Decision Support System for Addressing Demotivated Students: A Comparative Analysis of SAW and TOPSIS Methods

This study aims to determine the most optimal decision support system method and decision alternatives in the Accounting and Management study program at University of X, which has experienced a decrease in learning interest or demotivation. The objective is to provide special care to these students and encourage them to continue their studies. The Simple Additive Weighting (SAW) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) methods are employed, considering criteria such as GPA, study time, parents' income


Introduction
The running of a business or business will never be separated from the intervention of sellers and buyers.The seller sells both goods and services to the buyer then the buyer enjoys the product.However, there are times when the buyer is not satisfied with the products sold by the seller, causing a reduction in the seller's income.
The university is one of the places where business activities occur.In this case the university acts as a seller of services and students are buyers or users of services provided by the university related to learning subjects.Students have the right to be satisfied or dissatisfied with the services provided by the university so that in this case the university can lose students at any time and furthermore lose business income if the services provided to students are not satisfactory.
One of the private universities in Bogor City, also known as University of X, has been experiencing a decreasing trend in the number of students in the Faculty of Economics and Business since 2020.The number of students enrolling in the program has been decreasing year by year.Additionally, there have been several students who Piksel 11 (2): 241 -252 (September 2023) withdrew from their studies in the middle of the semester.One of the factors contributing to this decline is student demotivation.Demotivation is a motivational barrier that can hinder existing motivation (Islam, 2015).This has led to a decrease in the faculty's revenue.
Decision support systems can be a tool that can help make alternative decisions on business problems that are being faced by companies.One method that can be applied to this model is the ranking method, which is a method for finding the optimal value of a few alternatives with certain criteria (Utami et al., 2016), ( Ardiyanti & Mora, 2019).Two methods that are often combined or compared for decision making purposes are Simple Additive Weighting (SAW) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) (Ciardiello & Genovese, 2023).
The Faculty of Economics and Business, University of X can apply decision support system methods, such as the SAW and TOPSIS methods to find answers to the problem of declining student numbers being faced.By processing student data from several criteria such as GPA, study time, and others, it can be determined whether the student is experiencing motivation to study which is likely to eventually drop out of the Faculty of Economics and Business, University of X.
The research from (Veza & Arifin, 2020) developed a dashboard model for a decision support system and applied it to STT Ibnu Sina Batam.The results of this study the model was successfully applied optimally at STT Ibnu Sina.In addition, research related to determining the most optimal decision alternatives has been carried out by (Heriawan & Subawa, 2019) using the SAW and TOPSIS methods to determine the awarding of bidikmisi scholarships.The research resulted in an accuracy rate of 90% compared to manual calculations.The research (Fauziah & Sunardiyo, 2015) also applied a decision support system model to junior high school and high school computer laboratories to determine the feasibility of the infrastructure in these laboratories.The results of the research show that the model built has an accuracy value of 88.67% from 3 software experts, 90% from 2 admins, and 86.23% from 3 visitors.This shows that the application of a decision support system model using various programming languages can be applied to several organizations including schools or campuses.
Based on the research conducted by (Veza & Arifin, 2020) and (Heriawan & Subawa, 2019) on determining alternative inactive students, this study aims to further develop their research by focusing on demotivated students as the research subject.In the study by (Veza & Arifin, 2020) impact on potential inactive students.However, this research will focus more on students and use criteria such as study duration, GPA, parental income, and distance from home to campus.The main theme of this study is to determine the most optimal decision support system method between the SAW and TOPSIS methods to identify demotivated students and reduce the likelihood of a decrease in the number of students due to demotivation issues.

Simple Additive Weighting (SAW) Method
The Simple Additive Weighting method is used to determine alternative decisions to determine demotivated students.The following are the stages in the Simple Additive Weighting method.(Fauzan et al., 2018): g) Setelah melakukan proses normalisasi akan terbentuk matrik ternormalisasi (R) h) After the normalization process, a normalized matrix (R) is obtained.
i) The final step will result in preference values (Vi) obtained by summing the product of R and the preference weights (W).The largest Vi value can be selected as the alternative (Ai).
(2) Piksel 11 (2): 241 -252 (September 2023) The TOPSIS method is used to determine the best alternative based on the range of distances for positive and negative ideal solutions (Bera et al., 2022).The initial steps in the TOPSIS method are the same as the SAW method, namely by determining alternatives (Ai), criteria (Ci), and criteria weights (W).The complete steps in the TOPSIS (3) e) Create a normalized weight rating matrix as follows.Through this formula, positive and negative ideal solutions can be determined in formulas 5 and 6.

Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) Method
=     (4) with Yij having the greatest value or benefit with Yij having the lowest value or cost f) Determine the value of the distance between each alternative value with the positive and negative ideal solution matrices. (7) g) The final step is to determine the preference value to determine the ranking of each alternative.
In general, the work steps for this research can be seen in Figure 1.The work steps start from collecting data sourced from the student unit database at the Faculty of Economics and Business, University of X to obtain the necessary criteria.Then the data is processed respectively using the SAW and TOPSIS methods according to the steps and stages of the method and ends by comparing the two methods to determine the most optimal method that can be applied by the Faculty of Economics and Business.The most optimal method is the method that has the highest preference value between the two methods (NurFaddillah et al., 2023), (Sunarti, 2019).

Data Source
The data for this study used data from 1885 students of the Faculty of Economics and Business, University of X.The data was obtained from the student unit database at University of X in the form of a spreadsheet file.The data consists of criteria for length of study, GPA, parents' income, and distance from home to campus.

Collection and Determination of Data Criteria
This research is intended to find out students who experience demotivation in college learning.Based on research (Damanik, 2020), (Widarto, 2017), and (Andriani, 2010), several criteria were taken that could affect student learning motivation such as GPA scores, length of study, parents' income, and distance from home.The amount of data used is 1,885 student data and their criteria which have been collected into a template in the form of a spreadsheet file obtained from the student affairs database of the Faculty of Economics and Business, University of X to form a matrix as shown in

Data Conversion
The data in table 1 is still in text form, namely in the columns of parental income and distance from home.To simplify the process of calculating the model and referring to research (Fauzan et al., 2018), (Utami et al., 2016) which requires a numeric data type, the data can be converted into a scale form from 1 to 3.An explanation of the scale can be seen in the table below.After the data is converted, the text data type will be changed to a scale of 1 to 3. The data matrix after conversion can be seen in the table below.The data type in the table can already be included in the decision-making system model because it is already in the form of a numeric data type.

Category Weighting
After the data conversion is carried out, the next process can determine the weight value that will be used in the Simple Additive Weighting method.In this study the weighting was determined on a scale of 1-5 which refers to research (Veza & Arifin, 2020).The description of this scale can be seen in table 5 below.Both the SAW and TOPSIS methods have a stage of determining weight values so that after going through the analysis stage of determining weight values and criteria, normalization calculations can be carried out using the SAW and TOPSIS methods.

Calculation Method SAW
Calculation of normalization is done by identifying the use of attributes in each category.After that, the normalization process will be applied using formula 1 above.The results of the normalization produce new values in the range 0 to 1.The data from the normalization results are presented in table 7 below.The results of the process after normalization can be referred to as a matrix after normalization (Fauzan et al., 2018).Piksel 11 (2): 241 -252 (September 2023) 2114236 have the highest preference value so that these students can be included in the category of demotivated students.

TOPSIS Calculation Method
Calculation of normalization in the TOPSIS method starts from the stage of determining the weights which can be seen in table 6.Then it is necessary to carry out the normalization process by making a normalization matrix using formula 3.
In table 9 it can be seen the results of calculating the normalization for each alternative.
The next process is to make a normalized weight matrix using formula 4. The normalized matrix is multiplied by the weight values listed in table 6.
Table 10 presents the normalized weight matrix along with positive and negative ideal solution values for each criterion as follows.Next, it is necessary to determine the distance between each alternative with positive and negative ideal solutions using formulas 7 and 8 respectively.Tables 11 and   12 present the values of alternative distances for positive and negative ideal solutions as follows.Table 12 shows the distances of the negative ideal solution for each student.The last stage of the TOPSIS method is to determine the preference value based on predetermined positive and negative ideal distances.The preference value will be used as the basis for ranking each alternative using formula 9.The alternative with the highest value is used as the best alternative.Table 13 presents the results of calculating preference values using the TOPSIS method.It can be seen that the alternatives at rank one to five have similarities with the results of the SAW method.

Comparison of Results and Preferences of Both Methods
After the alternatives are tested with both methods, to determine the most optimal method, a comparison is made to the two methods.The comparison method is by looking for the highest preference value between the two methods.In Table 14, the rankings of the alternatives between the two methods are consistent from ratings 1 to 4. However, there are differences in the preference values.For Rank 1, TOPSIS obtained a preference value of 0.983, while SAW obtained a value of 0.992.

Conclusion
The SAW method has the highest preference value of 0.992 while the TOPSIS method has the highest preference value of 0.983.The results show that SAW has the highest preference value compared to the TOPSIS method, therefore the Faculty of Economics and Business, X University, Bogor City can implement a decision support system using the SAW (simple additive weighting) method to obtain an alternative for demotivated students.If this alternative is used, the student can be given guidance by the guidance and counseling unit or other related units so that the student can be motivated to return to study, hence, the finances of each study program can gradually improve.
a) Determining alternatives (Ai) b) Determining criteria to be used as the basis for evaluation (Cj) c) Determining preference weights for each criterion (W) d) Determining the fitness value of the criteria e) Constructing a decision matrix (X) that represents the level of compatibility between each alternative (Ai) and each criterion (Cj) f) Performing normalization on the decision matrix (X) by calculating the normalized performance level (rij) of each alternative (Ai) on each criterion (Cj).Maxij is used when the criterion weight uses the maximum value, and minij is used when the criterion weight uses the minimum value.
method are as follows(Azhari et al., 2018)    a) Determine alternatives (Ai) b) Determine the criteria that will be used as the basis for testing (Cj) c) Determine the preference weight for each criterion (W) d) Determine the normalized decision matrix using the following formula.

Table 1 .
Data Matrix Before Conversion

Table 2 .
Information on Parental Income Categories

Table 3
provides information about the categorization of home distance based on scores.

Table 3 .
Information on Home Distance Categories PIKSEL status is accredited by the Directorate General of Research Strengthening and Development No. 225/E/KPT/2022 with Indonesian Scientific Index (SINTA) journal-level of S3, starting from Volume 10 (1) 2022 to Volume 14 (2) 2026.247

Table 5 .
Weight Category The Simple Additive Weighting method has assessment attributes in the form of maximum and minimum values or commonly referred to as Benefit and Cost which are used to calculate normalized values.The weighting values of the criteria and attributes were obtained through observations and interviews with the Faculty of Economics and Business, X University.Table6describes the details of the weights and attributes used in this study.

Table 6 .
Implementation of Weights and Attributes in Data

Table 7 .
Data Normalization Results After normalization is carried out, the final stage is to calculate the preference value as the basis for ranking each of the existing alternatives.Preference calculations can use formula 2 above.A larger Vi value indicates that the alternative can be used as a decision alternative.The results are sorted from the highest preference value to the lowest.It can be seen in table 8 that students with Student Identification Number (NPM)

Table 8 .
Simple Additive Weighting Test Results

Table 9 .
Data Normalization Results

Table 10 .
Normalized Weight Matrix

Table 11 .
Positive Ideal Solution Distance

Table 12 .
Negative Ideal Solution Distance

Table 13 .
Preference Value and Rating

Table 14 .
Comparison of TOPSIS and SAW Preferences Decision Support System for Addressing Demotivated Students: A Comparative Analysis of SAW and TOPSIS Methods PIKSEL status is accredited by the Directorate General of Research Strengthening and Development No. 225/E/KPT/2022 with Indonesian Scientific Index (SINTA) journal-level of S3, starting from Volume 10 (1) 2022 to Volume 14 (2) 2026.251