It is very common chromatographic practice to optimize resolution by making changes in selectivity by systematically varying key retention controlling factors. In many instances, a change in conditions merely results in monotonic, systematic variation in the relative retention of all pairs of peaks. Useful or 'effective' changes in selectivity generally result when we see peak crossovers, changes in elution order or differential changes in band position of three or more peaks upon changing some operating condition. In this work, we demonstrate that changes in what we now call the effective selectivity can only take place when retention depends on a minimum of two solute molecular properties and further the dependencies must differ for the two sets of conditions. To verify our concept, real chromatographic data are examined from the viewpoint of linear solvation energy relationships (LSERs) and linear solvent strength theory. Five different RPLC stationary phases in different eluents are compared to elucidate the similarities and differences in their effective selectivities. Of major importance is our finding that the effective selectivity can only be understood when it is viewed in terms of the ratios of system-dependent interaction coefficients, such as the LSER coefficients, and not merely the absolute values of the coefficients. We confirm, both theoretically and experimentally, that a change in mobile- phase volume fraction and in column temperature is not as powerful a mechanism for tuning the effective selectivity as is a change in stationary- phase type.