Search Results

Three principles of observational design

Marian Stamp Dawkins

in Observing Animal Behaviour: design and analysis of quantitative data

The same principles that apply to the design of experiments also apply to the design of an observational study, but instead of manipulating the animals, the observer controls the way he or she takes ... More

The same principles that apply to the design of experiments also apply to the design of an observational study, but instead of manipulating the animals, the observer controls the way he or she takes observations. The three principles are: independent replication; not confounding variables; and removing known sources of variation by blocking or matching. The chapter shows how these three principles can be applied to the design of observations, and discusses issues that often get in the way of valid designs such as pseudoreplication, experimenter bias, and choice of which animals to observe. Some simple observational designs are introduced.Less

In Praise of Randomisation: The Importance of Causality in Medicine and its Subversion by Philosophers of Science

DAVID COLQUHOUN

in Evidence, Inference and Enquiry

The job of scientists is to try to distinguish what is true from what is false by means of observation and experiment. That job has been made difficult by some philosophers of science who appear to ... More

The job of scientists is to try to distinguish what is true from what is false by means of observation and experiment. That job has been made difficult by some philosophers of science who appear to give academic respectability to relativist, and even postmodernist, postures. This chapter suggests that the contributions of philosophers to causal understanding have been unhelpful. It puts the case for randomised studies as the safest guarantee of the reliability of scientific evidence. It uses the case of hormone replacement therapy to illustrate the importance of randomisation, and the case of processed meat and cancer to illustrate the problems that arise in the absence of randomised tests. Finally, it discusses the opposition to randomisation that has come from some philosophers of science.Less

 Mixing Things Up

Ken Binmore

in Playing for Real: Game Theory

This chapter develops the idea of a mixed strategy using the entry into a sealed-bid auction as a non-trivial example. Reaction curves are first illustrated for the case of pure strategies and then ... More

This chapter develops the idea of a mixed strategy using the entry into a sealed-bid auction as a non-trivial example. Reaction curves are first illustrated for the case of pure strategies and then applied to computing mixed Nash equilibria. The Hawk-Dove Game is equivalent either to the Prisoner's Dilemma or Chicken, depending on parameter values. The mixed-strategy reaction curves are plotted in each case. The interpretation of mixed Nash equilibria as polymorphic equilibria in a game played by a large population is considered. The matrix algebra necessary for handling mixed strategies is reviewed and illustrated with O'Neill's Card Game. Convexity ideas are reviewed and applied to the geometric representation of mixed strategies. Cooperative and noncooperative payoff regions are introduced and illustrated using Chicken and the Battle of the Sexes. Correlated equilibria are introduced after a discussion of self-policing agreements, cheap talk, and preplay randomization. The possibility of correlation without a referee using techniques from cryptography is discussed.Less

Mendelian randomization: the contribution of genetic epidemiology to elucidating environmentally modifiable causes of disease

George Davey Smith and Shah Ebrahim

in Human Genome Epidemiology, 2nd Edition: Building the evidence for using genetic information to improve health and prevent disease

Conventional risk factor epidemiology — directly studying environmentally modifiable exposures that may influence disease risk — and genetic epidemiology have similarities and differences. The ... More

Conventional risk factor epidemiology — directly studying environmentally modifiable exposures that may influence disease risk — and genetic epidemiology have similarities and differences. The case-control design is, for example, more popular in genetic epidemiology than it currently is in conventional risk factor epidemiology, and while the importance of sample size is recognized in conventional epidemiology, the huge collaborative ventures currently being undertaken in genetic epidemiology have not been the norm, since special attention has, appropriately, been paid to detailed exposure and outcome measurement. In genetic epidemiology there has recently been much attention paid to false-positive findings generated by multiple hypothesis testing against a background of inadequate statistical power, whereas in risk factor epidemiology, problems generated by confounding and bias have been to the forefront. This chapter deals with Mendelian randomization, a principle that underlies some of the differences between conventional risk factor and genetic epidemiology, and also renders genetic epidemiology a useful tool for improving the identification of environmentally modifiable risk factors that are causally related to disease outcomes, and therefore targets for therapeutic or preventative intervention.Less

Statistical Alternatives and Supplements to Random Sampling

Patrick Dattalo

in Strategies to Approximate Random Sampling and Assignment

This chapter describes the following alternatives and complements to RS in terms of their assumptions, implementations, strengths, and weaknesses: (1) randomization tests; (2) multiple imputation; ... More

This chapter describes the following alternatives and complements to RS in terms of their assumptions, implementations, strengths, and weaknesses: (1) randomization tests; (2) multiple imputation; and (3) mean-score logistic regression. Randomization tests are statistical alternatives to RS. Multiple imputation is a statistical supplement RS. Mean-score logistic regression is a statistical alternative or supplement to RS.Less

Analysis of Population-Based Survey Experiments

Diana C. Mutz

in Population-Based Survey Experiments

This chapter addresses three common problems that emerge in the analysis stage. These include the misguided practice of doing randomization checks, whether and how to use survey weights, and the use ... More

This chapter addresses three common problems that emerge in the analysis stage. These include the misguided practice of doing randomization checks, whether and how to use survey weights, and the use and misuse of covariates. Investigators tend to approach the analysis and interpretation of population-based experiments from the perspective of usual practices in whatever their home discipline happens to be. However, as it turns out, some of these choices go hand in hand with common errors and faulty assumptions about the most appropriate way to analyze results from population-based experiments. Usual practices are not always appropriate, particularly with disciplines that favor observational methods. The chapter aims to provide guidance on the best practices and to explain when and why they make a difference.Less

Randomization in Decision-Making

ROBERT V. DODGE

in Schelling's Game Theory: How to Make Decisions

This chapter considers randomization, beginning with its history as “fairness” in making decisions. It then examines “mixed strategies” and von Neumann's Minimax Theorem. A segment from Poe's The ... More

This chapter considers randomization, beginning with its history as “fairness” in making decisions. It then examines “mixed strategies” and von Neumann's Minimax Theorem. A segment from Poe's The Purloined Letter makes the point of randomizing one's choice when against an opponent who will outsmart you if you make a decision. A presentation of Schelling's class materials concerns randomizing to outsmart a burglar. The value of randomizing decisions is explained. These illustrations make the concept easy to comprehend without the complex mathematics often involved. The final section introduces the Nash Equilibrium, easily recognized in a 2 × 2 matrix, with examples. The Nash equilibrium often applies to bargaining and a matrix illustrates that Nash equilibria can exist and lead firms to form duopolies. There can be more than one Nash equilibrium. One solution, posited by Schelling, is that certain solutions have prominence and are selected among Nash equilibria. Schelling called them focal points.Less

Conclusion

Neil Duxbury

in Random Justice: On Lotteries and Legal Decision-Making

This book has illustrated the advantages and drawbacks of chance and randomized social decision-making, claiming that arguments concerning such decision-making more often than not require ... More

This book has illustrated the advantages and drawbacks of chance and randomized social decision-making, claiming that arguments concerning such decision-making more often than not require qualification. Detailed scrutiny of randomized legal decision-making compels us to confront difficult, sometimes uncomfortable, questions concerning the role of reason in law and how we conceptualize justice. Although we are often understandably wary of resorting to lotteries to determine outcomes of legal significance, the idea that randomization might be employed more extensively within legal decision-making contexts ought not to be dismissed cursorily. Rigid application of some particular decision-making criterion to settle disputes might render adjudication less fraught with complexity and ambiguity. Depending on the criterion used, such application might even make the process of adjudication less partial. If the criterion is easy to apply, moreover, the costs of decision-making are likely to be reduced. One criterion which offers all of these qualities is random selection.Less

Research designs for evaluation studies

Carolyn Emery

in Sports Injury Research

A multitude of study designs are used in order to establish the efficacy of a preventive measure, e.g. randomised controlled trials, time-trend analyses, prospective trials, etc. Each method has its ... More

A multitude of study designs are used in order to establish the efficacy of a preventive measure, e.g. randomised controlled trials, time-trend analyses, prospective trials, etc. Each method has its benefits and drawbacks. This chapter provides a clear overview of the available designs and shows, using existing literature, the pros and cons of various designs.Less

The Virtues of Randomization

in Against Prediction: Profiling, Policing, and Punishing in an Actuarial Age

The critiques set forth in this book reflect problems with the actuarial approach more generally—not just with specific types of stereotyping or profiles. This chapter sketches the contours and ... More

The critiques set forth in this book reflect problems with the actuarial approach more generally—not just with specific types of stereotyping or profiles. This chapter sketches the contours and benefits of a more randomized universe of crime and punishment. Randomization is the only way to achieve a carceral population that reflects the offending population. Randomization in this context is a form of random sampling: random sampling on the highway, for instance, is the only way that the police would obtain an accurate reflection of the offending population. And random sampling is the central virtue behind randomization. What randomization achieves, in essence, is to neutralize the perverse effects of prediction, both in terms of the possible effects on overall crime and of the other social costs.Less

Confounding

J. Mark Elwood

in Critical Appraisal of Epidemiological Studies and Clinical Trials

Confounding is the most challenging issue in the interpretation of studies. This chapter is divided into three parts. The first part defines confounding and shows what effects it can produce. The ... More

Confounding is the most challenging issue in the interpretation of studies. This chapter is divided into three parts. The first part defines confounding and shows what effects it can produce. The second part deals with how confounding can be controlled. The third part considers some further applications of the logic of confounding. Self-test questions are provided at the end of the chapter.Less

Choice of experimental design

Sheila Bird

in Effective Sexual Health Interventions: Issues in Experimental Evaluation

This chapter considers the strengths and limitations of a range of evaluation study designs, focusing on the consumer principle of randomization. Although much of the chapter is devoted to randomized ... More

This chapter considers the strengths and limitations of a range of evaluation study designs, focusing on the consumer principle of randomization. Although much of the chapter is devoted to randomized controlled trials (RCTs), non-randomized studies such as disease registries and cohort studies are included because of their importance in the evaluation of cost-effectiveness. The chapter also considers issues of informed consent that are relevant to choice of experimental design, and the need for database linkage to overcome informative loss to follow-up that might otherwise undermine randomized allocation.Less

Theoretical underpinning for the use of intergenerational studies in life course epidemiology

Debbie A Lawlor, Sam Leary, and George Davey Smith

in Family matters: Designing, analysing and understanding family based studies in life course epidemiology

Intergenerational studies have been widely used in life course epidemiology both to examine primary research hypotheses and to explore underlying mechanisms for established associations. This chapter ... More

Intergenerational studies have been widely used in life course epidemiology both to examine primary research hypotheses and to explore underlying mechanisms for established associations. This chapter describes the theoretical underpinning for using different types of intergenerational studies in life course epidemiology and discusses how results from such studies should be interpreted. Specifically, it considers the use and interpretation of cousin, sibling, and twin intergenerational studies; egg donation/surrogate mother intergenerational studies; maternal-paternal comparisons; and intergenerational migrant studies and Mendelian randomization in intergenerational studies in life course epidemiology.Less