The Influence of Investment Decisions and Funding Decisions on the Value of Companies with Ownership Structure as Moderated Variables in Manufacturing Companies Listed in Indonesia Stock Exchange (BEI)
The Influence of Investment Decisions and Funding Decisions on the Value of Companies with Ownership Structure as Moderated Variables in Manufacturing Companies Listed in Indonesia Stock Exchange (BEI)
The Influence of Investment Decisions and Funding Decisions on the Value of Companies with Ownership Structure as Moderated Variables in Manufacturing Companies Listed in Indonesia Stock Exchange (BEI)
ISSN No:-2456-2165
The data that has been collected in the study, is processed Autocorrelation Test Analysis
using statistical data processing applications, namely Eviews Autocorrelation can be defined as a correlation between
software use multiple linear regression with descriptive interference variables one with other variable disorders
research, normality test analysis, multicollinearity test (Ghozali, 2016). The autocorrelation test aims to test whether
analysis, heteroskedasticity test analysis, autocorrelation test in a linear regression model there is a correlation between
analysis, multiple regression analysis, coefficient of disruptor errors in certain periods with disruptor errors in the
determination (R2), simultaneous regression analysis, previous period. If there is a correlation between disruptor
individual parameter significant test. errors, it can be said that in a linear model there is an
autocorrelation. The autocorrelation test in this study uses the
The data analysis methods used in this study use multiple LM-Test test where there is no autocorrelation if the
linear regression with the help of Eviews programs. Based on significance value of Obs * Square is greater than 0.05.
the hypothesis in this study the method of data analysis used is
quantitative analysis to calculate or estimate quantitatively Heteroskedasticity Test Analysis
from several factors individually – alone or together – equal to The heteroskedastisity test aims to test whether in a
variables Related. The functional relationship between a single regression model there is a variance inequality of the residual
variable associated with a free variable can be done with (error) observation to another observation. If the variance of
multiple linear regression. The method of data analysis in this the residual from one observation to another remains, it is
study uses deskriftive statistics with quantitative data aimed at called homoskedastisity, and if different is called
obtaining a comprehensive picture of the direct or indirect heteroskedastisity. A good regression Model is that
influence between variables in this study. homoskedastisity does not happen heteroskedastisity (Ghozali,
2016) Heteroskedastitas test is done to test regression model in
The classic assumption test in the study used a classical case of variance inequality from the residual of one
assumption test before testing the hypothesis using multiple observation to another observation. If variance occurs
regression analyses. Test the classic assumptions to be used in differences then it is called Heteroskedastitas and if the
this study include: variance is fixed then it is called Homoskedastitas. The
Heteroskedastitas test in this study used the White test where
Normality Test Analysis there is no heterogeneity if the significance value of Obs *
Test data normality is performed to determine whether Square is greater than 0.05.
the residual regression model that is researched is distributed
normally or not. A good regression Model is that has a normal Regression Analysis
or close data distribution. This normality test aims to test Regression analysis is essentially a study of the
whether in a variable regression model bonded and free dependency of dependent variables with one or more
variables have a normal distribution or not. In this study the independent variables with the aim to estimate and predict the
normality test was done using the Jarque-Bera method where average population or average value of dependent variables
the data was said to be a normal distribution if the value of based on value Known independent variables (Ghozali,2013).
significance was greater than 0.05.
Moderated Regression Analysis (MRA) or interaction
Multicollinearity Test Analysis test is a special application of linear multiple regression where
Then the assumption of normality was met The the regression equation contains interaction elements
multicolinearity test aims to test whether a regression model is (multiplication of two or more independent variables) with the
found to be correlated between free (independent) variables following equation formula (Liana, 2013):
(Ghozali, 2016). In case of strong correlation, there is a
problem with multicolinearity to know the presence or absence Y = 𝛼 + 𝛽1 CATA + 𝛽 2 FATA + 𝛽 3 DER + 𝛽 4 CATA* KI +
of multicolinearity in the regression model can be seen from: 𝛽 5 FATA* KI+ 𝛽 6 DER* KI + ê
oth of these sizes indicate each of the independent variables
described by another independent variable. So the low
It can therefore be said that the data in this research is Obs*Rsquared Significance Information
already a normal distribution. The normality test can also be Non
done using the histogram chart present in the attachment with 32,778 0,3798 Heteroskedastisitas
the result on the histogram chart, indicating the normal Table 6:- Heteroscedasticity Value
distribution pattern is to follow or approach the shape of the Source: Secondary data processed in years 2019
bell, so that the regression model meets the normality
assumption. The test results of normality based on significance The above results obtained significance value of Obs *
testing and the histogram chart indicate that the data in this Rsquared is 0.3798 so the value is greater than 0.05 so there is
study is already under normal distribution. no heteroscedasticity, so it can be said that the regression
model does not contain heteroscedasticity symptoms.
Multicollinearity Test Analysis
Multicolinearity tests are performed to see the presence Autocorrelation Test Analysis
or absence of the correlation between independent variables in Heteroskedastitas test is done to test regression model in
testing the model of multiple linear regression. A regression case of variance inequality from the residual of one
Model is good if there is no high correlation between its free observation to another observation. If variance occurs
variables, because if there is a correlation then the variables differences then it is called Heteroskedastitas and if the
are not orthogonal or there are similarities. This test was variance is fixed then it is called Homoskedastitas. The
conducted to avoid partial influences of each free variable Heteroskedastitas test in this study used the White test where
against the bound in the decision making process. This test in there is no heterogeneity if the significance value of Obs *
regression models can be seen by means of the tolerance value Square is greater than 0.05.
and the value of VIF (Variance Inflation Factor). To view the
multicholinerity by using Eviews is to see a value of VIF no Obs*Rsquared Significance Information
more than 10. Non
3,040 0,2186 Autokorelasi
Table 7:- Autocorrelation Values
Source: Secondary data processed in years 2019