Una aproximación al número de repeticiones en modelos de diseño experimental de una y dos vías de clasificación con factores de efecto fijo y una observación por celda
Abstract: This work refers to the approximation proposal for the estimation of the number of repetitions in experimental design models of one and two classification ways, with fixed effect factors and one observation per cell that allows the researcher to determine their value, prior to the experime...
Gorde:
| Argitaratua: |
2017
|
|---|---|
| Gaiak: | |
| Sarrera elektronikoa: | http://hdl.handle.net/10872/17432 |
| Etiketak: |
Etiketarik gabe, Izan zaitez lehena erregistro honi etiketa jartzen!
|
| Gaia: | Abstract: This work refers to the approximation proposal for the estimation of the number of repetitions in experimental design models of one and two classification ways, with fixed effect factors and one observation per cell that allows the researcher to determine their value, prior to the experiment being performed considering in addition the type II error rate the power and magnitude of the differences that could be detected. Based on the procedure established by PC Tang (1938), it is proposed to determine the value of the non-centrality parameter of the non-central F distribution using information that does not depend on the estimates of 2 and Σ2. Which allows to calculate the approximation to the number of repetition to be made in the experiment and the estimation of the power of test associated to this one. In order to establish comparisons to observe the behavior of the approach proposal, experimental data were taken from illustrative examples of existing literature in experimental designs and field trials, as well as simulated tests, the power of the test was determined for these using The methodology established by PC Tang (1938), to then apply the proposed method to the same data. The results showed that the methodology of the proposed approach produces an underestimation in the value of the parameter of non-centrality as well as in the power, there were cases in which the determined values were very similar. |
|---|