Based on Figure 4.1. above shows that the scattered dots are interconnected to form a pattern that
follows the diagonal line. From the picture it does not look too out of the pattern variables (outlier), so the
researchers assume that the data used in this study were normally distributed.
5.2. Data Panel Modeling
5.2.1. Ordinary Least Square Model (common effect)
This technique is no different than making a regression with cross section data or time series. However, for panel
data, before making regression, first must incorporate cross section data with time series data (the data pool).
Here are the results of data analysis using Ordinary Least Square model (common effect) :
Table 4.1. Ordinary Least Square Model (common effect)
Based on the calculation above, it is known that the EPS and ROA equally positive effect on the stock
price with a coefficient of 9.609557 EPS and ROA at 443.3091. R-Squared value also seems quite large: 0.762,
meaning that the independent variables (EPS and ROA) is able to describe the dependent variable (stock price)
of 76.20%. The above results also show that the value of the t-statistic greater in the EPS variable which is equal
to 12.31, but is relatively low at 0.0822 the ROA variable.
5.2.2.Fixed Effect Model (Fixed Effect)
Fixed Effects Model is the same as that used dummy variable regression as independent variables, to distinguish
one object with another object. The effects are still here means that one object, has remained constant magnitude
for various periods of time. Likewise, the regression coefficient, fixed magnitude over time (time invariant).
Here are the results of the analysis of the data using the Fixed Effects Model (Fixed Effect):