Relationship between the independent variables and dependent variable Essay

Hypothesis statement can be well-defined as a statement of the relationship between two or more variables expressed in the form of a testable statement, which carry the clear implication for testing the stated relations. The hypotheses are divided by two elements which are null hypothesis (H0) and alternate hypothesis (H1). The null hypothesis is statements that express no relationship significant between variables while the alternate hypothesis is statements that express relationship is significant between variables.
Hypothesis 1
H0 There is no relationship between Capital Ratio (X1) and Return on Asset.
H1 There is a relationship between Capital Ratio (X1) and Return on Asset.
Hypothesis 2
H0 There is no relationship between Liquidity Ratio (X2) and Return on Asset.
H1 There is a relationship between Liquidity Ratio (X2) and Return on Asset.
Hypothesis 3
H0 There is no relationship between Gross Domestic Products (X3) and Return on Asset.
H1 There is a relationship between Gross Domestic Products (X3) and Return on Asset.
Hypothesis 4
H0 There is no relationship between Inflation Rate (X4) and Return on Asset.
H1 There is a relationship between Inflation Rate (X4) and Return on Asset.
3.4 STATISTICS/ECONOMETRIC METHODS
This study improves the observed method to detect any relationship between the independent variables and dependent variable. In this study, Multiple Linear Regression (MLR) models will be applied to time series data to test the data. This research paper involves the steps taken for analyzing time series data by using Multiple Linear Regression are
i. Check on the assumptions.
ii. Perform correlation matrix of all variables.
iii. Perform regression.
3.4.1 Test on Assumption
i. Normality Test
To test the normality, the Jarque-Bera is used to identify whether the error is normally distributed or not distributed. The hypothesis as follows H0 Error term normally distributed H1 Error term is not normally distributed If the p - value is more than 5% level of significance the null hypothesis is accepted which is meant the data for all variables (ROA, X1, X2, X3 and X4) is normally distributed which fulfilled the assumption of a good regression.
In testing for serial or autocorrelation problem, Breach-Godfrey Serial correlation test is used. This estimation will investigate whether there is serial interdependence for the error term of the variables (ROA, X1, X2, X3, and X4). Besides that, it is also to use to determine the correlation between two different time series data. In addition, autocorrelation implies that the error term from one time period depends on upon other time periods. The condition of this test is when residual are related to each other and it can be confirmed from the p-value of Chi-square of Obs*R-Squared. Using 5% level of significance, if the p-value is more than 5%, the null hypothesis is accepted, but if it is less, the alternate hypothesis are accepted the hypotheses are H0 Error term is serially independent (no autocorrelation problem) H1 Error term is not serially independent (have autocorrelation problem) .
iii. Heteroscedasticity Test Variance of Error Term Test