2013- 5- 16
|
#816
|
|
متميزفي كلية ألأداب_قسم الأنجلش
|
رد: هنـــــا المراجعه الاخيرة طرق البحث والتصميم
جرح هذا الرد
b4) Measures of relationship. These quantify the amount of relationship between two (or more) variables as measured in the same group of people or whatever. They are usually on a scale 0-1 (in some instances they run from -1 through 0 to +1). I.e. if such a measure comes out near 1 (or -1 where relevant), that indicates that those cases that scored a particular value on one variable also tended to score a particular value on the other. E.g. those who scored high on motivation also scored high on proficiency. If it comes out near 0, that indicates that cases that scored a particular way on one variable scored all over the other variable, and vice versa. Examples are the Pearson 'r' Correlation Coefficient, the Spearman 'rho' Correlation Coefficient, Kendall's W, the 'phi' Correlation Coefficient, Kruskal's 'gamma'. (Remember that relationship and difference are really the same thing looked at from different points of view. If there is a difference between men and women - the two values of the 36 | P a g e Heart story
gender variable - in attitude to RP accent, then there is a relationship between the variables gender and attitude to RP accent. It is just that for technical reasons sometimes statistics approaches the matter more via measuring difference, sometimes via measuring relationship).
اقروه كله اذا تبون 
|
|
|
|
|
|