Assessing Critical Factors Affecting the Mass Adoption of IoT in Bangladesh

Abstract

Internet of Things (IoT) is a system in which objects can exchange data among themselves by being interconnected with the help of the internet over a wireless connection without human intervention. IoT is a buzz word in the modern era. A developing country like Bangladesh may face different problems while trying to adopt IoT due to the lack of technological knowledge, underdeveloped infrastructure and deficiency of relevant resources. But to stay apace with the fast-growing world, Bangladesh should also cast aside existing systems and accept IoT with open arms. This research aims to find out the most important factors which might be responsible for IoT adoption in Bangladesh keeping in mind about its socio-economic conditions. Some hypotheses were formed to create a questionnaire for the survey. Indicator variables found from the survey were grouped into 5 (five) factors using Exploratory Factor Analysis (EFA). A measurement model was created based on the hypotheses formed previously. Structural Equation Modelling (SEM) technique was used to find out the estimates of different factors associated with IoT adoption. From the structural model, it was found that Affordability and Ability had the highest regression weight. This suggests that in developing countries like Bangladesh people give priority to products being affordable rather than thinking about its usefulness or the positive changes it would bring to society. Besides a huge portion of people are not highly educated because of which they have expressed concerns over their ability to use IoT devices efficiently and safely. So all these factors were incorporated in this study which explains its significance and relevance.

Full text

International Conference on Mechanical, Industrial and Energy Engineering 2020
19-21 December, 2020, Khulna, BANGLADESH
* Corresponding author. Tel.: +88-01764282138
E-mail addresses: toukir.ahmed@ipe.ruet.ac.bd
ICMIEE20-261
Assessing Critical Factors Affecting the Mass Adoption of IoT in Bangladesh
Souvik Chakraborty1, Nabila Khayer2, Toukir Ahmed3,*
1 Department of Industrial & Production Engineering, Rajshahi University of Engineering & Technology, Rajshahi-6204,
BANGLADESH
2 Department of Industrial & Production Engineering, Rajshahi University of Engineering & Technology, Rajshahi-6204,
BANGLADESH
3,* (Corresponding Author) Lecturer, Department of Industrial & Production Engineering, Rajshahi University of Engineering
& Technology, Rajshahi-6204, BANGLADESH
ABSTRACT
Internet of Things (IoT) is a system in which objects can exchange data among themselves by being interconnected with the
help of the internet over a wireless connection without human intervention. IoT is a buzz word in the modern era. A developing
country like Bangladesh may face different problems while trying to adopt IoT due to the lack of technological knowledge,
underdeveloped infrastructure and deficiency of relevant resources. But to stay apace with the fast-growing world, Bangladesh
should also cast aside existing systems and accept IoT with open arms. This research aims to find out the most important
factors which might be responsible for IoT adoption in Bangladesh keeping in mind about its socio-economic conditions. Some
hypotheses were formed to create a questionnaire for the survey. Indicator variables found from the survey were grouped into 5
(five) factors using Exploratory Factor Analysis (EFA). A measurement model was created based on the hypotheses formed
previously. Structural Equation Modelling (SEM) technique was used to find out the estimates of different factors associated
with IoT adoption. From the structural model, it was found that Affordability and Ability had the highest regression weight.
This suggests that in developing countries like Bangladesh people give priority to products being affordable rather than
thinking about its usefulness or the positive changes it would bring to society. Besides a huge portion of people are not highly
educated because of which they have expressed concerns over their ability to use IoT devices efficiently and safely. So all these
factors were incorporated in this study which explains its significance and relevance.
Keywords: Internet of Things, Affordability, Adoption, Exploratory Factor Analysis, Structural Equation Modelling.
1. Introduction
Bangladesh is a densely populated developing
country. Despite of all the obstacles Bangladesh is
encountering, it has achieved nearly 8% gross domestic
product (GDP) growth and is one of the fastest-growing
economies in the world [1]. The basic fixing behind this
development has been the savvy utilization of ICT in
pretty much every sector. Its economy largely relies on
the working-class people that are bereft of the basic
needs of life. Bangladesh is focused to ensure 100%
internet connectivity by 2021 and steps have been taken
to ensure 5G network throughout the country for fastest
speed [2]. It introduces the term ‘IoT’ (Internet of
Things).
Internet of Things (IoT) has acquired outstanding
attention within the last decade. This phenomenal
innovation makes a new world where a good range of
services are being offered by billions of smart,
interacting gadgets to close and remote entities. Humans
have been introduced as smart operators by IoT to
regulate and supervise activities. IoT significantly
expands productivity, lowers health costs, reduces
carbon footprints, enlarges access to education in
remote places and among underserved communities,
improves transportation safety, encourages smart homes
and cities and thus improves the standard of life [3]. IoT
not only makes life easier and reduces the restrictions
but also provides innovative changes for the lifestyle of
people [4]. So IoT is a must for Bangladesh to realize its
vision ‘Digital Bangladesh’.
Firstly, there is a literature review of the previous
works about the adoption of IoT in different countries.
Then some hypotheses were developed based on which
the survey would be performed. Next, the results were
analyzed to know about the impact of different factors
on IoT adoption in Bangladesh. Lastly, the conclusion
recapitulates the importance of the research and advises
future research prospects.
2. Literature Review
In recent years a lot of work has been done regarding
the smooth transition of IoT from the existing market
leaders for mass people in different countries. Roy,
Zalzala and Kumar (2016) experimented on the
acceptance of IoT among the urban poor in India and
developed a model to facilitate the IoT adoption process
among them. They observed the lifestyle of the urban
poor to identify possible sectors in which IoT can be
adopted [5]. Hernandez (2019) observed the constraints
which restrain the poor people from accessing digital
services in Bangladesh. He singled out availability,
awareness, affordability, ability, agency and gender as the
key barriers for adoption in Bangladesh [6]. AlHogail and
AlShahrani (2019) developed a conceptual model based
on the Technology Acceptance Model (TAM) and
identified trust as the key component responsible for the

ICMIEE20-261-
2
adoption of IoT and emphasized on building trust [7].
Wireko, Hiran and Doshi (2018) in their research
discussed about the cultural and social bias as the main
hindrance in IoT acceptance in Ethiopia. The Unified
Theory of Acceptance and Use of Technology
(UTAUT) was used in their study to analyze the
response from the participants [8]. Gao and Bai (2014)
considered the impact of technology factors, social
context factors and individual user characteristics to
build an integrated model on the IoT adoption in China
[9]. Luqman and Belle explored the non-technical
factors associated with acceptance of IoT in rural areas
and considered basic needs, skepticism, care for the
community, safety and security as the main factors [10].
Hsu and Yeh (2016) used the TOE model to develop a
base model for assessment and the DEMATEL method
to evaluate the cause and effects of the factors of the
base model [11]. Hsu and Lin (2016) examined the
relationship between benefits and sacrifices made to the
adoption of IoT. They applied the SEM technique to
analyze the results of the survey [12]. Kamble et.al
(2019) depicted the inter-relation between barriers using
a two-staged integrated ISM and DEMATEL
methodology in a food-retail supply chain [13].
Atayero, Oluwatobi and Alege (2016) compared the
readiness of IoT adoption of Sub-Saharan Africa with
other parts of the world using the Global Innovation
Index, Global Competitiveness Index and Knowledge
Economic Index [14]. Shafique, Ali and Salman (2019)
mentioned about the needs of implementation of IoT for
the reduction of poverty in the rural areas of Pakistan
[15]. Swami and Bhargava (2019) explained the
importance of digital security and the need to protect
data from hackers while adopting digital services [16].
Hopalı and Vayvay (2018) explored the usability
challenges of IoT in developing countries. They
described lack of standardization, underdeveloped
infrastructure, security issues etc. as the main hurdles
for a developing country [4]. Kautsarina and
Kusumawati (2018) used PRISMA protocol to
understand the various uses of IoT in the rural areas of
Indonesia and studied the potential implications of IoT
adoption in that region [17]. AlHogail (2018)
investigated on product-related factors, social influence
related factors and security-related factors in building
trust which in turn increases the tendency to adopt IoT
[18]. Sharma et.al (2020) applied TISM approach, the
Fuzzy MICMAC model, and the DEMATEL method to
analyze the 15 IoT adoption barriers that were identified
for waste management in developing economies [19].
Janssen et.al (2019) pointed out the challenges of
adopting and implementing IoT in smart cities and
analyzed them using integrated MICMAC-ISM
approach [20].
Most of the previous studies done in this area of
research concentrated on building a conceptual
framework of the barriers by using either technical or
non-technical factors. The prior studies have been
conducted in different parts of the world but no such
work has been done in Bangladesh in the past where IoT
is relatively a new word. As sooner or later Bangladesh
has to adopt IoT to achieve its agenda of ‘Digital
Bangladesh’, so this research aims to find out the
barriers of IoT adoption in the context of Bangladesh
where a large section of people lack technological
knowledge. The factors which are most relevant in
relation to Bangladesh were considered for getting
accurate estimates.
3. Hypothesis Development
It should be considered that nearly 20% of the
people in Bangladesh live below the poverty line when
speaking about mass adoption [21]. Around 26% of the
adult do not know how to read and write [22]. So
keeping these in mind, some critical factors for mass
adoption of IoT were taken into account.
3.1 Affordability and Ability (A.A):
Almost 24% of the population of the country do not
have access to electricity and a lot of people do not even
have a stable 4g internet connection. Besides the cost of
the internet is too high for people living hand to mouth.
Most of the places do not have free public Wi-Fi though
the use of IoT technologies require continuous
connection to the internet [6].
Lack of content in Bengali creates great problems
in communication for people having less acquaintance
with English. Besides the use of IoT requires a
significant level of technological knowledge.
3.2 Awareness and Trust (A.T):
Almost 67% of the people of Bangladesh lack
awareness about the process of using the internet [6].
They don’t know what IoT is and wouldn’t even
recommend their known ones to change their prevalent
lifestyle.
Less educated people in Bangladesh believe that
sharing their data in the IoT cloud will make them
vulnerable both financially and personally [7]. Because
of the great level of IT involvement people feel a source
of greater uncertainty [9].
3.3 Usefulness (U):
People tend to underestimate the usefulness of IoT
as the impacts are intangible. They tend to neglect the
benefits of using IoT in the long run. IoT can increase
the efficiency of a system by multiple folds. Consumers
are only interested in adopting IoT if it provides a
distinct advantage than existing systems almost
immediately. Usefulness can play a significant role in
the tendency to adopt IoT [9].
3.4 Ease of Use (E.U):
Ease of use refers to giving minimal effort required
to satisfactorily use a product. Convenience in use can
lead to an increased tendency to use a technology [7].

ICMIEE20-261-
3
For people to adopt IoT, they need to have a sense of
feeling that they can use it comfortably [9].
So it can be assumed that all the factors mentioned
from 3.1-3.4 have a strong positive impact on consumer
tendency for Adoption of IoT (A). To test the
hypotheses a survey was performed and based on the
result it was decided whether to accept or reject the
hypotheses.
4. Methodology
Data Collection:
The survey was conducted over people from
different classes, income groups and different age
groups in Bangladesh over the course of 4 weeks.
Participants of the study were provided with sufficient
information and a description of how IoT devices work
in general so that they could form a perspective about
their use of IoT technology in the future. This approach
was taken so that the participants could avoid any
unclear understanding of IoT devices due to the lack of
technological knowledge that could lead to faulty data
collection. Then the respondents were asked to fill the
questionnaire on a Likert scale ranging from strongly
disagree (1) to strongly agree (5).
Among the 400 questionnaires distributed 380 were
used for empirical analysis. 20 questionnaires were
discarded due to invalid responses or missing data.
Concerning the age group of the participants, 29.3
percent were between 20 and 25 years old, 36.8 percent
were aged between 26-35 years, 17.7 percent between
36-45 years, 8.9 percent between 46 to 55 and the rest
above 56 years of age.
With the help of Structural Equation Modelling
(SEM), the results obtained from the survey were
analyzed.
5. Result Analysis and Discussion
The results obtained were analyzed in three parts.
Firstly, an Exploratory Factor Analysis was performed
to reduce the redundant factors associated with the
measurement model. Then the validity and reliability of
the measurement model were tested. Lastly, the
structural model was examined to find the impact of the
residual factors on the Adoption of IoT and to find the
model fitness.
5.1 Exploratory Factor Analysis:
5.1.1 Measuring Sampling Adequacy:
First, the sampling adequacy was examined to
determine the appropriateness of factor analysis using
the Kaiser Meyer Olkin (KMO) and Bartlett’s Test
measure.
Table 1 KMO and Bartlett’s Test.
Kaiser-Meyer-Olkin Measure of Sampling
Adequacy
0.905
Bartlett’s Test of
Sphericity
Approx. Chi-Square
7087.298
df
276
Sig.
0.000
If KMO > 0.5, the sample is adequate. Here, KMO
= 0.905 which indicates that the sample data is adequate
and the factor analysis can proceed. Taking a 95% level
of significance, α = 0.05, the p-value (Sig.) is 0.000.
The p-value is significant as it should be less than 0.05.
Therefore the Factor Analysis is valid [23].
5.1.2 Factor reduction:
Initially, 24 variables were used in Factor Analysis.
However, not all of 24 variables were retained.
Variances of the factors are represented by Eigenvalues.
The first factor always has the highest variance and
hence the highest Eigenvalue. Factors with Eigenvalue
greater than 1 should be considered for the study [23].
The cumulative percentage gives the summation of the
percentage of variance accounted by preceding factors
and the present one. In this case study, the first 5 factors
explain almost 75% of the variance.
Table 2 Total Variance.
Factor
Initial Eigenvalues
Total
% of
Variance
Cumulative
%
1
9.188
38.284
38.284
2
3.042
12.674
50.958
3
2.331
9.714
60.672
4
1.710
7.126
67.798
5
1.643
6.847
74.645
6
0.948
3.949
78.593
7
0.535
2.231
80.825
8
0.483
2.013
82.837
9
0.460
1.915
84.753
10
0.422
1.757
86.510
11
0.370
1.542
88.052
12
0.366
1.526
89.577
13
0.314
1.309
90.886
14
0.297
1.240
92.126
15
0.282
1.174
93.299
16
0.248
1.035
94.334
17
0.212
0.884
95.218
18
0.201
0.836
96.054
19
0.196
0.818
96.873
20
0.185
0.770
97.643
21
0.169
0.704
98.347
22
0.157
0.653
99.000
23
0.144
0.599
99.600
24
0.096
0.400
100.000
Based on Promax Rotation with Kaiser
Normalization, 5 (five) factors have been extracted by
maximum likelihood. The percentage of non-redundant

ICMIEE20-261-
4
residuals is 4% which is considered good as the
percentage should be less than 5% [23]. 19 variables
were clubbed into 5 (five) factors. These 5 (five)
extracted factors explain 74.645% of the total variability.
5.2 Measurement Model:
Overall, the measurement model produces an
excellent fit with Minimum Discrepancy = 2.051, GFI =
0.956, CFI = 0.972, TLI = 0.963, NFI = 0.948, RMSEA
= 0.053, P-value = 0.000 [12, 24].
5.2.1 Measuring Convergent Validity:
As displayed in the Table 3 below, to assess
Convergent Validity the Average Variance Extracted
(AVE) was examined from the standardized loadings of
the indicator variables. All the standardized loadings are
greater than 0.7 and all AVEs exceed 0.5. Thus the
model has good convergent validity [9].
5.2.2 Measuring Discriminant Validity:
Discriminant validity was calculated by the square
root of AVEs. From Table 4 it can be seen that the
square root of AVEs ranging between 0.776 and 0.891
is consistently greater than the off-diagonal entries (co-
relation between latent variables) which are in the range
0.239 to 0.653. This shows good discriminant validity
[12].
5.2.3 Measuring Composite Reliability:
Table 3 shows the value of Composite Reliability
(CR). All the CRs exceed 0.7. This indicates that there
is good internal consistency reliability between the
respective latent and indicator variables [9, 12].
Table 3 Table for Measuring Convergent Validity and
Composite Reliability.
Ind.
Variables
Latent
Variables
Std.
Loadings
AVE
CR
Q5
E.U
0.905
0.694
0.900
Q4
E.U
0.881
Q2
E.U
0.731
Q1
E.U
0.804
Q13
U
0.814
0.794
0.939
Q12
U
0.948
Q11
U
0.929
Q10
U
0.868
Q18
A.T
0.822
0.718
0.927
Q17
A.T
0.808
Q16
A.T
0.887
Q15
A.T
0.843
Q14
A.T
0.873
Q8
A.A
0.725
0.712
0.880
Q7
A.A
0.907
Q6
A.A
0.887
Q22
A
0.821
0.603
0.820
Q23
A
0.750
Q24
A
0.756
Fig.1 Values of AVEs and CRs of Latent Variables
Fig.1 is the graphical representation of Table 3
where each latent variable has AVEs greater 0.5. And
the values of CRs exceed 0.7.
Table 4 Table for Measuring Discriminant Validity.
Latent
Variable
U
E.U
A.T
A.A
A
U
0.891
E.U
0.335
0.833
A.T
0.476
0.497
0.847
A.A
0.520
0.239
0.358
0.844
A
0.544
0.371
0.523
0.653
0.776
Fig.2 Assessment of Discriminant Validity
Discriminant validity can be clearly observed from
Fig.2 as we can see that the diagonal entries (Square
root of AVEs) from Table 4 are far greater than the off-
diagonal (Co-relation between Latent Variables) ones.
5.3 Structural Model:
To measure the impact of the antecedent factors on
the adoption of IoT technologies in Bangladesh the
Structural Equation Modelling (SEM) technique was
used. The SEM model displayed an absolute fit as P-
value = 0.000, GFI = 0.945, RMSEA = 0.072 were
within the acceptable range. The model also had an
incremental fit as the value of CFI (0.948), NFI (0.923)

ICMIEE20-261-
5
and TLI (0.932) all exceeded 0.9 and parsimonious fit
as Minimum Discrepancy (2.950) was less than 5 [25].
Fig.3 shows the structural model in which we can see
the impact of different factors on IoT adoption.
Fig.3 Structural Model.
Table 5 shows that Ease of use (E = 0.106 and p-
value = 0.030), Awareness & Trust (E = 0.274 and p-
value = 0.000), Usefulness (E = 0.166 and p-value =
0.006) and Affordability & Ability (E = 0.502 and p-
value = 0.000) all have a positive influence on the
Adoption of IoT technologies in Bangladesh.
Table 5 Standardized Regression Weights.
Estimate(E)
P-value
A
<---
E.U
0.106
0.030
A
<---
A.T
0.274
0.000
A
<---
U
0.166
0.006
A
<---
A.A
0.502
0.000
However, Affordability & Ability is the factor that
strongly influences the IoT adoption and Ease of use has
the lowest impact on Adoption. So, it can be concluded
that the overall SEM model was significant in
identifying the impact of different factors on the
Acceptance of IoT.
6 Conclusion
This study tried to incorporate different non-
technical aspects of IoT adoption such as Affordability
and Ability, Awareness and Trust which cannot be
ignored because of the background of Bangladesh.
These factors needed attention thinking of the socio-
economic condition of Bangladesh. As these factors
were taken into consideration, the results obtained were
also significant enough to suggest that the study was
relevant. It was found that Affordability and Ability had
the highest estimate of 0.502 and Ease of Use had the
lowest impact in the lot with an estimate of 0.106. All
the other factors too had a positive impact on IoT
adoption in Bangladesh which reinstates the
assumptions made in the study. Practitioners and
researchers have acknowledged the importance of IoT
adoption as a key element for social and economic
development in developing countries. This study serves
as a theoretical base for understanding the implications
of IoT adoption in Bangladesh. The factors found in the
study can be taken into consideration to ensure
participation of mass people in practical applications of
IoT in Bangladesh in the near future. However, this
research has some limitations. Firstly, the study was
based on only Bangladesh though it intended to be more
consistent with respect to the context of other
developing countries. As other countries might have
different IT regulations and socio-economic conditions,
this research should enlarge its boundaries in the future.
Secondly, a generalized survey was created describing
the unique characteristics of IoT to the consumers. In
future, the survey might be narrowed down to a
particular product or IoT product category to improve
the rigor of the study. Thirdly, the survey was based on
the tendency to use IoT in the future rather than asking
for customer experience about using IoT devices in
personal. In future researches, user experience should be
taken into account so that other factors may be
considered which might not have been used in this study.
Still, this research provides a good understanding of the
Adoption of IoT in Bangladesh and provides an
incentive for future researches.
7 References
[1]https://www.weforum.org/agenda/2019/11/banglades
h-gdp-economy-asia/ (20.08.2020)
[2]https://en.m.wikipedia.org/wiki/Vision_2021
(20.08.2020)
[3]https://www.gsma.com/iot/iot-opportunities-impacts/
(20.08.2020)
[4] Hopalı, E. & Vayvay, Ö., Internet of Things (IoT)
and its Challenges for Usability in Developing
Countries, International Journal of Innovation
Engineering and Science Research, vol. 2, pp. 6-9, 2018.
[5] Roy, A., Zalzala, A.M.S & Kumar, A., Disruption of
things: a model to facilitate adoption of IoT-based
innovations by the urban poor, Humanitarian
Technology: Science, Systems and Global Impact 2016,
vol. 159, pp. 199-209, 2016.
[6] Hernandez, K., Barriers to Digital Services
Adoption in Bangladesh, K4D Helpdesk Report 573.
Brighton, UK: Institute of Development Studies, 2019.
[7] AlHogail, A., AlShahrani, M., Building Consumer
Trust to Improve Internet of Things (IoT) Technology
Adoption, International Conference on Applied Human
Factors and Ergonomics, Advances in
Neuroergonomics and Cognitive Engineering. AHFE
2018, pp. 325-334, 2019.
[8] Wireko, J.K., Hiran, K.K. & Doshi, R., Culturally
Based User Resistance to New Technologies in the Age
of IoT In Developing Countries: Perspectives From
Ethiopia, International Journal of Emerging Technology
and Advanced Engineering, vol. 8, pp. 96-105, 2018.
[9] Gao, L. & Bai, X., A unified perspective on the
factors influencing consumer acceptance of internet of
things technology, Asia Pacific Journal of Marketing
and Logistics, vol. 26, pp. 211 – 231, 2014.
[10] Luqman, A. & Belle, J.V., Analysis of human
factors to the adoption of Internet of Things-based
services in informal settlements in Cape Town, 1st

ICMIEE20-261-
6
International Conference on Next Generation
Computing Applications (NextComp), Mauritius, pp. 61-
67, 2017.
[11] Hsu, C.W., & Yeh, C.C., Understanding the factors
affecting the adoption of the Internet of Things,
Technology Analysis & Strategic Management, vol. 29,
pp. 1089-1102, 2016.
[12] Hsu, C.L. & Lin, J.C.C, Exploring Factors
Affecting the Adoption of Internet of Things Services,
Journal of Computer Information Systems, vol. 58, pp.
49-57, 2016.
[13] Kamble, S.S., Gunasekaran, A., Parekh, H. & Joshi,
S., Modeling the internet of things adoption barriers in
food retail supply chains, Journal of Retailing and
Consumer Services, vol. 48, pp. 154-168, 2019.
[14] Atayero, A., Oluwatobi, S.O., & Alege, P.O., An
Assessment of the Internet of Things (IoT) Adoption
Readiness of Sub-Saharan Africa, Journal of South
African Business Research, vol. 2016, 2016.
[15] Shafique, M.U., Ali, W. & Salman, M., Rural
Development of Pakistan with IoT, Asian Journal of
Research in Computer Science, vol. 3, pp. 1-9, 2019.
[16] Swami, A.C. & Bhargava, D.R., Digital Security
for Smart Cities in India: Challenges and Opportunities,
IOSR Journal of Engineering (IOSRJEN), vol. 9, pp. 63-
71, 2019.
[17] Kautsarina & Kusumawati, D., The Potential
Adoption of the Internet of Things in Rural Areas, 2018
International Conference on ICT for Rural Development
(IC-ICTRuDev), Badung Regency, Indonesia, pp. 124-
130, 2018.
[18] AlHogail, A., Improving IoT Technology Adoption
through Improving Consumer Trust, Technologies 2018,
vol. 6, 2018.
[19] Sharma, M., Joshi, S., Kannan, D., Govindan, K.,
Singh, R. & Purohit, H.C., Internet of Things (IoT)
adoption barriers of smart cities’ waste management:
An Indian context, Journal of Cleaner Production,
122047, 2020.
[20] Janssen, M., Luthra, S., Mangla S.K., Rana, N.P.,
& Dwivedi, Y.K., Challenges for adopting and
implementing IoT in smart cities: An integrated
MICMAC-ISM approach, Internet Research. 29(6):
1589-1616, University of Bradford, 2019.
[21]https://www.aljazeera.com/news/2020/01/banglades
h-people-live-poverty-line-200126100532869.html
(22.08.2020)
[22]https://www.thedailystar.net/frontpage/literacy-rate-
in-bangladesh-2019-100-pc-still-far-cry-1796734
(22.08.2020)
[23]https://shodhganga.inflibnet.ac.in/bitstream/10603/9
094/10/10_chapter%208.pdf (23.08.2020)
[24] Hoyle, R.H., Evaluating Model Fit, Structural
Equation Modeling Concepts, Issues and Applications,
SAGE Publication, New Delhi, pp. 76-99, 1995.
[25]https://www.youtube.com/watch?v=nFml5AQwwH
Y&list=PLijnwxDUqRrx3EvR1nJerZMSbOwEg82d&i
ndex=6 (23.08.2020)