Diff for "FAQ/paranp" - CBU statistics Wiki
location: Diff for "FAQ/paranp"
Differences between revisions 7 and 8
Revision 7 as of 2013-08-28 13:10:08
Size: 2691
Editor: PeterWatson
Comment:
Revision 8 as of 2013-08-28 13:11:00
Size: 2699
Editor: PeterWatson
Comment:
Deletions are marked like this. Additions are marked like this.
Line 23: Line 23:
Hodges, J. L. and Lehmann, E. L. (1956). "The efficiency of some nonparametric competitors of the t-test". ''Annals of Mathematical Statistics'' '''27(2)''' 324–335. Hodges, J. L. and Lehmann, E. L. (1956). "The efficiency of some nonparametric competitors of the t-test". ''Annals of Mathematical Statistics'', '''27(2)''', 324–335.
Line 25: Line 25:
Johnson, D. H. (1995). Statistical Sirens: The Allure of Nonparametrics. ''Ecology'' '''76''' 1998–2000. Johnson, D. H. (1995). Statistical Sirens: The Allure of Nonparametrics. ''Ecology'', '''76''', 1998–2000.
Line 27: Line 27:
Tanizaki, H. (1997). "Power Comparison of Nonparametric Tests: Small Sample Properties from Monte-Carlo Experiments," ''Journal of Applied Statistics'' '''24(5)''' 603-632. Tanizaki, H. (1997). "Power Comparison of Nonparametric Tests: Small Sample Properties from Monte-Carlo Experiments," ''Journal of Applied Statistics'', '''24(5)''', 603-632.
Line 29: Line 29:
Zimmerman, D. W. (2003). A warning about the large sample Wilcoxon-Mann-Whitney test. ''Understanding Statistics'' '''2(4)''' 267-280. Zimmerman, D. W. (2003). A warning about the large sample Wilcoxon-Mann-Whitney test. ''Understanding Statistics'', '''2(4)''', 267-280.

Some advantages to using parametric tests

The two main advantages to using parametric tests are greater precision and power. Field (2012), however, notes that there are some situations where nonparametric tests have greater power than parametric ones. The advent of permutations and randomisation tests have also enabled confidence intervals to be computed for test statistics based upon distribution-free tests.

(The below is taken from Alex Yu's reply to the edstat discussion list on here which discussed advantages of parametric tests).

Edgington (1995, p.85-86) ‘it is unwise to degrade the precision by transforming the measurements into ranks for conducting a statistical test’. In addition, generally speaking, the statistical power of non-parametric tests are lower than that of their parametric counterpart except on a few occasions (Hodges & Lehmann, 1956; Tanizaki, 1997).

Further, non-parametric tests are criticized for being incapable of answering the focused question. For example, the WMW procedure tests whether the two distributions are different in some way but does not show how they differ in mean, variance, or shape. Based on this limitation, Johnson (1995) preferred robust procedures and data transformation to non-parametric tests.

Fagerland and Sandvik (2009) suggests using the Welch-Satterthwaite correction to the t-test instead of the Mann-Whitney test if the two groups being compared differ on standard deviations and skewnesses. Fagerland and Sandvik (2009) and Zimmerman (2003) also found in addition to Mann-Whitney giving inflated Type I errors when groups have unequal standard deviations it can also give misleading results when comparing large groups.

References:

Edgington, E. S. (1995). Randomization tests. Third Edition. Marcel Dekker Inc:New York.

Fagerland, M. W. and Sandvik, L. (2009). The Wilcoxon-Mann-Whitney test under scrutiny. Statistics in Medicine, 28, 1487-1497. A pdf copy is here.

Hodges, J. L. and Lehmann, E. L. (1956). "The efficiency of some nonparametric competitors of the t-test". Annals of Mathematical Statistics, 27(2), 324–335.

Johnson, D. H. (1995). Statistical Sirens: The Allure of Nonparametrics. Ecology, 76, 1998–2000.

Tanizaki, H. (1997). "Power Comparison of Nonparametric Tests: Small Sample Properties from Monte-Carlo Experiments," Journal of Applied Statistics, 24(5), 603-632.

Zimmerman, D. W. (2003). A warning about the large sample Wilcoxon-Mann-Whitney test. Understanding Statistics, 2(4), 267-280.

None: FAQ/paranp (last edited 2013-08-28 13:12:11 by PeterWatson)