The Application of Discriminant Analysis to Determine the

Classifications of Human Trafficking Cases in East Nusa Tenggara

(ENT) Province

Maria Agustina Kleden

, Astri Atti

and Uda Gerardus

Department of Mathematics, Nusa Cendana University, Jln Adisucipto, Kupang, Indonesia

Department of Guidance and Counseling, Nusa Cendana University, Jln Adisucipto, Kupang, Indonesia

Keywords: Human Trafficking, Discriminant Analysis, East Nusa Tenggara.

Abstract: Human trafficking is a common problem to many parts of the world. This worldwide problem is sometimes

perceived as a modern slavery. East Nusa Tenggara (ENT) Province is the region with the highest number

of human trafficking cases in Indonesia which is now alarming both the national and provincial government.

Some factors such as poverty, culture, shortage of income, unemployment, lack of information and not well-

informed of the regulations are believed to be the triggers of this problem. This study aims at classifying the

human trafficking cases in the migrant labours and children trafficking groups based upon the above

mentioned factors. Discriminant analysis was employed in this study. The results shows that the human

trafficking discriminant model is Zscore = 1.204 Economy and culture – 0.878 Work opportunity and

incomes + 0.639 Early-age marriage. The higher the perception score which implies that the more positive

the perception of the respondents or in other words the lower the economic level and culture knowledge of

the respondents then the higher the score of discriminant function. Likewise, the younger the age of the

respondents when they get married resulting in the higher the score of discriminant function.

1 INTRODUCTION

Human trafficking is also known as modern slavery.

International Organization for Migration (IOM) says

that from March 2005 to December 2014, the

number of human trafficking in Indonesia reached

6,651 people. Of that number, 82 percent are women

who work in Indonesia and abroad as informal

workers (Fellowship untuk Jurnalis: Liputan

Investigasi Perdagangan Manusia, n.d.). This shows

that women are the main target of human trafficking.

Human trafficking has become a world problem.

This problem is a form of organized crime that is

very complex to solve. This is in line with the

statement of the international humanitarian agency

(sorooptimist.org) which confirms that “Organized

crime is largely responsible for the spread of

international human trafficking” (Sex

Slavery/Trafficking, n.d.). Human trafficking is a

type of crime with high profits and low risk to be

known. The same thing is emphasized by (Nugroho

& Imelda, 2001). Human trafficking is a growing

criminal activity that involves the movement of

victims by force or coercion for sexual exploitation

or labor. Human trafficking is often facilitated

without being realized by the tourism business

(Paraskevas & Brookes, 2018).

Human trafficking, forced labor and slavery have

become important issues in the current era, but

efforts to counter these problems have also been

severely criticized (McGrath & Watson, 2018). The

global crisis of human trafficking involves

exploitation of people for personal gain and affects

millions of people (Thompson & Haley, 2018).

Human trafficking, including sex trafficking and

labor, is a global problem affecting 20.9 million

people worldwide. The National Human Trafficking

Hotline identified 36,270 human trafficking cases in

the United States since 2007 (Vietor & Hountz,

n.d.). As an extraordinary crime, human trafficking

has a negative impact on individuals, families,

communities and even on national honor. To address

this, the government established RI Law No. 21 of

2007 concerning Eradication of Human Trafficking

Crimes.

The problem of Indonesian workers in the

Province of East Nusa Tenggara (ENT) is a concern

222

Agustina Kleden, M., Atti, A. and Gerardus, U.

The Application of Discriminant Analysis to Determine the Classiﬁcations of Human Trafﬁcking Cases in East Nusa Tenggara (ENT) Province.

DOI: 10.5220/0010139500002775

In Proceedings of the 1st International MIPAnet Conference on Science and Mathematics (IMC-SciMath 2019), pages 222-226

ISBN: 978-989-758-556-2

of the Indonesian Ministry of Social Affairs. There

are very many Indonesian workers who cannot be

absorbed by the foreign labor market. That causes,

Indonesian workers have a very low bargaining

position in the market. The low bargaining position

results in violence against workers abroad. Cases of

violence in Indonesian workers have reached a very

bombastic rate. Cases of violence in Indonesian

workers have reached 4,381 cases. From this

number, ENT has a significant number (Pos Kupang,

2016b). This must receive serious attention from the

Indonesian government and regions that supply

workers abroad.

ENT Regional Police revealed that of 2,279

trafficking victims sent in 2015 and 2016 as many as

451 people were identified as being related to the

handling of human trafficking cases by the ENT

Regional Police (Pos Kupang, 2016a). This

condition makes the ENT Province often referred to

as human trafficking emergency, it was even said to

be ranked first in the problem of human trafficking

in Indonesia.

This problem causes concern and discomfort

from many people. The handling of human

trafficking cases seems very complex, resulting in

victims becoming increasingly helpless in

demanding justice. Most of the people consider the

problem of human trafficking to be a legal and

humanitarian problem. A legal approach in dealing

with the problem of human trafficking is certainly

needed. Without denying the efforts made above, in

the opinion of the researchers it has not touched the

root of the problem of human trafficking in ENT.

In general, factors that cause human trafficking

throughout the world are poverty, globalization, the

sex tourism industry, women's rights, and general

global education levels (Betz, 2009). In ENT

Province, there are several problems that cause

human trafficking cases such as culture, low income,

unemployment, and lack of regulations information

about human trafficking.

An analysis to deal with human trafficking cases

in ENT appropriately needs to be done immediately

considering this case has been very disturbing for

the community. What factors are the most influential

and what factors differentiate groups of victims of

human trafficking. Discriminant models of the forms

of human trafficking cases in ENT can help

determine the factors that distinguish between

discriminant models or groups of human trafficking

cases that are formed.

Discriminant analysis is a multivariate technique

whose main purpose is to separate groups of objects

from two or more populations and allocate an

unknown object from which population to a

predetermined group. This grouping is mutually

exclusive. This means that if object A has been

included in group 1 then it is not possible to also be

included in group 2. Analysis can then be developed

on any variable that makes group 1 different from

group 2, what percentage is included in group 1,

what percentage is enter group 2 and so on (Johnson

& Wichern, 2002).

This discriminant function is used to explain

differences between groups and in classification

problems. In other words the purpose of

discriminant analysis is to arrange the distinguishing

function between group 1, group 2, ..., group k. With

this function, we will get value boundaries between

groups 1, ..., group k. The function can be known the

level of accuracy (what percentage of errors), and

the function is used to group new members into

which groups (Johnson & Wichern, 2002).

Based on the above background, the research

problems are formulated as follow (1) what is the

discriminant model of the forms of human

trafficking in ENT?; (2) What factors distinguish

between the discriminant models of human

trafficking that are formed?. The purpose of this

study were to identify: (1) creating discriminatory

models of the forms of human trafficking cases in

ENT; (2) To determine the factors that differentiate

between discriminant models or groups of human

trafficking cases that are formed.

2 METHOD

The study population was victims of human

trafficking in the province of East Nusa Tenggara

(ENT). The research sample was 48 victims of

human trafficking in South Central Timor District

(SCT) and Belu. The instrument used was a

questionnaire using a Likert scale which contained

37 questions. These questions are also research

variables (attributes). Each question is given a value

of 1 (strongly disagree) to a value of 5 (strongly

agree). The 37 questions were extracted into seven

factors which were used as independent variables,

while the response variable consisted of two groups

namely Y1 = migrant workers and Y2 = child

trafficking.

Data were analyzed using the discrete method to

determine the discriminant model. Discriminant

analysis is a multivariate technique aimed at

separating several groups of objects from two or

more populations. In addition, allocating an

unknown object comes from which population into a

The Application of Discriminant Analysis to Determine the Classiﬁcations of Human Trafﬁcking Cases in East Nusa Tenggara (ENT)

Province

223

predetermined group boundary. In addition,

allocating an unknown object comes from which

population into a predetermined group boundary.

Analysis can then be developed on which variables

make group 1 different from group 2, what

percentage goes to group 1, what percentage goes to

group 2 and so on (Johnson & Wichern, 2002).

3 RESULTS AND DISCUSSION

3.1 Test Homogeneity of Variance

The first assumptions that must be fulfilled in

discriminant analysis are Variants of the

independent variables for each group are

homogeneous and the variants among the variables

are homogeneous. Homogeneity testing of variance

uses the Box’s M value shows in Table 1.

Table 1: Tests Box’s M.

Box's M 33.003

F Approx. 1.026

df1 28

df2 1.161E4

. 0.427

Table 1 shows the sig value = 0.427 > 0.05. This

means that the group covariance matrix is the same.

These results indicate that discriminant analysis is

feasible for this data.

3.2 Analyze the Summary of Canonical

Discriminant Functions

In order to develop a discriminatory function, first

we examine the variables that are influential or

variable those distinguish the two groups. Table 2

shows the results of the feasibility testing of the

variables that will be used in the discriminant

function.

Based on Table 2, seven variables included to

make discriminant models, there are only three

significant. These variables are economic and

cultural (X1), employment and income (X2) and

early marriage (X3). This result shows that the two

forms of human trafficking in ENT occur because of

economic and cultural factors, employment and

income opportunities and early marriage factor.

Table 2: Variables in the Analysis.

ML: Migrant labors

CT: Children Trafficking

Eigenvalue is the ratio between the number of

squares between groups and the number of squares

in a group. The greater eigenvalue indicates better

discriminant function. The eigenvalues in this case

are shown in Table 3.

Table 3: Eigenvalues.

Functi

Eigen

value

% of

Variance

Cumulative

Canonical

Correlation

1 0.493 100 100 0.575

Table 3 shows that the discriminant function has

an eigenvalue of 0.493 and a variance of 100%. This

shows that the discriminant function has been able to

group accurately. Similarly, the canonical

correlation value is 0.575. This value indicates that

there is a strong correlation between the scores of

discriminant functions with groups (child trafficking

and international immigrant workers). This means

that the independent variables used in discriminant

functions can discriminate / differentiate objects into

groups.

Table 4: Wilks' Lambda.

Test of

Function(s)

Wilks'

Lambda

Chi-square df Sig.

1 0.670 22.660 1 0.000

Based on Table 4, it can be seen that the

significance value of Wilks Lambda for discriminant

functions is 0.000 or it can be said to be very

significant. This shows that the independent variable

has a significant effect on the group of objects. In

other words, the result shows a clear difference

between the two groups of respondents (child

trafficking and immigrant labor) based on economy

and culture (X1), employment and income (X2),

early marriage (X3).

The Structure Matrix table shows the correlation

between the independent variables with the

discriminant function that is formed. Variables that

IMC-SciMath 2019 - The International MIPAnet Conference on Science and Mathematics (IMC-SciMath)

224

are not included in discriminant analysis are

variables with low correlation values and are given

the symbol "a" next to each of these variables.

Table 5: Structure Matrix

Function

0.708

0.091

0.511

Table 5 is a matrix structure table. This table

explains the correlation between the independent

variables and the discriminant functions that are

formed. It can be seen that the Economic and

cultural variable (X1) is the variable that has the

highest correlation with the discriminant function

that is formed. The next highest correlation is early

marriage variable (X3) and follows by employment

and income (X2) variable.

3.3 Model of Discriminant

To construct the discriminant model, the

discriminant test is carried out as shown in Table 6.

From the Canonical Function Coefficient table, an

overview of the discriminant model can be obtained.

Table 6: Canonical Discriminant Function Coefficient.

Function

X1 1.204

X2 -0.878

X3 0.639

(Constant) 0.000

Based on Table 6 it can be seen that the

coefficients X1, X2 and X3 are 1,204, -0,878 and

0.639, respectively. Thus the discriminant function

formed is: Zscore = 1204 Economy and culture -

0887 Work opportunities and income + 0.639

Early marriage. This function is used to see new

cases whether they will be classified as child

trafficking groups or immigrant workers.

The discriminant function above shows that

economic and cultural variables and early marriage

are positive. This positive sign indicates that these

variables affect the increase in discriminant scores.

This means that the greater the perception score the

more positive the respondent's perception. In other

words, the lower the economic level and cultural

knowledge of the respondent, the more discriminant

the function score will be. Likewise, the earlier the

respondent's age at marriage, the discriminatory

score of the function will be even greater.

In general, those who are victims of trafficking

are those who have a weak economy. Low education

makes them lack the knowledge and skills to get

decent jobs. Victims with poor economic

background usually have poor social growth. This

condition can affect lifestyles that are manifested in

the desire to become well-off economically.

Association in the environment of people who are

able to meet all their needs, making victims obsessed

with working to meet the needs of their lives that

have not been fulfilled because of the conditions of

their parents or household. The crush of life causes

victims to find solutions through various efforts to

meet their needs without thinking about the risks of

the path they are taking. The desire to earn money

easily makes them find and accept any job without

seeing the risk of the job. Economic factors are also

caused by low education. Victims only have

elementary or junior high school education and find

it difficult to find work that meets their daily needs.

Most of the population of the ENT Province

adheres to patriarchal culture. The social system that

places men as the main authority and dominates

women, children and property is the main cause of

human trafficking. The dominance of men over

women poses no threat to women if there is no

psychological, material or physical harm to women.

In addition, the dominance of men in getting proper

education causes women to lack opportunities to get

higher education.

Other than that, assumption that women are

better able to find work outside the village

encourages parents and husbands to support women

in finding work in cities and even abroad. Women

are considered to have the skills to do household

chores because only that job is suitable for their

level of education. The neglect of the educational

needs of women increases the number of victims of

trafficking. Gender inequality in society causes a lot

of domestic violence.

In addition to economic and cultural factors,

early marriage is also a factor causing trafficking in

persons. In article 7 of Law Number 1 of 1974

concerning marriage, a woman is at least 16 years

old to get married. In Indonesia in 2016 there were

750,000 child marriages. Early marriage like this

lasts the longest two years. This makes the

fulfillment of the family economy more difficult.

This is in line with Dian Kartika Sari, who

emphasized that early marriage was the root cause of

trafficking in persons. Weak economic conditions

The Application of Discriminant Analysis to Determine the Classiﬁcations of Human Trafﬁcking Cases in East Nusa Tenggara (ENT)

Province

225

put women at a disadvantage compared to men

(Kompas, 2017).

Low education of women has an impact on the

lack of job opportunities. The work done is only

limited to domestic helpers. Low levels of

employment caused by low levels of economy and

education encourage urbanization to other cities or

countries as places that are considered easy to find

work. The recruitment of women from villages as a

form of providing legitimate employment

opportunities has an impact on increasing the

number of young women for prostitution. The

suffering and even death experienced by women

who work abroad as domestic workers as revealed

by various media lately is the suffering they are

experiencing because they are women.

4 CONCLUSIONS

Based on the research results described above, the

following conclusions are presented:

1. The discriminant functions obtained are: Zscore

= 1,204 Economy and culture - 0,878

Employment and income opportunities +

0.639 Early marriage.

2. Variables that distinguish the two groups of

Trafficking in Persons are Economic and cultural

(X1), employment and income (X2) and early

marriage (X3).

3. The variable that has the strongest correlation

with the discriminant function that is formed is

the Economic and cultural (X1) followed by the

Early Marriage (X3) and then the smallest

correlation with the discriminant function is the

employment and income (X2).

REFERENCES

Betz, D. (2009). Human Trafficking in Southeast Asia:

Causes and Policy Implications [Naval Postgraduate

School].

http://www.dtic.mil/dtic/tr/fulltext/u2/a501444.pdf

Fellowship untuk Jurnalis: Liputan Investigasi

Perdagangan Manusia. (n.d.).

http://jaring.id/id/enactivities/fellowship/fellowship -

untuk-jurnalis-liputan-investigasi-

perdaganganmanusia/

Johnson, R. A., & Wichern, D. W. (2002). Applied

Multivariate Statistical Analysis. Prentice Hall.

https://books.google.co.id/books?id=VlcZAQAAIAAJ

Kompas. (2017, March 2). Perkawinan Dini Dinilai

sebagai Akar Masalah Perdagangan Orang.

https://nasional.kompas.com/read/2017/03/02/175627

91/perkawinan.dini.dinilai.sebagai.akar.masalah.perda

gangan.orang

McGrath, S., & Watson, S. (2018). Anti-slavery as

development: A global politics of rescue. Geoforum,

93, 22–31.

Nugroho, I. F., & Imelda, J. D. (2001). Perdagangan Anak

di Indonesia. International Labour Office.

Paraskevas, A., & Brookes, M. (2018). Nodes, guardians

and signs: Raising barriers to human trafficking in the

tourism industry. Tour Manag, 67, 147–156.

Pos Kupang. (2016a, December 7). Jumlah Kasus Human

Trafficking Meningkat.

Pos Kupang. (2016b, December 23). Darurat human

trafficking di NTT.

Sex Slavery/Trafficking. (n.d.). Retrieved November 5,

2018, from

http://www.soroptimist.org/trafficking/faq.html

Thompson, J., & Haley, M. (2018). Human Trafficking:

Preparing Counselors to Work with Survivors. Int J

Adv Couns, 40(3), 298–309.

Vietor, T., & Hountz, R. (n.d.). RIGHTS: Response

Initiative Guiding Human Trafficking Services. J

Forensic Nurs.

IMC-SciMath 2019 - The International MIPAnet Conference on Science and Mathematics (IMC-SciMath)

226