Testing for Exponentiality
| dc.contributor.advisor | Tiku, M. L. | |
| dc.contributor.author | Rai, Kamta | |
| dc.contributor.department | Statistics | en_US |
| dc.date.accessioned | 2016-08-10T14:10:55Z | |
| dc.date.available | 2016-08-10T14:10:55Z | |
| dc.date.issued | 1973-08 | |
| dc.description.abstract | <p> Several test statistics, which are known, can be used for testing for exponentiality. A new test statistic TE is proposed. TE is based on a censored sample and is similar to Tiku's T statistic for testing for normality. The distribution of TE tends to normality with increasing sample size. Besides, TE is easy to compute and is both origin and scale invariant. The power of TE for non-exponential distributions is comparable with Shapiro & Wilk statistic W-exponential. </p> | en_US |
| dc.description.degree | Master of Science (MSc) | en_US |
| dc.description.degreetype | Thesis | en_US |
| dc.identifier.uri | http://hdl.handle.net/11375/20091 | |
| dc.language.iso | en | en_US |
| dc.subject | exponentiality | en_US |
| dc.subject | normality | en_US |
| dc.subject | scale invariant | en_US |
| dc.subject | non-exponential distributions | en_US |
| dc.title | Testing for Exponentiality | en_US |