Multiple Regression Analysis Hypothesis test

64 Sugiyono, 2012: 230 The conclusion is to compare t value with t table . If t count greater than or equal to the level t table significant 5 then these variables significantly affect the dependent variable. Conversely, if t value smaller than t table the effect of the variable is not significant. 4. Make a simply linear regresi line Formula : Y = aX + K Descriptions : Y = Accounting Learning Achievement a = Coefficient number X = Locus of ControlStudent Perception of Teacher Teaching Methods Parents Concern K = Constant number Sutrisno Hadi, 2004: 1

b. Multiple Regression Analysis

This analysis is used to examine the independent variables together with the dependent variable. This analysis is used to test the fourth hypothesis that the effect of Locus of Control, Student Perceptions of Teachers Teaching Methods and Parents Concern hypothesis 4. These Steps in the multiple regression analysis are: 1 Finding coefficient double correlation between variables X1, X2 and X3 and dependent variable Y, using the formula: R y1,2 = √ 65 Descriptions : R y1,2 = Correlation Coefficient between Accounting Learning Achievement and Locus of Control, Student Perceptions of Teacher Teaching Methods, Parents Concern a 1 = Coefficient of Locus Control a 2 = Coefficient of Student Perceptions of Teacher Teaching Methods a 3 = Coefficient of Parents Concern x 1 y = Total of Product between X1 and Y x 2 y = Total of Product between X2 and Y x 3 y = Total of Product between X3 and Y y 2 = Total of Accounting Learning Achievement Square Sutrisno Hadi, 2004: 22 Directions correlation will be positive if the result of the correlation calculation sign =. If the minus sign -, then the direction of the correlation is negative Suharsimi Arikunto, 2010: 213 2 Finding determination coefficient R 2 between independent X1, X2 and X3 and dependent variables Y Formula : R 2 = R 2 Descriptions : R 2 = determination coefficient R = multiple correlation coeffiecient Darwyan Shah, et al., 2009: 94 Therefore the effect of independent variables X1, X2 and X3 together with the dependent variable Y by the square of the multiple correlation coefficient. Furthermore, the coefficient of determination multiplied by 100 to determine 66 the level of effect of two independent variables on the dependent variable in terms of percentage. 3 Examine multiple regresi significant with F test, the formula is Descriptions : Freg = F price of regression N = Number of cases m = Number of predictor R = Correlations of coefficient between predictor criteria Sutrisno Hadi, 2004: 26 The conclusion taken by comparing F value with significant F table at the level of 5. If F value is equal to or greater than F table the independent variables on the dependent ones significantly. Conversely, when F value is smaller than F table , then the effect of the independent variables on the dependent variable is not significant. 4 Finding Relative Contributions sumbangan relatif SR dan Effective Contribution sumbangan efektif SE each predictor on kriterium with formula: a Relative Contribution, Sumbangan Relatif SR Predictor X 1 = Predictor X 2 = Predictor X 3 = 67 Descriptions : SR = relative contribution of predictor a1 = predictor coefficient X1 a2 = predictor coefficient X2 a3 = predictor coefficient X3 ∑x 1 y = total product between X1 and Y ∑x 2 y = total product between X2 and Y ∑x 3 y = total product between X3 and Y Jkreg = total of regression square Sutrisno Hadi, 2004: 42 b Effective Contributions Sumbangan Efektif SE Prediktor X 1 = SE= SRX 1 x R 2 Prediktor X 2 = SE= SRX 2 x R 2 Prediktor X 3 = SE= SRX 3 x R 2 Descriptions : SEX1 = effective contribution of X1 predictor SEX2 = effective contribution of X2 predictor SEX3 = effective contribution of X3 predictor SRX1 = relative contribution of X1 predictor SRX2 = relative contribution of X2 predictor SRX3 = relative contribution of X3 predictor R 2 = determinant coefficient Sutrisno Hadi, 2004: 46

4. Locus of Control Category Determination