Appraisal of linear plots in chemical engineering and metallurgy
Jing Li · Journal of Xi'an University of Architecture & Technology · 2001
This paper analyses the advantages and disadvantages of several methods of established linear relations used widely in research of chemical engineering and meatllurgy.It concludes that the t -distribution method is most scientific for defining linear correlation parameters in that it operates in an integrated way with the linear correlating coefficient r ,correlating coefidence level a , N ,the number of experimental points,and r c,the critical linear correlating coefficient.The paper further addresses how to use this method to examine the declared linear relationships in literatures.The results have shown that 15 groups are not statistically significant at a 95 % confident level of the 81 ones,taking 19 % of the total events.Majority of the 15 groups are composed of the data groups their numbers are less than 5.