials: quo vadis in the age of covid?
Clinical trials: quo vadis in the age of covid?Stephen Senn
 
In search of the lost loss function
In search of the lost loss function In search of the lost loss function
In search of the lost loss function Stephen Senn
 
To infinity and beyond
To infinity and beyond To infinity and beyond
To infinity and beyond Stephen Senn
 
To infinity and beyond v2
To infinity and beyond v2To infinity and beyond v2
To infinity and beyond v2Stephen Senn
 
Numbers needed to mislead
Numbers needed to misleadNumbers needed to mislead
Numbers needed to misleadStephen Senn
 
Seventy years of RCTs
Seventy years of RCTsSeventy years of RCTs
Seventy years of RCTsStephen Senn
 
Audit and stat for medical professionals
Audit and stat for medical professionalsAudit and stat for medical professionals
Audit and stat for medical professionalsNadir Mehmood
 
Is ignorance bliss
Is ignorance blissIs ignorance bliss
Is ignorance blissStephen Senn
 
Understanding randomisation
Understanding randomisationUnderstanding randomisation
Understanding randomisationStephen Senn
 
The Seven Habits of Highly Effective Statisticians
The Seven Habits of Highly Effective StatisticiansThe Seven Habits of Highly Effective Statisticians
The Seven Habits of Highly Effective StatisticiansStephen Senn
 

Similar to What is the point of point estimates (20)

What is your question
What is your questionWhat is your question
What is your question
 
What is your question
What is your questionWhat is your question
What is your question
 
The challenge of small data
The challenge of small dataThe challenge of small data
The challenge of small data
 
Yates and cochran
Yates and cochranYates and cochran
Yates and cochran
 
Thinking statistically v3
Thinking statistically v3Thinking statistically v3
Thinking statistically v3
 
Clinical trials: quo vadis in the age of covid?
Clinical trials: quo vadis in the age of covid?Clinical trials: quo vadis in the age of covid?
Clinical trials: quo vadis in the age of covid?
 
MD poverty indexes
MD poverty indexesMD poverty indexes
MD poverty indexes
 
In search of the lost loss function
In search of the lost loss function In search of the lost loss function
In search of the lost loss function
 
To infinity and beyond
To infinity and beyond To infinity and beyond
To infinity and beyond
 
To infinity and beyond v2
To infinity and beyond v2To infinity and beyond v2
To infinity and beyond v2
 
Numbers needed to mislead
Numbers needed to misleadNumbers needed to mislead
Numbers needed to mislead
 
Overview of statistical tests: Data handling and data quality (Part II)
Overview of statistical tests: Data handling and data quality (Part II)Overview of statistical tests: Data handling and data quality (Part II)
Overview of statistical tests: Data handling and data quality (Part II)
 
Seventy years of RCTs
Seventy years of RCTsSeventy years of RCTs
Seventy years of RCTs
 
Basics of SPSS, Part 1
Basics of SPSS, Part 1Basics of SPSS, Part 1
Basics of SPSS, Part 1
 
Audit and stat for medical professionals
Audit and stat for medical professionalsAudit and stat for medical professionals
Audit and stat for medical professionals
 
Is ignorance bliss
Is ignorance blissIs ignorance bliss
Is ignorance bliss
 
TCI in primary care - SEM (2006)
TCI in primary care - SEM (2006)TCI in primary care - SEM (2006)
TCI in primary care - SEM (2006)
 
6 sr and meta analysis-ayurved
6 sr and meta analysis-ayurved6 sr and meta analysis-ayurved
6 sr and meta analysis-ayurved
 
Understanding randomisation
Understanding randomisationUnderstanding randomisation
Understanding randomisation
 
The Seven Habits of Highly Effective Statisticians
The Seven Habits of Highly Effective StatisticiansThe Seven Habits of Highly Effective Statisticians
The Seven Habits of Highly Effective Statisticians
 

Recently uploaded

Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...dajasot375
 
Low Rate Call Girls Bhilai Anika 8250192130 Independent Escort Service Bhilai
Low Rate Call Girls Bhilai Anika 8250192130 Independent Escort Service BhilaiLow Rate Call Girls Bhilai Anika 8250192130 Independent Escort Service Bhilai
Low Rate Call Girls Bhilai Anika 8250192130 Independent Escort Service BhilaiSuhani Kapoor
 
From idea to production in a day – Leveraging Azure ML and Streamlit to build...
From idea to production in a day – Leveraging Azure ML and Streamlit to build...From idea to production in a day – Leveraging Azure ML and Streamlit to build...
From idea to production in a day – Leveraging Azure ML and Streamlit to build...Florian Roscheck
 
Aminabad Call Girl Agent 9548273370 , Call Girls Service Lucknow
Aminabad Call Girl Agent 9548273370 , Call Girls Service LucknowAminabad Call Girl Agent 9548273370 , Call Girls Service Lucknow
Aminabad Call Girl Agent 9548273370 , Call Girls Service Lucknowmakika9823
 
Call Girls in Defence Colony Delhi 💯Call Us 🔝8264348440🔝
Call Girls in Defence Colony Delhi 💯Call Us 🔝8264348440🔝Call Girls in Defence Colony Delhi 💯Call Us 🔝8264348440🔝
Call Girls in Defence Colony Delhi 💯Call Us 🔝8264348440🔝soniya singh
 
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...soniya singh
 
定制英国白金汉大学毕业证(UCB毕业证书) 成绩单原版一比一
定制英国白金汉大学毕业证(UCB毕业证书)																			成绩单原版一比一定制英国白金汉大学毕业证(UCB毕业证书)																			成绩单原版一比一
定制英国白金汉大学毕业证(UCB毕业证书) 成绩单原版一比一ffjhghh
 
{Pooja: 9892124323 } Call Girl in Mumbai | Jas Kaur Rate 4500 Free Hotel Del...
{Pooja:  9892124323 } Call Girl in Mumbai | Jas Kaur Rate 4500 Free Hotel Del...{Pooja:  9892124323 } Call Girl in Mumbai | Jas Kaur Rate 4500 Free Hotel Del...
{Pooja: 9892124323 } Call Girl in Mumbai | Jas Kaur Rate 4500 Free Hotel Del...Pooja Nehwal
 
Call Girls In Mahipalpur O9654467111 Escorts Service
Call Girls In Mahipalpur O9654467111  Escorts ServiceCall Girls In Mahipalpur O9654467111  Escorts Service
Call Girls In Mahipalpur O9654467111 Escorts ServiceSapana Sha
 
Ukraine War presentation: KNOW THE BASICS
Ukraine War presentation: KNOW THE BASICSUkraine War presentation: KNOW THE BASICS
Ukraine War presentation: KNOW THE BASICSAishani27
 
Unveiling Insights: The Role of a Data Analyst
Unveiling Insights: The Role of a Data AnalystUnveiling Insights: The Role of a Data Analyst
Unveiling Insights: The Role of a Data AnalystSamantha Rae Coolbeth
 
Delhi Call Girls Punjabi Bagh 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Call
Delhi Call Girls Punjabi Bagh 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip CallDelhi Call Girls Punjabi Bagh 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Call
Delhi Call Girls Punjabi Bagh 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Callshivangimorya083
 
Full night 🥵 Call Girls Delhi New Friends Colony {9711199171} Sanya Reddy ✌️o...
Full night 🥵 Call Girls Delhi New Friends Colony {9711199171} Sanya Reddy ✌️o...Full night 🥵 Call Girls Delhi New Friends Colony {9711199171} Sanya Reddy ✌️o...
Full night 🥵 Call Girls Delhi New Friends Colony {9711199171} Sanya Reddy ✌️o...shivangimorya083
 
VIP High Class Call Girls Bikaner Anushka 8250192130 Independent Escort Servi...
VIP High Class Call Girls Bikaner Anushka 8250192130 Independent Escort Servi...VIP High Class Call Girls Bikaner Anushka 8250192130 Independent Escort Servi...
VIP High Class Call Girls Bikaner Anushka 8250192130 Independent Escort Servi...Suhani Kapoor
 
04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationships04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationshipsccctableauusergroup
 
PKS-TGC-1084-630 - Stage 1 Proposal.pptx
PKS-TGC-1084-630 - Stage 1 Proposal.pptxPKS-TGC-1084-630 - Stage 1 Proposal.pptx
PKS-TGC-1084-630 - Stage 1 Proposal.pptxPramod Kumar Srivastava
 
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdfMarket Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdfRachmat Ramadhan H
 

Recently uploaded (20)

Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
 
Low Rate Call Girls Bhilai Anika 8250192130 Independent Escort Service Bhilai
Low Rate Call Girls Bhilai Anika 8250192130 Independent Escort Service BhilaiLow Rate Call Girls Bhilai Anika 8250192130 Independent Escort Service Bhilai
Low Rate Call Girls Bhilai Anika 8250192130 Independent Escort Service Bhilai
 
From idea to production in a day – Leveraging Azure ML and Streamlit to build...
From idea to production in a day – Leveraging Azure ML and Streamlit to build...From idea to production in a day – Leveraging Azure ML and Streamlit to build...
From idea to production in a day – Leveraging Azure ML and Streamlit to build...
 
Delhi 99530 vip 56974 Genuine Escort Service Call Girls in Kishangarh
Delhi 99530 vip 56974 Genuine Escort Service Call Girls in  KishangarhDelhi 99530 vip 56974 Genuine Escort Service Call Girls in  Kishangarh
Delhi 99530 vip 56974 Genuine Escort Service Call Girls in Kishangarh
 
Aminabad Call Girl Agent 9548273370 , Call Girls Service Lucknow
Aminabad Call Girl Agent 9548273370 , Call Girls Service LucknowAminabad Call Girl Agent 9548273370 , Call Girls Service Lucknow
Aminabad Call Girl Agent 9548273370 , Call Girls Service Lucknow
 
꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...
꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...
꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...
 
Call Girls in Defence Colony Delhi 💯Call Us 🔝8264348440🔝
Call Girls in Defence Colony Delhi 💯Call Us 🔝8264348440🔝Call Girls in Defence Colony Delhi 💯Call Us 🔝8264348440🔝
Call Girls in Defence Colony Delhi 💯Call Us 🔝8264348440🔝
 
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...
High Class Call Girls Noida Sector 39 Aarushi 🔝8264348440🔝 Independent Escort...
 
定制英国白金汉大学毕业证(UCB毕业证书) 成绩单原版一比一
定制英国白金汉大学毕业证(UCB毕业证书)																			成绩单原版一比一定制英国白金汉大学毕业证(UCB毕业证书)																			成绩单原版一比一
定制英国白金汉大学毕业证(UCB毕业证书) 成绩单原版一比一
 
{Pooja: 9892124323 } Call Girl in Mumbai | Jas Kaur Rate 4500 Free Hotel Del...
{Pooja:  9892124323 } Call Girl in Mumbai | Jas Kaur Rate 4500 Free Hotel Del...{Pooja:  9892124323 } Call Girl in Mumbai | Jas Kaur Rate 4500 Free Hotel Del...
{Pooja: 9892124323 } Call Girl in Mumbai | Jas Kaur Rate 4500 Free Hotel Del...
 
Call Girls In Mahipalpur O9654467111 Escorts Service
Call Girls In Mahipalpur O9654467111  Escorts ServiceCall Girls In Mahipalpur O9654467111  Escorts Service
Call Girls In Mahipalpur O9654467111 Escorts Service
 
VIP Call Girls Service Charbagh { Lucknow Call Girls Service 9548273370 } Boo...
VIP Call Girls Service Charbagh { Lucknow Call Girls Service 9548273370 } Boo...VIP Call Girls Service Charbagh { Lucknow Call Girls Service 9548273370 } Boo...
VIP Call Girls Service Charbagh { Lucknow Call Girls Service 9548273370 } Boo...
 
Ukraine War presentation: KNOW THE BASICS
Ukraine War presentation: KNOW THE BASICSUkraine War presentation: KNOW THE BASICS
Ukraine War presentation: KNOW THE BASICS
 
Unveiling Insights: The Role of a Data Analyst
Unveiling Insights: The Role of a Data AnalystUnveiling Insights: The Role of a Data Analyst
Unveiling Insights: The Role of a Data Analyst
 
Delhi Call Girls Punjabi Bagh 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Call
Delhi Call Girls Punjabi Bagh 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip CallDelhi Call Girls Punjabi Bagh 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Call
Delhi Call Girls Punjabi Bagh 9711199171 ☎✔👌✔ Whatsapp Hard And Sexy Vip Call
 
Full night 🥵 Call Girls Delhi New Friends Colony {9711199171} Sanya Reddy ✌️o...
Full night 🥵 Call Girls Delhi New Friends Colony {9711199171} Sanya Reddy ✌️o...Full night 🥵 Call Girls Delhi New Friends Colony {9711199171} Sanya Reddy ✌️o...
Full night 🥵 Call Girls Delhi New Friends Colony {9711199171} Sanya Reddy ✌️o...
 
VIP High Class Call Girls Bikaner Anushka 8250192130 Independent Escort Servi...
VIP High Class Call Girls Bikaner Anushka 8250192130 Independent Escort Servi...VIP High Class Call Girls Bikaner Anushka 8250192130 Independent Escort Servi...
VIP High Class Call Girls Bikaner Anushka 8250192130 Independent Escort Servi...
 
04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationships04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationships
 
PKS-TGC-1084-630 - Stage 1 Proposal.pptx
PKS-TGC-1084-630 - Stage 1 Proposal.pptxPKS-TGC-1084-630 - Stage 1 Proposal.pptx
PKS-TGC-1084-630 - Stage 1 Proposal.pptx
 
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdfMarket Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
 

What is the point of point estimates

  • 1. What is the point of point estimates? Stephen Senn Consultant Statistician Edinburgh (C) Stephen Senn 2022 1 “Superficially, point estimation may seem a simpler problem to discuss than that of interval estimation” Cox and Hinckley, p250
  • 2. Apologies • This was originally submitted to the Communicating Statistical Methods session • But it was not accepted  • Due to a cancellation from the Cluster Trials session I am able to present it here • I have departed from the abstract somewhat to introduce a cluster trial example • By doing this I shall annoy two sets of people • Those who wanted to hear what was in the abstract • Those who want to hear about up-to-date research on cluster trials (C) Stephen Senn 2022 2
  • 3. Being befuddled about balance • A matter that often befuddles amateur (non-statistical) commentaries on randomised studies is the role of balance • If factors are balanced, adjusting for them (in the linear case) won’t change the estimate so what’s the point? • However, the statistical point of view is more or less the opposite. • If you have balanced for a factor you ought to have the factor in the model • We don’t analyse a matched pairs design like a completely randomised one • Why? (C) Stephen Senn 2022 3
  • 4. As we all know • The reason is to do with the standard error • Balancing by a factor eliminates its effect on the estimate • But it does not, of its own, eliminate its effect on the standard error • In fact it increases it slightly • To estimate how well you have done, you need to remove its effect on the standard error • This requires putting it in the model • We all know this but one still regularly encounters arguments that show we have forgotten • In my opinion this is because we obsess about point estimates (C) Stephen Senn 2022 4
  • 5. An incomplete block design in asthma Two formulations of formoterol given at different doses • Details need not concern us • Some patients received • ISF24 and MTA6 • Some patients received • ISF24 but not MTA6 • Some patients received • MTA6 but not ISF24 • I can calculate both within subject and between subject estimates of the effect ISF24-MTA6 • This is not the way I would analyse these data but I could (and have) (C) Stephen Senn 2022 5 Data from Senn et al, 1997
  • 6. More or Less More Less (C) Stephen Senn 2022 6
  • 7. The Lanarkshire Milk Experiment An incomplete block cluster allocated trial from 1930 • 67 Schools enrolled • In some schools children were allocated 1:1 either to act as controls or receive raw milk • In other schools children were allocated 1:1 either to act as controls or receive pasteurised milk • Just over 18,000 schoolchildren studied • Four months of treatment (C) Stephen Senn 2022 7
  • 8. Analysis What the authors did • Analysed data on weight and height • Pooled all controls • Not a good idea • But analysed each of 14 = 2 sex by 7 age groups separately • Calculated ‘probable errors’ • But it is not clear how • Concluded that milk was beneficial but that one could not decide whether raw or pasteurised was better (C) Stephen Senn 2022 8 Prominent commentators Fisher Student
  • 9. What I have done • Developed algebraic expressions for variances of contrasts for six different types of analysis • Four that might be reasonable • Two that the authors carried out • For each of the six analyses there are two variances • The true variance • What one might naively estimate • Simulated to check my formulae • Everything looks fine • This means • Either it is fine • Or I made the same mistakes in my simulation as in the theory (C) Stephen Senn 2022 9
  • 10. What I have not done • I have not got hold of the original data • In fact we don’t even know how many schools of each type there were • Therefore I have had to speculate (postulate, guess, fake) what the variance components might be • The key issue as all those interested in cluster allocate trials know is the ratio of the between to the within cluster variance • I don’t like ICCs (C) Stephen Senn 2022 10
  • 11. Milking some data Parameter settings are identical for all six cases Heights in inches are considered The formulae are for a given sex by age group and n=10 per group per school 33 schools per milk type in a & b the school effect is fixed in e & f the school effect is random in c & d the school effect is ignored
  • 12. Sacred cow The TARGET study Target study • One of the largest studies ever run in osteoarthritis • 18,000 patients • Randomisation took place in two sub- studies of equal size • Lumiracoxib versus ibuprofen • Lumiracoxib versus naproxen • Practical considerations dictated design • Purpose to investigate CV and GI tolerability of lumiracoxib • Sub-study effect explicitly dealt with in analysis Lanarkshire Milk Study • At the time one of the largest nutritional studies • 18,000 school children • Randomisation took place in two sub- studies of equal size • No milk versus raw milk • No milk versus pasteurised milk • Practical considerations dictated design • Purpose to investigate effect of milk on height and weight • Sub-study effect ignored in analysis (C) Stephen Senn 2022 12 (c) Stephen Senn 2012
  • 13. Baseline Demographics Sub-Study 1 Sub Study 2 Demographic Characteristic Lumiracoxib n = 4376 Ibuprofen n = 4397 Lumiracoxib n = 4741 Naproxen n = 4730 Use of low-dose aspirin 975 (22.3) 966 (22.0) 1195 (25.1) 1193 (25.2) History of vascular disease 393 (9.0) 340 (7.7) 588 (12.4) 559 (11.8) Cerebro- vascular disease 69 (1.6) 65 (1.5) 108 (2.3) 107 (2.3) Dyslipidaemias 1030 (23.5) 1025 (23.3) 799 (16.9) 809 (17.1) Nitrate use 105 (2.4) 79 (1.8) 181 (3.8) 165 (3.5) (C) Stephen Senn 2022 13 If this looks familiar it ought to. Bergen 2012
  • 14. Baseline Deviances Model Term Demographic Characteristic Sub-study (DF=1) Treatment given Sub- study (DF=2) Treatment (DF=2) Use of low-dose aspirin 23.57 0.13 13.40 History of vascular disease 70.14 5.23 47.41 Cerebro- vascular disease 13.54 0.14 7.75 Dyslipidaemias 117.98 0.17 54.72 Nitrate use 39.83 4.62 29.17 (C) Stephen Senn 2022 14
  • 15. Some final words • The very impressive and interesting causal inference school seems to be promoting identifiability rather than estimability • The former considers whether what happens asymptotically is correct • It is not always obvious what has to go to infinity for an asymptote to be reached • Schools/centres or pupils/patients • I think we are in danger of loosing a valuable insight from experimental design theory • How treatments are varied across the block structure matters • The lessons of the Rothamsted School should be heeded (C) Stephen Senn 2022 15 stephen@senns.uk @stephensenn http://www.senns.uk/Blogs.html “..the calculation of standard errors is idle and misleading, if the method of arrangement adopted fails to guarantee their validity…” RA Fisher, The Design of Experiments section 34 Fisher Yates Nelder