Namespaces
Variants
Actions

Difference between revisions of "Bradley-Terry model"

From Encyclopedia of Mathematics
Jump to: navigation, search
(Importing text file)
 
m (tex encoded by computer)
 
(One intermediate revision by the same user not shown)
Line 1: Line 1:
 +
<!--
 +
b1108201.png
 +
$#A+1 = 32 n = 0
 +
$#C+1 = 32 : ~/encyclopedia/old_files/data/B110/B.1100820 Bradley\ANDTerry model
 +
Automatically converted into TeX, above some diagnostics.
 +
Please remove this comment and the {{TEX|auto}} line below,
 +
if TeX found to be correct.
 +
-->
 +
 +
{{TEX|auto}}
 +
{{TEX|done}}
 +
 
The names of R.A. Bradley and M.E. Terry are associated with a model that is widely employed in paired comparisons. In paired comparison experiments, observations are made by presenting pairs of objects to one or more judges. This method is used in experimental situations where objects are judged subjectively (cf. also [[Paired comparison model|Paired comparison model]]). The basic model was presented in [[#References|[a1]]]. The model is mostly formulated as follows (see, e.g., [[#References|[a2]]]).
 
The names of R.A. Bradley and M.E. Terry are associated with a model that is widely employed in paired comparisons. In paired comparison experiments, observations are made by presenting pairs of objects to one or more judges. This method is used in experimental situations where objects are judged subjectively (cf. also [[Paired comparison model|Paired comparison model]]). The basic model was presented in [[#References|[a1]]]. The model is mostly formulated as follows (see, e.g., [[#References|[a2]]]).
  
The paired comparison experiment has <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b1108201.png" /> objects <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b1108202.png" /> with <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b1108203.png" /> comparisons of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b1108204.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b1108205.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b1108206.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b1108207.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b1108208.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b1108209.png" />. Some of the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082010.png" /> may be zero, but the design should be connected, i.e., there must not be any subset of the treatments such that no treatment in this subset is compared with any treatment of the complementary subset. The model postulates the existence of treatment parameters <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082011.png" /> for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082012.png" /> (<img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082013.png" />), <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082014.png" />, such that the probability of selecting <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082015.png" />, when compared with <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082016.png" />, is equal to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082017.png" />.
+
The paired comparison experiment has $  t $
 +
objects $  T _ {1} \dots T _ {t} $
 +
with $  n _ {ij }  $
 +
comparisons of $  T _ {i} $
 +
and $  T _ {j} $,
 +
$  n _ {ij }  \geq  0 $,  
 +
$  n _ {ii }  = 0 $,  
 +
$  n _ {ji }  = n _ {ij }  $,  
 +
$  i,j = 1 \dots t $.  
 +
Some of the $  n _ {ij }  $
 +
may be zero, but the design should be connected, i.e., there must not be any subset of the treatments such that no treatment in this subset is compared with any treatment of the complementary subset. The model postulates the existence of treatment parameters $  \pi _ {i} $
 +
for $  T _ {i} $(
 +
$  \pi _ {i} \geq  0 $),  
 +
$  \sum _ {i} \pi _ {i} = 1 $,  
 +
such that the probability of selecting $  T _ {i} $,  
 +
when compared with $  T _ {j} $,  
 +
is equal to $  { {\pi _ {i} } / {( \pi _ {i} + \pi _ {j} ) } } $.
  
The likelihood method can be used to estimate the parameters <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082018.png" /> (cf. [[Likelihood equation|Likelihood equation]]). If the comparisons are independent, the likelihood function <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082019.png" /> is
+
The likelihood method can be used to estimate the parameters $  \pi _ {i} $(
 +
cf. [[Likelihood equation|Likelihood equation]]). If the comparisons are independent, the likelihood function $  L $
 +
is
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082020.png" /></td> </tr></table>
+
$$
 +
L = {
 +
\frac{\prod _ { i } \pi _ {i} ^ {a _ {i} } }{\prod _ {i \leq  j } ( \pi _ {i} + \pi _ {j} ) ^ {n _ {ij }  } }
 +
} ,
 +
$$
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082021.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082022.png" /> is the number of times <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082023.png" /> has been preferred to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082024.png" /> when <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082025.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082026.png" /> were compared (<img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082027.png" />). In [[#References|[a3]]], large-sample results and the asymptotic distribution of the maximum-likelihood estimators are given. L.R. Ford has described an iterative solution of the likelihood equations (see [[#References|[a4]]]). R.J. Beaver has presented, [[#References|[a5]]], a weighted least-squares approach to paired comparison models. He has also used a method, described in [[#References|[a6]]], which presents a unified approach to the analysis of data resulting from an experiment involving multinomial populations. Thus, general results from weighted least-squares methods can be used. A. Springall (see [[#References|[a7]]]) has considered the case where the parameters <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082028.png" /> (<img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082029.png" />) are functions of continuous variables <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082030.png" />. He has formulated a model that is linear in the unknown parameters <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082031.png" />. In this case the parameters <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110820/b11082032.png" /> have to be estimated. He has given results concerning the asymptotic variance-covariance matrix of the estimators.
+
where $  a _ {i} = \sum _ {j} n _ {i.ij }  $
 +
and $  n _ {i.ij }  $
 +
is the number of times $  T _ {i} $
 +
has been preferred to $  T _ {j} $
 +
when $  T _ {i} $
 +
and $  T _ {j} $
 +
were compared ( $  n _ {i.ij }  + n _ {j.ij }  = n _ {ij }  $).  
 +
In [[#References|[a3]]], large-sample results and the asymptotic distribution of the maximum-likelihood estimators are given. L.R. Ford has described an iterative solution of the likelihood equations (see [[#References|[a4]]]). R.J. Beaver has presented, [[#References|[a5]]], a weighted least-squares approach to paired comparison models. He has also used a method, described in [[#References|[a6]]], which presents a unified approach to the analysis of data resulting from an experiment involving multinomial populations. Thus, general results from weighted least-squares methods can be used. A. Springall (see [[#References|[a7]]]) has considered the case where the parameters $  \pi _ {i} $(
 +
$  i = 1 \dots t $)  
 +
are functions of continuous variables $  x _ {1} \dots x _ {s} $.  
 +
He has formulated a model that is linear in the unknown parameters $  { \mathop{\rm log} } \pi _ {i} = \sum _ {k} x _ {ik }  \beta _ {k} $.  
 +
In this case the parameters $  \beta _ {k} $
 +
have to be estimated. He has given results concerning the asymptotic variance-covariance matrix of the estimators.
  
 
A linear model like the one above can be used to answer optimal design questions.
 
A linear model like the one above can be used to answer optimal design questions.

Latest revision as of 06:29, 30 May 2020


The names of R.A. Bradley and M.E. Terry are associated with a model that is widely employed in paired comparisons. In paired comparison experiments, observations are made by presenting pairs of objects to one or more judges. This method is used in experimental situations where objects are judged subjectively (cf. also Paired comparison model). The basic model was presented in [a1]. The model is mostly formulated as follows (see, e.g., [a2]).

The paired comparison experiment has $ t $ objects $ T _ {1} \dots T _ {t} $ with $ n _ {ij } $ comparisons of $ T _ {i} $ and $ T _ {j} $, $ n _ {ij } \geq 0 $, $ n _ {ii } = 0 $, $ n _ {ji } = n _ {ij } $, $ i,j = 1 \dots t $. Some of the $ n _ {ij } $ may be zero, but the design should be connected, i.e., there must not be any subset of the treatments such that no treatment in this subset is compared with any treatment of the complementary subset. The model postulates the existence of treatment parameters $ \pi _ {i} $ for $ T _ {i} $( $ \pi _ {i} \geq 0 $), $ \sum _ {i} \pi _ {i} = 1 $, such that the probability of selecting $ T _ {i} $, when compared with $ T _ {j} $, is equal to $ { {\pi _ {i} } / {( \pi _ {i} + \pi _ {j} ) } } $.

The likelihood method can be used to estimate the parameters $ \pi _ {i} $( cf. Likelihood equation). If the comparisons are independent, the likelihood function $ L $ is

$$ L = { \frac{\prod _ { i } \pi _ {i} ^ {a _ {i} } }{\prod _ {i \leq j } ( \pi _ {i} + \pi _ {j} ) ^ {n _ {ij } } } } , $$

where $ a _ {i} = \sum _ {j} n _ {i.ij } $ and $ n _ {i.ij } $ is the number of times $ T _ {i} $ has been preferred to $ T _ {j} $ when $ T _ {i} $ and $ T _ {j} $ were compared ( $ n _ {i.ij } + n _ {j.ij } = n _ {ij } $). In [a3], large-sample results and the asymptotic distribution of the maximum-likelihood estimators are given. L.R. Ford has described an iterative solution of the likelihood equations (see [a4]). R.J. Beaver has presented, [a5], a weighted least-squares approach to paired comparison models. He has also used a method, described in [a6], which presents a unified approach to the analysis of data resulting from an experiment involving multinomial populations. Thus, general results from weighted least-squares methods can be used. A. Springall (see [a7]) has considered the case where the parameters $ \pi _ {i} $( $ i = 1 \dots t $) are functions of continuous variables $ x _ {1} \dots x _ {s} $. He has formulated a model that is linear in the unknown parameters $ { \mathop{\rm log} } \pi _ {i} = \sum _ {k} x _ {ik } \beta _ {k} $. In this case the parameters $ \beta _ {k} $ have to be estimated. He has given results concerning the asymptotic variance-covariance matrix of the estimators.

A linear model like the one above can be used to answer optimal design questions.

See also Paired comparison model.

References

[a1] R.A. Bradley, M.E. Terry, "The rank analysis of incomplete block designs. I. The method of paired comparisons" Biometrika , 39 (1952) pp. 324–345
[a2] R.A. Bradley, A.T. El-Helbawy, "Treatment contrasts in paired comparisons: basic procedures with application to factorials" Biometrika , 63 (1976) pp. 255–262
[a3] R.A. Bradley, "Rank analysis of incomplete block designs. III. Some large-sample results on estimation and power for a method of paired comparisons" Biometrika , 42 (1955) pp. 450–470
[a4] L.R. Ford, Jr., "Solution of a ranking problem from binary comparisons" Amer. Math. Monthly , 64 (1957) pp. 28–33
[a5] R.J. Beaver, "Weighted least squares analysis of several univariate Bradley–Terry models" J. Amer. Statist. Assoc. , 72 (1977) pp. 629–634
[a6] J.E. Grizzle, C.F. Starmer, G.G. Koch, "Analysis of categorical data by linear models" Biometrics , 25 (1969) pp. 489–504
[a7] A. Springall, "Response surface fitting using a generalization of the Bradley–Terry paired comparison model" Appl. Statist. , 22 (1973) pp. 59–68
How to Cite This Entry:
Bradley-Terry model. Encyclopedia of Mathematics. URL: http://encyclopediaofmath.org/index.php?title=Bradley-Terry_model&oldid=17604
This article was adapted from an original article by E.E.M. van Berkum (originator), which appeared in Encyclopedia of Mathematics - ISBN 1402006098. See original article