ABSTRACT
This chapter takes a first look at the variation of production and productivity growth across different producers and different areas. Descriptive statistics are employed to scan the producer characteristics and regional variables that may be responsible for the variation in the growth. Correlation analysis and Chi Square statistics are used respectively to evaluate linear and nonlinear relationships between input and output growth and producer attributes. The correlation coefficient analysis may also identify the categories of companies who experienced higher growth or, conversely, lower or negative growth. To examine whether company attributes make a difference in technological inputs, frequency distribution of the technological inputs across different categories and Chi Square statistics for all distribution tables are computed. A univariate analysis is performed to determine the breaking points so the companies can spread in a reasonable number of categories. Since textile-apparel and computer-electronics companies have different distributions by size and input level, different breaking points are used in conversion.
