| Título: | The bootstrap : a simple approach to estimating standard errors and confidence intervals when theory fails |
| Autores: | THOMPSON, MICHAEL |
| Tipo de documento: | texto impreso |
| Editorial: | Cambridge [REINO UNIDO] : Royal Society of Chemistry, 2001 |
| Colección: | AMC Technical Brief, num. 8 |
| ISBN/ISSN/DL: | 29714 |
| Dimensiones: | 2 p. |
| Nota general: | ISBN: |
| Langues: | Inglés |
| Clasificación: | |
| Resumen: | Standard errors and confidence intervals for a simple statistic like the mean can be calculated by the use of an algebraic formula derived from familiar assumptions about the data, such as the normal distribution. For a more complex type of statistic (like the familiar relative standard deviation), or where the standard assumptions do not apply, we often find that an algebraic formula cannot be derived. In such instances, a simple alternative method based on re-sampling the data is becoming increasingly popular. This computer-intensive method, known as the bootstrap1, is widely applicable and is introduced here by two straightforward examples. |
Documentos electrónicos (1)
 27964 Adobe Acrobat PDF |
