Objective: To study the tendency of effects with no descriptions in Prescriptions of Peaceful Benevolent Dispensary. Method: Based on the attributes of medicines in formulae

the attribution distance matrix was developed and applied to compute the weights of attributes in a tuple vector. The methods of average value

steepest descent

Lagrange mean value

and information quantity were applied to induce the efficiency of formula based on component medicines. Result: To quantify the results

the precision measure was used to evaluate the effectiveness of effect attribution reduction among these four methods by using different formulae in Prescriptions of Peaceful Benevolent Dispensary. By experiments

the precision of efficiency attribution reduction among these four methods is 61.5%

79.2%

68.2%

70.8%

respectively. Conclusion: The results of average value method may be complex

the results of steepest descent method may be simple

the Lagrange mean value method may be a trade-off solution

and the results of information quantity method will be changeable based on thresholds. These methods can cross-refer and complement each other.