Available online at https://journal.
com/index.
php/ijqrm/index International Journal of Quantitative Research and Modeling e-ISSN 2721-477X p-ISSN 2722-5046 Vol.
No.
3, pp.
298-306, 2025
Multiple Linear Regression Analysis of Factors Influencing Human Development Index By Regency/City in East Java Province in 2024 Nurul Itsnaini1.
Agung Prabowo2*.
Diah Paramita Amitarwati3 Mathematics Study Program.
Faculty of Mathematics and Natural Sciences.
Universitas Jenderal Soedirman.
Purwokerto.
Indonesia Statistics Study Program.
Faculty of Mathematics and Natural Sciences.
Universitas Jenderal Soedirman.
Purwokerto.
Indonesia Sharia Economics Study Program.
Faculty of Islamic Economics and Business.
UIN Saifuddin Zuhri.
Purwokerto.
Indonesia *Corresponding author email: agung.
prabowo@unsoed.
Abstract The Human Development Index (HDI) is an indicator used to assess the success of human development.
The Human Development Index (HDI) is used to measure the impact of efforts to improve basic human capital capabilities.
Based on BPS data, the HDI in East Java has consistently increased and has reached a high category, however, when compared to DKI Jakarta and DI Yogyakarta, the HDI in East Java is still relatively low.
This study aims to determine the factors that influence the HDI in East Java Province.
The research data are the 2024 HDI data for East Java Province obtained from the BPS of Lamongan Regency and the BPS website of East Java Province.
This study uses the multiple linear regression method with RStudio Based on the results of the study, the multiple linear regression model with HLS.
RLS.
UHH, and GK factors has an influence of 97.
94% on the HDI in East Java Province, while the TPT does not show a significant influence on the HDI in East Java Province.
Keywords: Human Development Index.
East Java.
Multiple Linear Regression Introduction One of the government's efforts to improve people's welfare and improve the quality of a region is development (Puspitasari et al.
, 2.
Development is a process of change that is always attempted to improve the welfare of the people in a region (Alkhoiriyah & Sa'roni, 2.
According to Dwi et al.
In the development and economic process, the government has made the quality of human resources its basic module (Alkhoiriyah & Sa'roni, 2.
believes that human resources play an important role in creating development that aims to make society healthy and able to lead a productive life.
Human development basically has four main components, namely: empowerment, productivity, justice, and (Wahyudi et al.
, 2.
By optimizing these four factors, human development can be successful.
Successful human development means that a country's people have the opportunity to live longer and healthier lives, receive a decent education, and utilize their knowledge in productive activities.
One indicator that can be used to measure the success of development implementation is the Human Development Index (HDI).
According to Fitriyah et al.
The Human Development Index can be used as a benchmark to assess the success of efforts to improve the quality of life in a region.
The Human Development Index is crucial information used to measure government performance and allocate resources for the General Allocation Fund (DAU).
(Arum et al.
, 2.
East Java is one of six provinces on the island of Java.
According to Statistics Indonesia (BPS) data, the Human Development Index (HDI) of East Java Province has consistently increased from 2022 to 2024.
In 2024, the Human Development Index (HDI) in East Java reached 75.
This figure is higher than the HDI of Central Java .
and West Java .
However, when compared to the HDI of Yogyakarta Special Region .
Jakarta Special Region .
, and Banten .
East Java's HDI is still relatively lower.
Given the importance of the Human Development Index (HDI) in a region, particularly in East Java, an analysis of the HDI is necessary.
This analysis can be conducted statistically by identifying factors that have a significant influence on the HDI.
One statistical method that can be used to achieve this research objective is multiple linear regression analysis.
Itsnaini et al.
/ International Journal of Quantitative Research and Modeling.
Vol.
No.
3, pp.
298-306, 2025
There are several previous studies on HDI, including: (Arum et al.
, 2.
in his research examined the factors that influence the Human Development Index based on Regency/City in Central Java in 2022.
The research used multiple linear regression analysis with the variables used being the Human Development Index (Y).
Life Expectancy (X.
Expected Years of Schooling (X.
Average Years of Schooling (X.
, and Per Capita Expenditure (X.
The results of the study showed that the variables of life expectancy, expected years of schooling, average years of schooling, and per capita expenditure had a significant effect on the human development index in 2022 and the coefficient of determination or R-Square value was 99.
Other research, namely Puspitasari et al.
which examines the robust regression model for the Human Development Index in East Java with M Estimation in 2019.
The variables used in the study are HDI as the dependent variable and life expectancy, average years of schooling, expected years of schooling, and per capita income as independent variables.
In the study there are outlier data so that the distribution of the residuals is not normal, therefore a robust regression analysis was conducted to overcome the outliers.
The results of the study obtained a determination coefficient value of 99.
91% and all independent variables have a significant effect on the HDI.
The objectives of this research are:
Analyzing the factors that influence the human development index by Regency/City in East Java Province in .
Obtaining a multiple linear regression model for factors influencing the Human Development Index by Regency/City in East Java Province in 2024 using RStudio software.
Literature Review Human Development Index The Human Development Index is an indicator used to assess the success of human development (Susanti & Saumi, 2.
According to Hasibuan et al.
the concept of human development was first introduced by the United Nations Development Programmed (UNDP) in 1990, in its report "Global Human Development Report" which states that human development is defined as the process of expanding individual choices so that they have the opportunity to live a healthy and long life, acquire sufficient knowledge and skills, so that they can utilize these skills in productive activities to improve their quality of life.
The Human Development Index (HDI) is constructed through a three-dimensional approach: a long and healthy life, knowledge, and a decent standard of living.
According BPS 2025, human development achievements in a region at a certain time are grouped into four groups, namely:
Very High Group : HDI Ou 80
High Group
: 70 O HDI O 80
Medium Group
: 60 O HDI O 70
Lower Group
: HDI < 60
Regression Method Simple Linear Regression According to Wahyudi et al.
Simple linear regression analysis is an approach method for modeling the relationship between one dependent variable and one independent variable.
As for according to Montgomery et.
The form of the simple linear regression model equation is:
: vbound variables : constant .
: kregression coefficient .
: vindependent variables : standard error Multiple Linear Regression The multiple linear regression method is a development of simple linear regression.
In simple linear regression, only one independent variable is used, while in multiple linear regression, more than one independent variable is used (Arum et al.
, 2.
According to Maharadja et al.
Multiple linear regression is a method for making predictions involving two or more variables: an influencing variable and an affected variable.
These variables are interrelated or have a causal relationship (Daniya et al.
, 2.
The form of the multiple linear regression model equation is:
Itsnaini et al.
/ International Journal of Quantitative Research and Modeling.
Vol.
No.
3, pp.
298-306, 2025
: vbound variables : vindependent variables : kregression coefficient : error Classical Assumption Test Normality Test According to Mardiatmoko .
The normality test is used to determine whether residual values are normally A good regression model is one with normally distributed residual values.
Normality tests can be performed using statistical tests such as Kolmogorov-Smirnov.
Shapiro-Wilk, etc.
The hypotheses used are:
H0: Residuals are normally distributed .
H1: Residuals are not normally distributed Furthermore, the testing criteria are as follows:
If p-value > , then is accepted .
If p-value < , then is rejected Multicollinearity Test According to Anastashya et al.
Multicollinearity testing is a test that aims to determine whether there is a correlation between independent variables in a regression model.
To determine the presence of multicollinearity, the Variance Inflation Factor (VIF) value is used.
According to (Susanti & Saumi, 2.
if the VIF value is < 10 then it is free from multicollinearity and if then multicollinearity occurs.
Autocorrelation Test According to Susanti and Saumi .
the autocorrelation test aims to test whether in a linear regression model there is a correlation between the error of the disturbance in period t and the error in the previous t-1.
According to Mardiatmoko .
autocorrelation testing can be performed using the Durbin-Watson (DW) test.
Other methods include the Breusch-Pagan test, the Run Test, and others.
The criteria for autocorrelation testing in regression analysis are shown in Table 1:
Table 1 : Durbin-Watson Test Criteria Durbin-Watson Statistical Value Results positive autocorrelation occurs.
without decision.
there is no positive/negative correlation.
without decision.
negative autocorrelation occurs.
: Durbin-Watson Value : Durbin-Watson Lower Bound : Durbin-Watson Upper Limit Heteroscedasticity Test According to Mardiatmoko .
heteroscedasticity is a condition where there is inequality in the variance of the residuals for all observations in the regression model.
According to Susanti and Saumi .
heteroscedasticity testing can be done using the scatterplot method by observing the points in the image.
If the points in the image are spread out around 0 on the Y-axis and do not form a pattern, heteroscedasticity does not occur.
Model Feasibility Test Simultaneous Test (F Tes.
According to Susanti and Saumi .
the F test is used to determine whether or not there is a joint or simultaneous influence between independent variables on the dependent variable.
The hypothesis used is:
o significant effec.
: there is .
ignificant impac.
Itsnaini et al.
/ International Journal of Quantitative Research and Modeling.
Vol.
No.
3, pp.
298-306, 2025
Furthermore, the decision criteria for simultaneous testing are:
If , then is rejected .
If , then is accepted Coefficient of Determination (R.
According to Mardiatmoko .
determination analysis is a measure that shows how much variable X contributes to variable Y.
A small value means that the ability of the independent variable is very limited in explaining the variation in the dependent variable (Anastashya et al.
, 2.
Research Methods The method used in this research is a literature study method, namely the researcher conducted a literature study by searching, studying, and also understanding material on the Human Development Index (HDI) and multiple linear regression methods from various sources such as journals, e-books, theses, and other references.
The data used in this report is secondary data on the Human Development Index (HDI) in East Java Province obtained from the BPS Lamongan Regency and the official website of the BPS East Java Province.
The dependent variable used is the Human Development Index (Y), while the independent variables consist of 4 variables, namely the Open Unemployment Rate (X.
Expected Years of Schooling (X.
Average Years of Schooling (X.
Life Expectancy (X.
, and Poverty Line (X.
The data analysis used is multiple linear regression analysis with data processing using RStudio.
The stages in this research include data collection and determining the independent and dependent variables to be The next step is to conduct classical assumption tests, including normality, multicollinearity, autocorrelation, and heteroscedasticity.
After ensuring that all classical assumptions are met, model feasibility tests are conducted using simultaneous tests (F test.
, partial tests .
, and the coefficient of determination (R.
Results and Discussion Data The data used in this report is secondary data on the Human Development Index (HDI) in East Java Province in The data is available in Table 2.
Table 2 : Human Development Index (HDI) Data in East Java Province in 2024 No.
Regency/City Pacitan Regency Ponorogo Regency Trenggalek Regency Tulungagung Regency Blitar Regency Kediri Regency Malang Regency Lumajang Regency Jember Regency Banyuwangi Regency Bondowoso Regency Situbondo Regency Probolinggo Regency Pasuruan Regency Sidoarjo Regency Mojokerto Regency Jombang Regency Nganjuk Regency Madiun Regency Magetan Regency Ngawi Regency Bojonegoro Regency Tuban Regency Lamongan Regency Gresik Regency Bangkalan Regency Sampang Regency Pamekasan Regency HDI
(%)
TPT
(%)
HLS
RLS
UHH
Poverty Line (Rp/Month/Capit.
370,643 413,619 434,146 447,793 408,399 403,621 420,334 405,136 459,043 470,713
517,741
413,611
537,724
450,088
597,284
508,618
514,170
539,714
460,205
455,119
445,865
471,457
488,131
524,636
608,828
547,017
491,753
467,493
Itsnaini et al.
/ International Journal of Quantitative Research and Modeling.
Vol.
No.
3, pp.
298-306, 2025
No.
Regency/City HDI
(%)
Sumenep Regency Kediri City Blitar City Malang City Probolinggo City Pasuruan City Mojokerto City Madiun City Surabaya City Batu City TPT
(%)
HLS
RLS
UHH
Poverty Line (Rp/Month/Capit.
506,569 621,051 596,105 706,341 654,409 554,195 610,968 637,838 742,678 642,778
Classical Assumption Test Normality Test Hypothesis H0: Residuals are normally distributed H1: The residuals are not normally distributed.
Test Statistics The Shapiro-Wilk test with RStudio gives the results in Table 3.
Table 3.
Shapiro-Wilk Test Results with RStudio p-value Significance Level Decision Based on Table 3, p-value .
> .
, then H0 is accepted.
Conclusion It can be concluded that the residuals are normally distributed .
ormal data distributio.
Multicollinearity Test From the multicollinearity output results in Table 4, the VIF value for variable X1 is 1.
X2 is 3.
X4 is 2.
and X5 is 2.
The VIF value of each variable is less than 10, so it can be concluded that there is no multicollinearity in the data.
Table 4: Multicollinearity Test Results with RStudio
VIF
1,367052
3,102448
6,114509
Autocorrelation Test Hypothesis H0: Residuals are independent .
here is no autocorrelatio.
H1: Residuals are mutually dependent .
here is autocorrelatio.
Test Statistics The Durbin-Watson test with RStudio gives the results in Table 5.
Table 5: Durbin-Watson Test Results with RStudio
1,4279
In the Durbin-Watson table:
umber of dat.
= 38 k .
umber of independent variables .
= .
= 6 dL (Durbin Watson lower limi.
= 1.
dU (Durbin Watson upper limi.
= 1.
Significance Level p-value 2,725217 Itsnaini et al.
/ International Journal of Quantitative Research and Modeling.
Vol.
No.
3, pp.
298-306, 2025
Decision dL d dU Conclusion Thus, no conclusion can be drawn because it is an area of doubt/no decision Heteroscedasticity Test Based on the Residual vs Fitted graph in Figure 1 produced by RStudio, it can be seen that the data distribution is spread out and does not form a certain pattern or gather at a certain point, so it can be concluded that heteroscedasticity does not occur.
Model Feasibility Test Simultaneous Test (F Tes.
Hypothesis H0:
, meaning that the independent variables simultaneously do not have a significant H1:
, meaning that the independent variables simultaneously have a significant influence.
Test Statistics The F test with RStudio gives the results in Table 6.
Table 6: Test Results F with RStudio Fcount Ftable 2,512255 Figure 1: Residual vs Fitted Significance Level Decision Fcount .
> Ftable .
, then H0 is rejected.
Conclusion Thus, it can be concluded that the independent variables simultaneously have a significant influence in the Partial Test .
-Tes.
Hypothesis H0:
H1:
Test Statistics , meaning that there is no influence of the first independent variable on the dependent , meaning that there is an influence of the first independent variable on the dependent Itsnaini et al.
/ International Journal of Quantitative Research and Modeling.
Vol.
No.
3, pp.
298-306, 2025
The t-test with RStudio gives the results in Table 7.
Table 7: Test Resultst with RStudio
1,322
4,112
10,505
2,150
2,803
Significance Level Decision criteria If tcount > ttable or tcount < - ttable, then H0 is rejected.
Conclusion Based on Table 4.
10, the independent variable X1 shows that t count < t table, so H0 is accepted, meaning there is no influence of the independent variable X1 on the dependent variable Y.
Furthermore, variables X2.
X3.
X4 and X5 show that each t count > t table, so H0 is rejected, meaning there is an influence of the independent variables X2.
X3.
X4 and X5 on the dependent variable Y.
Coefficient of Determination Based on the output in Table 8 obtained from RStudio (Figure .
, the independent variables in the model have an influence of 0.
9794 or 97.
94% on the dependent variable, and the remainder is influenced by other variables not included in the model.
Table 8: Results of the Determination Coefficient Test with RStudio Multiple R-squared Adjusted R-squared Figure 2: RStudio Output for Regression Coefficient and Determination Coefficient Multiple Linear Regression Based on the results of processing and analysis carried out with RStudio (Figure .
, the multiple linear regression model obtained is as follows:
It can be interpreted that:
The intercept value obtained is -2.
This indicates that when all independent variables are zero, the predicted value of Y is approximately -2.
The coefficient value for the TPT (X.
variable is 0.
This shows that every 1% increase in TPT (X.
will increase the HDI by 0.
Itsnaini et al.
/ International Journal of Quantitative Research and Modeling.
Vol.
No.
3, pp.
298-306, 2025
The coefficient value for the HLS variable (X.
This shows that every 1-year increase in HLS (X.
will increase the HDI by 0.
The coefficient value for the RLS variable (X.
This shows that every 1-year increase in RLS (X.
will increase the HDI by 1.
The coefficient value for the UHH variable (X.
as big as 0.
This shows that every 1-year increase in UHH (X.
will increase the HDI by 0.
The coefficient value for the GK variable (X.
This shows that every 1 unit increase in GK (X.
will increase the HDI by 0.
Conclussion and Suggestions Based on the results of multiple linear regression analysis and model feasibility test, it was found that the variables that significantly influence the HDI at a 5% significance level are the Expected Years of Schooling (HLS).
Average Years of Schooling (RLS).
Life Expectancy (UHH), and Poverty Line (GK).
Meanwhile, the Open Unemployment Rate (TPT) variable did not show a significant influence on the HDI because the calculated t value was smaller than the t table.
This study shows that increases in variables that significantly influence the HDI tend to have a positive impact, particularly in education and health.
The higher the expectations, average years of schooling, and life expectancy, the higher the HDI value in a Conversely, an increase in the poverty line has a negative impact on the HDI, making poverty alleviation efforts crucial.
The suggestions for this research are:
The addition of other relevant variables, such as labor force participation rate, per capita expenditure, or economic growth, can be considered in subsequent research.
Use of data over a longer time span to provide more comprehensive and dynamic results.
The use of other analysis methods, such as panel data regression or logistic regression with each variable adjusted References