In multiple regression a hypotheses is used to determine if what you want to predict is possible or not possible whereas in simple regression the value to be predicted is often one of the two variables and you use the second one to forecast the dependant variable.
How multiple regression can be used with the collected data.
Multiple regressions can be used on the collected data to forecast the future value of a given dependant variable for example given the Last Trade as a dependant variable we can choose the Nyse Only Close, Volume, and DJIA as an independent variable.
Y will represent the values of The Last Trade from 10th FEB to 12th and X1 can represent the values of NYSE Close only from 10th FEB to 12th FEB and X2 can represent the Volumes . The coefficients for variables can be expressed as b1 andb2. As always in any equation we must have a constant variable. And in this case it is a. so our equation can be as follows;
Y = a+ b1X1 + b2X2
Using the equation above we can use multiple regression to forecast the future value of the Last Trade. The last trade being the dependant variable represented by Y.
The applicability of multiple regressions can be deficient because in the market the last trade value and the NYSE only close are determined in the market itself and not by the DJIA from another market. Hence it can not be used in multiple regressions.
In multiple regressions we use real integers and not percentages and negative value thus the values for change in this data cannot be used in multiple regressions (Lane, D. n.d.)
In general multiple regression is applicable with part of the collected data which is important to predict the required value and the one that is not needed is not used. It can be done with part of the data without changing it like the Last Trade and the NYSE only Close and the Volume but cannot be applied with the DJIA and the change values.