|
David de Vaus' Web Page: SOCIAL-RESEARCH.ORG
|
|
Links for Chapter 14: Bivariate Analysis for Nominal and Ordinal variables
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Web pointer 14.3 Bivariate inferential analysis |
||
Description |
URL in book |
Last Updated |
|
A good brief discussion of significance levels |
http://www.statsoftinc.com/textbook/ |
|
|
Clear discussion of the logic of hypothesis testing and significance levels. |
||
|
Good discussion of relationship between sample size, strength of correlations and statistical significance. |
http://www.statsoftinc.com/textbook/ |
|
|
An explanation of why only large samples will produce weak relationships that are statistically significant. |
||
|
A very good discussion of confidence intervals, sample size, significance levels and confidence intervals. |
http://www.sportsci.org/ |
|
|
Calculators for confidence intervals of some ordinal measures of association (but only for very limited range of table sizes) |
Download the program SISA tables from the following web site: http://home.clara.net/sisa/ This program can calculate statistics for 2 x 2, 2 x 7 and 3 x 3 tables. It calculates confidence intervals for Kendall’s tau, gamma and lambda. Alternatively the site
http://home.clara.net/ enables the same online analysis but is restricted to 2 x 2 and 2 x 5 tables
|
|
|
A calculator for confidence interval for a Pearson’s or Spearman’s correlation |
||
|
A spreadsheet calculator for (among other things) calculating confidence limits for Pearson’s r and other normally distributed statistics |
||
|
A series of quotes from statisticians who object to the slavish (and often inappropriate) use of significance tests – especially when used on their own. |
||
New Links |
||
|
A
set of articles and debates regarding
significance testing |
http://roberts.ed.psu.edu/users/droberts/sigtest.htm | |
| Do you know of more good links on this topic? Email me and I will add them | ![]() |
|
Site last updated 2 Decmeber 2003 |