Correlation and Regression as applied in Green Concrete Research

Correlation and regression were used as statistical tools to analyse the mechanical and durability properties of green concrete. The use of these analytical tools showed that the degree of relationship that exists between a green concrete property (for example compressive strength or water absorption or resistance to sulphate attack) and percentage replacement of cement with a civil engineering material can be established using correlation; the nature and extent of the relationship can be known using regression; and the green concrete property analysed using regression can be predicted with the regression equation. The results of correlation analysis showed that Pearson product-moment is a better method for analysing green concrete properties when compared with Spearman’s correlation coefficient, since Pearson’s method takes cognizance of the actual data values unlike Spearman’s use of data positions. The results of regression analysis in section 3.1 and 3.2 have also shown that, for data whose trendline is polynomial in shape with the highest power of its assumed equation more than two (that is higher than quadratic equation), it is better to analyse such data using least square regression line method for linear relationship than least square regression line method for non-linear relationship
The use of Microsoft Excel aids the simplification of the calculation in regards to the nonlinear regression analysis as it reduces the computational time and effort needed to carry out the analysis. So it is recommended for use in carrying out nonlinear regression analysis in relation to green concrete research.