Select your statistics for data analysis

PSYBER NEWS observes that some manuscripts are good in framing hypotheses, collection of data etc but poor in data analysis. In this connection, authors are requested to choose suitable statistics according to own data quality (scaling properties, distribution of data), sample distribution (random or purposive) and specific hypotheses.

Some good sources are:

Non-parametic test:

http://psybernews.blogspot.com/2010/12/non-parametic-test-of-significance.html

Stepwise or hierarchical multiple regression analysis:

http://www.drivehq.com/web/bepperiva/Articoli/Wampold%20(1987)%20Multiple%20Regression%20in%20Consulting%20Psychology.pdf