The purposes of this study were: (1) to illustrate the applications of sequential analysis in education and the behavioral sciences, (2) to compare the power of the sequential configural rank (SCR) test, the modified sequential configural rank (MSCR) test, and the sequential t-test for the two sample situation, and (3) to compare the average sample size of the SCR-test, the MSCR-test and the sequential t-test for the two-sample situation. / Applications of five sequential tests were illustrated with data from education and the behavioral sciences and with the aid of computer software written in Applesoft BASIC. The efficiency of the sequential tests with respect to sample size was also demonstrated. / In order to compare the power and average sample size of the SCR-test, the MSCR-test, and the sequential t-test, the present study used computer generated Monte Carlo methods as its primary means of investigation. The power and the average sample size of the three tests were compared under three distributions: normal ((mu) = 0, (sigma) = 1), uniform (interval 0, 1), and exponential ((lamda) = 1). / Results of the Monte Carlo study suggested that the sequential t-test has appreciable power advantage (about 1 to 23 percent) over the SCR-test and the MSCR-test under normal and uniform distributions. The magnitude of the power advantage depends on the effect size tested (k(,1)) and the true effect size (k). The SCR-test is more powerful than the MSCR-test under normal and uniform distributions. However, under exponential distribution the MSCR-test shows modest power advantage over the SCR-test. Under exponential distribution the SCR-test and the MSCR-test have power superiority over the sequential t-test in the range 1 to 11 percent and 1 to 21 percent respectively. / The MSCR-test requires substantially smaller number of observations (5 to 53 percent smaller) than the SCR-test and the sequential t-test under the three distributions. The sequential t-test requires fewer observations than the SCR-test under the three distributions. / Source: Dissertation Abstracts International, Volume: 47-05, Section: A, page: 1535. / Thesis (Ph.D.)--The Florida State University, 1986.
| Identifer | oai:union.ndltd.org:fsu.edu/oai:fsu.digital.flvc.org:fsu_75810 |
| Contributors | PAULSON, DAVE., Florida State University |
| Source Sets | Florida State University |
| Detected Language | English |
| Type | Text |
| Format | 174 p. |
| Rights | On campus use only. |
| Relation | Dissertation Abstracts International |
Page generated in 0.0016 seconds