Thermal de Broglie wavelength: Difference between revisions
en>Khazar2 m →General definition of the thermal wavelength: clean up using AWB |
en>Jason from nyc fill in citation |
||
Line 1: | Line 1: | ||
{{More footnotes|date=November 2010}} | |||
'''Correction for attenuation''' is a statistical procedure, due to [[Charles Spearman|Spearman]] (1904), to "rid a [[correlation]] coefficient from the weakening effect of measurement error" (Jensen, 1998), a phenomenon also known as [[regression dilution]]. In [[measurement]] and [[statistics]], it is also called '''disattenuation'''. The [[correlation]] between two sets of parameters or measurements is estimated in a manner that accounts for measurement error contained within the [[estimator|estimates]] of those parameters. | |||
==Background== | |||
Correlations between parameters are diluted or weakened by measurement error. Disattenuation provides for a more accurate estimate of the correlation between the parameters by accounting for this effect. | |||
===Derivation of the formula=== | |||
Let <math>\beta</math> and <math>\theta</math> be the true values of two attributes of some person or [[statistical unit]]. These values are regarded as [[random variables]] by virtue of the statistical unit being selected randomly from some [[statistical population|population]]. Let <math>\hat{\beta}</math> and <math>\hat{\theta}</math> be estimates of <math>\beta</math> and <math>\theta</math> derived either directly by observation-with-error or from application of a measurement model, such as the [[Rasch model]]. Also, let | |||
::<math> | |||
\hat{\beta} = \beta + \epsilon_{\beta} , \quad\quad \hat{\theta} = \theta + \epsilon_\theta, | |||
</math> | |||
where <math>\epsilon_{\beta}</math> and <math>\epsilon_\theta</math> are the measurement errors associated with the estimates <math>\hat{\beta}</math> and <math>\hat{\theta}</math>. | |||
The correlation between two sets of estimates is | |||
:<math> | |||
\operatorname{corr}(\hat{\beta},\hat{\theta})= \frac{\operatorname{cov}(\hat{\beta},\hat{\theta})}{\sqrt{\operatorname{var}[\hat{\beta}]\operatorname{var}[\hat{\theta}}]} | |||
</math> | |||
:::::<math> | |||
=\frac{\operatorname{cov}(\beta+\epsilon_{\beta}, \theta+\epsilon_\theta)}{\sqrt{\operatorname{var}[\beta+\epsilon_{\beta}]\operatorname{var}[\theta+\epsilon_\theta]}}, | |||
</math> | |||
which, assuming the errors are uncorrelated with each other and with the estimates, gives | |||
:<math> | |||
\operatorname{corr}(\hat{\beta},\hat{\theta})= \frac{\operatorname{cov}(\beta,\theta)}{\sqrt{(\operatorname{var}[\beta]+\operatorname{var}[\epsilon_\beta])(\operatorname{var}[\theta]+\operatorname{var}[\epsilon_\theta])}} | |||
</math> | |||
:::::<math> | |||
=\frac{\operatorname{cov}(\beta,\theta)}{\sqrt{(\operatorname{var}[\beta]\operatorname{var}[\theta])}}.\frac{\sqrt{\operatorname{var}[\beta]\operatorname{var}[\theta]}}{\sqrt{(\operatorname{var}[\beta]+\operatorname{var}[\epsilon_\beta])(\operatorname{var}[\theta]+\operatorname{var}[\epsilon_\theta])}} | |||
</math> | |||
:::::<math> | |||
=\rho \sqrt{R_\beta R_\theta}, | |||
</math> | |||
where <math>R_\beta</math> is the ''separation index'' of the set of estimates of <math>\beta</math>, which is analogous to [[Cronbach's alpha]]; this is, in terms of [[Classical test theory]], <math>R_\beta</math> is analogous to a reliability coefficient. Specifically, the separation index is given as follows: | |||
:<math> | |||
R_\beta=\frac{\operatorname{var}[\beta]}{\operatorname{var}[\beta]+\operatorname{var}[\epsilon_\beta]}=\frac{\operatorname{var}[\hat{\beta}]-\operatorname{var}[\epsilon_\beta]}{\operatorname{var}[\hat{\beta}]}, | |||
</math> | |||
where the mean squared standard error of person estimate gives an estimate of the variance of the errors, <math>\epsilon_\beta</math>. The standard errors are normally produced as a by-product of the estimation process (see [[Rasch model estimation]]). | |||
The disattenuated estimate of the correlation between two sets of parameters or measures is therefore | |||
:<math> | |||
\rho = \frac{\mbox{corr}(\hat{\beta},\hat{\theta})}{\sqrt{R_\beta R_\theta}}. | |||
</math> | |||
That is, the disattenuated correlation is obtained by dividing the correlation between the estimates by the square root of the product of the separation indices of the two sets of estimates. Expressed in terms of Classical test theory, the correlation is divided by the square root of the product of the reliability coefficients of two tests. | |||
Given two [[random variable]]s <math>X</math> and <math>Y</math>, with [[correlation]] <math>r_{xy}</math>, and a known [[Reliability (statistics)#Classical test theory|reliability]] for each variable, <math>r_{xx}</math> and <math>r_{yy}</math>, the correlation between <math>X</math> and <math>Y</math> corrected for attenuation is | |||
<math>r_{x'y'} = \frac{r_{xy}}{\sqrt{r_{xx}r_{yy}}}</math>. | |||
How well the variables are measured affects the correlation of ''X'' and ''Y''. The correction for attenuation tells you what the correlation would be if you could measure ''X'' and ''Y'' with perfect reliability. | |||
If <math>X</math> and <math>Y</math> are taken to be imperfect measurements of underlying variables <math>X'</math> and <math>Y'</math> with independent errors, then <math>r_{x'y'}</math> measures the true correlation between <math>X'</math> and <math>Y'</math>. | |||
==See also== | |||
* [[Regression dilution]] | |||
* [[Errors-in-variables model]] | |||
== References == | |||
*Jensen, A.R. (1998). ''The ''g'' Factor: The Science of Mental Ability'' Praeger, Connecticut, USA. ISBN 0-275-96103-6 | |||
*Spearman, C. (1904) "The Proof and Measurement of Association between Two Things". ''The American Journal of Psychology'', 15 (1), 72–101 {{JSTOR|1412159}} | |||
==External links== | |||
*[http://www.rasch.org/rmt/rmt101g.htm Disattenuating correlations] | |||
*[http://pareonline.net/getvn.asp?v=8&n=11 Disattenuation of correlation and regression coefficients: Jason W. Osborne] | |||
{{DEFAULTSORT:Correction For Attenuation}} | |||
[[Category:Measurement]] | |||
[[Category:Covariance and correlation]] | |||
[[Category:Psychometrics]] |
Revision as of 01:49, 10 December 2013
Correction for attenuation is a statistical procedure, due to Spearman (1904), to "rid a correlation coefficient from the weakening effect of measurement error" (Jensen, 1998), a phenomenon also known as regression dilution. In measurement and statistics, it is also called disattenuation. The correlation between two sets of parameters or measurements is estimated in a manner that accounts for measurement error contained within the estimates of those parameters.
Background
Correlations between parameters are diluted or weakened by measurement error. Disattenuation provides for a more accurate estimate of the correlation between the parameters by accounting for this effect.
Derivation of the formula
Let and be the true values of two attributes of some person or statistical unit. These values are regarded as random variables by virtue of the statistical unit being selected randomly from some population. Let and be estimates of and derived either directly by observation-with-error or from application of a measurement model, such as the Rasch model. Also, let
where and are the measurement errors associated with the estimates and .
The correlation between two sets of estimates is
which, assuming the errors are uncorrelated with each other and with the estimates, gives
where is the separation index of the set of estimates of , which is analogous to Cronbach's alpha; this is, in terms of Classical test theory, is analogous to a reliability coefficient. Specifically, the separation index is given as follows:
where the mean squared standard error of person estimate gives an estimate of the variance of the errors, . The standard errors are normally produced as a by-product of the estimation process (see Rasch model estimation).
The disattenuated estimate of the correlation between two sets of parameters or measures is therefore
That is, the disattenuated correlation is obtained by dividing the correlation between the estimates by the square root of the product of the separation indices of the two sets of estimates. Expressed in terms of Classical test theory, the correlation is divided by the square root of the product of the reliability coefficients of two tests.
Given two random variables and , with correlation , and a known reliability for each variable, and , the correlation between and corrected for attenuation is .
How well the variables are measured affects the correlation of X and Y. The correction for attenuation tells you what the correlation would be if you could measure X and Y with perfect reliability.
If and are taken to be imperfect measurements of underlying variables and with independent errors, then measures the true correlation between and .
See also
References
- Jensen, A.R. (1998). The g Factor: The Science of Mental Ability Praeger, Connecticut, USA. ISBN 0-275-96103-6
- Spearman, C. (1904) "The Proof and Measurement of Association between Two Things". The American Journal of Psychology, 15 (1), 72–101 Glazier Alfonzo from Chicoutimi, has lots of interests which include lawn darts, property developers house for sale in singapore singapore and cigar smoking. During the last year has made a journey to Cultural Landscape and Archaeological Remains of the Bamiyan Valley.