SJS SDI_51 Design of Statistical Investigations Stephen Senn 5. Orthogonal Designs Randomised Blocks.

Slides:



Advertisements
Similar presentations
Analysis by design Statistics is involved in the analysis of data generated from an experiment. It is essential to spend time and effort in advance to.
Advertisements

SJS SDI_71 Design of Statistical Investigations Stephen Senn 7. Orthogonal Designs Two (plus) Blocking Factors.
SJS SDI_61 Design of Statistical Investigations Stephen Senn 6. Orthogonal Designs Randomised Blocks II.
SJS SDI_81 Design of Statistical Investigations Stephen Senn 8 Factorial Designs.
SJS SDI_101 Design of Statistical Investigations Stephen Senn 10 Random Effects.
SJS SDI_161 Design of Statistical Investigations Stephen Senn Random Sampling I.
SJS SDI_11 Design of Statistical Investigations Stephen Senn 1 General Introduction.
SJS SDI_31 Design of Statistical Investigations Stephen Senn 3. Design of Experiments 1 Some Basic Ideas.
SJS SDI_111 Design of Statistical Investigations Stephen Senn 11 Nested Factors.
SJS SDI_141 Design of Statistical Investigations Stephen Senn 14 Case Control Studies.
SJS SDI_151 Design of Statistical Investigations Stephen Senn Introduction to Sampling.
SJS SDI_21 Design of Statistical Investigations Stephen Senn 2 Background Stats.
SJS SDI_171 Design of Statistical Investigations Stephen Senn Random Sampling 2.
1 By: Chris Hartl A statistics learning project: 0.
Experimental Statistics - week 5
Multiple Comparisons in Factorial Experiments
DESIGNING EXPERIMENTS
Block designs and matched pairs designs
Experiments. Types of experiments ‘so far’ Paired comparison Happy experiment watching Goon video Two independent groups Different treatments for each.
Split-Plot Experiment Top Shrinkage by Wool Fiber Treatment and Number of Drying Revolutions J. Lindberg (1953). “Relationship Between Various Surface.
Sample Selection Issues in Experiment Random sampling (difficult) Convenience & purposive sampling Volunteers External validity Representativeness & generalizability.
Study Design Data. Types of studies Design of study determines whether: –an inference to the population can be made –causality can be inferred random.
7/6/2015 Orthogonal Functions Chapter /6/2015 Orthogonal Functions Chapter 7 2.
Introduction to Experimental and Observational Study Design KNNL – Chapter 16.
Introduction to the design (and analysis) of experiments James M. Curran Department of Statistics, University of Auckland
Discussion on Randomisation and Blinding
+ Controlled User studies HCI /6610 Winter 2013.
Section 3. Double Blind - Neither the subjects nor the people working with them know which treatment each subject is receiving Block Design - A block.
Types of clinical studies Thomas Abraham. Three broad types of studies: used for different purposes 1. Observational (observe groups of people, gather.
Much of the meaning of terms depends on context. 1.
The success or failure of an investigation usually depends on the design of the experiment. Prepared by Odyssa NRM Molo.
1 Experimental Design. 2  Single Factor - One treatment with several levels.  Multiple Factors - More than one treatment with several levels each. 
The Research Design. Experimental Design Definition A description of what a researcher would like to find out and how to find it out. Pre-requisites 1.Identification.
Designing Experiments Purpose for experiments – to study the response of one variable to the changes in other variables. Experimental Units (Subjects)
Producing Data 1.
Copyright ©2013, 2010, 2007, 2004 by W. H. Freeman and Company The Basic Practice of Statistics, 6 th Edition David S. Moore, William I. Notz, Michael.
Section 1.6 The Design of Experiments. Definitions Experiment = a controlled study conducted to determine the effect that varying one or more explanatory.
Understanding Experiments Lecture 12 Section 3.5 Tue, Feb 6, 2007.
Simple Comparative Experiments Section 2.3. More on Experiments An experiment is a planned intervention undertaken to observe the effects of one or more.
Designing Experiments 5.2. Vocabulary Experimental Units: the individuals on which the experiment is done Subjects: when the experimental units are humans.
Chapter 13: Part III AP Statistics.
 Get out homework and discuss with neighbor.  Be prepared with any questions you might have.  Get out materials for notes.
Collecting Sample Data Chapter 1 Section 4 Part 2.
Analytical Interventional Studies
Comparing Multiple Factors:
EXPERIMENTAL RESEARCH
Probability and Statistics
Statistical Analysis of the Randomized Block Design
Design of Experiments.
Design of Experiments.
MATH 2311 Sections 6.2 & 6.3.
مدیریت استراتژيک منابع انسانی
An activity-based journal club to help staff & students improve confidence at reading scientific papers Cornwall Health Library: Katy Oak Catriona Organ.
Introduction to Experimental and Observational Study Design
Section 2.2: Sampling.
Review Questions III Compare and contrast the components of an individual score for a between-subject design (Completely Randomized Design) and a Randomized-Block.
Ռազմավարական կառավարում
Experimental Design Project
Chapter 11 Principles of Experimental Design.
MATH 2311 Sections 6.2.
Designing Experiments
Principles of Experimental Design
Introduction to the design (and analysis) of experiments
Experimental Design Statistics.
Similarities Differences
Design of Experiments.
Designing experiments - keeping it simple
Principles of Experimental Design
Probability and Statistics
University of Florida College of Medicine
Presentation transcript:

SJS SDI_51 Design of Statistical Investigations Stephen Senn 5. Orthogonal Designs Randomised Blocks

SJS SDI_52 Blocks So far we have ignored differences in experimental units Some subsets of units may be similar to each other but different from other subsets Such similar subsets are called blocks The presence of blocks can be exploited –By design –And by analysis

SJS SDI_53 Randomised Block Design We identify blocks of experimental material We allocate treatments to block at random in such a way that –each treatment appears in every block –if a treatment appears m times in one block it appears m times in all blocks –but subject to no further restriction Referred to as a randomised block design

SJS SDI_54 # to create randomized blocks n.b<-6 # number of blocks n.t<-3 # number of treatments n.r<-2 # number of replicates #create vector of treatments treat<-c(rep(seq(1,n.t),n.r)) #creat vector of blocks block<-seq(1,n.b) #create one permuted block unit<-sample(treat) #create other permuted blocks #and join them for(i in 1:(n.b-1)){ unit<-rbind(unit,sample(treat)) } design.frame<-data.frame(block,unit) design.frame #print design Note use of sample function

SJS SDI_55 Randomised Blocks SPlus Output > design.frame block unit.1 unit.2 unit.3 unit.4 unit.5 unit

SJS SDI_56 Exp_5 Graff-Lonnevig and Browaldh (1990), Senn and Auclair (1990) Cross-over trial of single doses of 12 g formoterol compared with 200 g salbutamol in 13 asthmatic children. Main outcome measure peak expiratory flow (PEF) 8 hours after treatment. Two sequences used with wash-out in between.

SJS SDI_57 Design Points Treatments are given in two periods Washout is used to allow possible carry-over to disappear Two sequences were used –Permits blinding –A voids associating particular treatment with particular period We shall assume patients were randomised to the two sequences

SJS SDI_58 Exp_5 The Data We shall ignore the sequence information for the moment. If we have assigned patients at random to the two possible sequences, this is a randomised blocks design

SJS SDI_59 Questions What do we note about the precision of measurement? –What possible explanation is there? What do we note about the patient numbers? –What possible explanation is there?

SJS SDI_510 Blocks in a Cross-over In this design the units are episodes of treatment As the graphs that follow will show, there is a correlation between results from the same patient Patients form the blocks of the experiment –Naturally –And by design

SJS SDI_511

SJS SDI_512

SJS SDI_513 Points and Questions The graph plots the salbutamol reading against the formoterol reading Each point represents a patient –triangles salbutamol/formoterol sequence –squares formoterol/salbutamol sequence All the points except one are to the right of the line of equality –What does this suggest?

SJS SDI_514 Blocking From field trials in agriculture A block was a set of plots of presumed similar fertility Design trick was to use each treatment within a given block –Compare like with like –Eliminate a source of variation Now used to describe any set of similar units use in design

SJS SDI_515 Blocks - Examples Centres in a multi-centre trial –Units are patients Cars in a fuel consumption experiment –Units are runs Patients in a cross-over trial –Units are episodes of treatment Fermentation tanks in a plant –Units are runs

SJS SDI_516 Model for Randomised Blocks Assume for simplicity every treatment appears once in each block Quantity to be minimised Normal equations obtain by differentiating sum of squares with respect to unknown parameters and setting equal to zero Basic model

SJS SDI_517 Some Notation Total of all observations Number of observations Total on treatment i Total in block j

SJS SDI_518 Solutions

SJS SDI_519 Identifiability (5.1) Is over-parameterised Not all effects are identifiable However contrasts of the form below are uniquely identifiable These are of particular interest These are not

SJS SDI_520 Indentifiability continued Predictions are also identifiable

SJS SDI_521 Exp_5: Means

SJS SDI_522 Exp_5: Means

SJS SDI_523 Predicted Value and Residual Patient 7, Formoterol Note that the data are laid out in columns for treatments and rows for blocks ( patients) for convenience but that our notation suggested rows for treatments columns for blocks. Our subscripts reflect this latter convention. Note also that since patient 8 is missing we have a potential ambiguity regarding subscripts for patients 9 onwards.

SJS SDI_524 Exp_5 Predicted Values

SJS SDI_525 Exp_5 Residuals

SJS SDI_526 Sums of Squares Expanding we get….

SJS SDI_527 Hence we get...

SJS SDI_528 ANOVA Identity

SJS SDI_529 ANOVA Table

SJS SDI_530 Computational Approaches Reminder In general

SJS SDI_531 Computational Approaches (cont) Thus to calculate the Total Sum of Squares we may proceed as follows

SJS SDI_532 Computational Approaches (cont) Sum of squares between blocks

SJS SDI_533 Computational Approaches (cont) Sum of squares between treatments

SJS SDI_534 Exp_5: Calculation 1

SJS SDI_535 Exp_5: Calculation 2

SJS SDI_536 Exp_5: Calculation 3

SJS SDI_537 Exp_5 Analysis using Excel This uses the data analysis menu of Excel

SJS SDI_538 Exp_5 ANOVA Analysis using SPlus (Data input details omitted) #ANOVA just fitting treat fit1<-aov(pef~treat) summary(fit1) #ANOVA fitting treat and patient fit2<-aov(pef~patient+treat) summary(fit2)

SJS SDI_539 Exp_5 SPlus Results > summary(fit1) Df Sum of Sq Mean Sq F Value Pr(F) treat Residuals > #ANOVA just fitting treat and patient fit2 <- aov(pef ~ patient + treat) > summary(fit2) Df Sum of Sq Mean Sq F Value Pr(F) patient treat Residuals

SJS SDI_540 Questions Has the treatment sum of squares changed in fitting patient? Are the degrees of freedom for treatment different? What has changed? Why has it changed? What is the net effect?