NUR 80A/B Lecture Notes - Lecture 5: Scatter Plot, Descriptive Statistics, Interval Ratio
Document Summary
Bivariate descriptive statistics: describe relationships b/w 2 variables. Two-dimensional frequency distribution in which the frequencies of 2 variables are cross-tabulated. Usually used with nominal/ ordinal data that have few levels/ ranks. Relationships b/w 2 variables usually described through correlation procedures. Can be answered quantitatively by calculating correlation coefficient, which describes intensity & direction of a relationship. Weight & height have a positive relationship b/c greater height tends to be associated with greater weight. When 2 variables are unrelated, cc is 0. Cc b/w . 00 and -1. 00 express a negative/ inverse relationship. The higher the absolute value (value disregarding the sign) of the coefficient, the stronger the relationship; eg. correlation of -. 80 much stronger than +. 20. Most commonly used correlation index is product-moment correlation coefficient (pearson"s r), which is computed with interval/ ratio measures. One correlation index for ordinal measures in spearman"s rank-order correlation ( as spearman"s rho.
