APPENDIX A: CORRELATION AND REGRESSION ANALYSIS 



1. Correlation and regression analysis deals with the investigation and 

 inference of dependencies between variables as derived solely from statistical 

 techniques. Even if a high degree of association between variables is found 

 by this method, the resultant relation does not necessarily have, and usually 

 will not have, a clear physical basis. However, these statistical methods can 

 provide an indication of possible relationships to be accepted or rejected 

 after physical considerations. Information in this appendix may be obtained 

 from any textbook on regression analysis, e.g., Ostle and Mensing (1975)*, and 

 is included here to facilitate understanding of the statistical techniques and 

 terminology extensively employed in the data analyses presented in the main 

 text of this report. 



Correlation Coefficient 



2. The correlation coefficient r expresses the degree of linearity 

 between variables. For a set of discrete data, r is defined as: 



(Al) 



where 



= ^ ^ (Xi - x)(yi - y) (A2) 



i=l 



i=l 



) (A3) 



References cited in the Appendix can be found in the Reference list 

 at the end of the main text. 



Al 



