Evaluating Mobile Learning Adoption in Higher Education based on

New Hybrid MCDM Models

Ming-Tsang Lu

, Gwo-Hshiung Tzeng

1,2

Kua-Hsin Peng

and Shu-Kung Hu

Institute of Project Management, Kainan University, No. 1, Kainan Road, Luchu, Taoyuan 338, Taiwan

Institute of Management of Technology, National Chiao Tung University, 1001 Ta-Hsueh Road, Hsinchu 300, Taiwan

Institute of Leisure and Recreation Management, Kainan University, No. 1, Kainan Road, Luchu, Taoyuan 338, Taiwan

Department of Business and Entrepreneurial Management, Kainan University,

No. 1, Kainan Road, Luchu, Taoyuan 338, Taiwan

Keywords: Mobile Learning, MCDM (Multiple Criteria Decision-Making), DEMATEL, DANP (DEMATEL-based

ANP), VIKOR.

Abstract: This study investigated the mobile learning adoption of evaluation in higher education. Mobile learning is a

new form of learning utilizing the unique of mobile devices. However, students’ readiness for mobile

learning has yet to fully explore in Taiwan. The purpose of this study is to address this issue using a hybrid

MCDM (multiple criteria decision-making) approach that includes the DEMATEL (decision-making trial

and evaluation laboratory) for constructing influential network relationship, DANP (DEMATEL-based

ANP) for finding the influential weights, and VIKOR methods combining the influential weights of DANP

for evaluating the performance gaps in each criterion and then how based on influential network relationship

map (INRM) to reduce gaps for achieving aspiration level. An empirical case as example is illustrated to

show that these hybrid MCDM. By evaluating the influential interrelationships between criteria related to

mobile learning, this approach can be used to solve interdependence and feedback problems, allowing for

greater satisfaction of the actual needs of mobile learning behaviour.

1 INTRODUCTION

This study contributes in higher education in three

ways. First, the adoption of mo-bile learning is

explored from a multi-faceted perspective including

attitude-related behaviors to mobile learning,

perceived behavioral control, and trust-related

behaviors. This implies that university practitioners

should consider these three fac-tors before

employing m-learning. Second, the current study

shows the relative im-portance of perceived

behavior control (i.e., perceptions of internal and

external con-straints on behavior) (Taylor and Todd,

1995) in the decision to adopt mobile learning. That

is, students who are confident with mobile devices

are likely to adopt mobile learning. Hence,

universities need to provide students with training

opportunities about the basic functions and

applications of mobile learning technologies. Lastly,

the current findings reveal that usefulness and ease

of use affect students’ attitude for adopting mobile

learning. Thus, to facilitate the acceptance of mobile

learning, the learning environment should be

perceived as useful and easy to use. A better

understanding of the process of mobile learning

adoption will help researchers and decision makers

work together to implement proper strategies for

mobile learning.

Most of the conventional multi-criteria decision

analysis (MCDA) models cannot handle the analysis

of complex relationships among different

hierarchical levels of criteria. Yet the decision to

adopt mobile learning requires decision model that

does just that. The purpose of the present study is to

address these issues; we develop a hybrid MCDM

model that combines DEMATEL, DANP, and

VIKOR. The hybrid method overcome the

limitations of existing decision models and can be

used to help us analyze the criteria that influence

mobile learning issue. In particular, we use Tai-

wan’s college students as an example to study the

interdependence among the factors that influence the

267

Lu M., Tzeng G., Peng K. and Hu S..

Evaluating Mobile Learning Adoption in Higher Education based on New Hybrid MCDM Models.

DOI: 10.5220/0004389502670271

In Proceedings of the 5th International Conference on Computer Supported Education (CSEDU-2013), pages 267-271

ISBN: 978-989-8565-53-2

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

user behavior of mobile learning in the higher

education as well as evaluate alternative user

behavior processes to achieve the aspired levels of

performance from mobile learning.

2 METHODLOGY

This Section comprises four parts: the first part

presents the DEMATEL technique for building an

influential network relationship; the second part

calculates the influential weights using DANP

(DEMATEL-based ANP); the third, the last part

uses VIKOR to evaluate total accreditation

performance; finally, describes the data collection.

2.1 DEMATEL for Establishing an

Influential Network Relationship

DEMATEL is mainly used to solve complex

problems to clarify their essential nature.

DEMATEL uses matrix and related mathematical

theories (Boolean operation) to calculate the cause

and effect relationships involved in each element.

This technique is widely used to solve various

complex studies, and particularly to understand

complex problem structures and provide viable

problem-solving methods (Tzeng et al., 2007).

DEMATEL is based on the concept of influential

relation map, which can distinguish the

direct/indirect influential relationship of the criteria,

allowing decision-makers to identify the key

criterion for developing strategies for improving

accreditation performance in higher education of this

study.

2.2 Find the Influential Weights using

the DANP

This study not only uses the DEMATEL technique

to confirm the interactive relationship among the

various dimensions/criteria, but also seeks the most

accurate influential weights. This study found that

ANP can serve this purpose. This study used the

basic concept of ANP (Saaty, 1996), which

eliminates the limitations of Analytic Hierarchy

Process (AHP) and is applied to solve nonlinear and

complex network relations (Saaty, 1996). ANP is

intended to solve interdependence and feedback

problems of criteria. This study thus applies the

characteristics of influential weights ANP and

combines them with DEMATEL (call DANP,

DEMATEL-based ANP) to solve these kind of

problems based on the basic concept of ANP. This

approach yields more practical results.

2.3 Evaluating Competitiveness Gaps

using VIKOR

Opricovic and Tzeng (2004) proposed the

compromise ranking method (VIKOR) as a suitable

technique for implementation within MCDM (Tzeng

et al., 2005; Opricovic and Tzeng, 2004; Opricovic

and Tzeng, 2007; Liu and Tzeng, 2012). VIKOR

uses the class distance function (Yu, 1973) based on

the concept of the Positive-ideal (or we adopt the

Aspiration level) solution and Negative-ideal (or we

adopt the Worst level) solution and puts the results

in order. For normalized class distance function it is

better to be near the positive-ideal point (the

aspiration level) and far from the negative-ideal

point (the worst value) for normalized class distance

function.

2.4 Data Collection

Table 1 descripts the framework of dimensions and

criteria. And the data was collected from 32

education experts who understand mobile learning

trend and usage (in consensus, significant

confidence is 96.375%, more than 95%; i.e., gap

error =3.265%, smaller less 5%). Most of the

education experts have teaches more than ten years

in higher education. Expert perspectives on all

criteria within the criteria were collected via

personal interviews and a questionnaire. Expert

elicitation was conducted in Nov., 2012, and it took

60 to70 minutes for each subject to complete a

survey.

Table 1: Framework of dimensions and criteria.

Dimensions Criteria

Attitude-related behaviours

Relative advantage

Compatibility

Complexity

Perceived behavioural

control D

Self-efficacy

Resource facilitating

conditions C

Technology facilitating

conditions C

Trust-related behaviours D

Disposition to trus

Structural assurance

Trust belief

CSEDU2013-5thInternationalConferenceonComputerSupportedEducation

268

3 EMPIRICAL STUDY ANALYSIS

FOR MOBILE LEARNING

ISSUE

In this section, an empirical study is displayed to

illustrate the application of the proposed model for

evaluating and selecting the best method that can

help decision makers to understand how to improve

their evaluations of mobile learning user-behaviour.

3.1 Analysis of Result

In this paper, we confirmed DEMATEL decision-

making structure, and analysed from three

dimensions with 9 criteria of the user-behaviour

perspective on mobile learning. According to the

expert questionnaires, we obtain the total influence

matrix

T of dimensions and criteria shown in Table

2 to Table 3. We find the cognition and opinion from

experts in three dimensions, and the relationship

between the extents of the impact can also be found

which is compared to other dimensions as show in

Table 2.

Table 2: The total effect matrix of

T and sum of effects

on dimensions.

D D

0.827 0.813 0.817 2.457 2.532 4.989 -0.075

0.888 0.784 0.822 2.494 2.338 4.832 0.156

0.817 0.741 0.767 2.325 2.406 4.730 -0.081

According to the total influence prominence

()

d+s , “attitude-related behaviours

()D ” has the

highest influence of the strength of relationship that

means the most important influencing dimensions;

in addition, “trust-related behaviours

()D ” is all the

factors that affect the least degree of other

dimensions. According to the influence relationship

()

ds− , we can also find “perceived behavioural

control

()D

” is the highest degree of influence

relationship that affects other dimensions directly.

Otherwise, “trust-related behaviours

()D ” is the

most vulnerable to influence that compare with other

dimensions. According to Table 3, we can obtain all

the criteria of the impact of relations with each

criterion. And then, from Table 4 shows the

relationship between the extents of the direct or

indirect influences and compares them with other

criteria. “Technology facilitating conditions

()C ” is

the most important considerations criteria; in

addition, “structural assurance

()C

” is the influence

of all criteria in the least degree of other criteria.

Furthermore, we can also find in Table 4 that shows

“self-efficacy

()C ” is the highest degree of

influence relationship in all the criteria. And,

“technology facilitating conditions

()C ”, is the most

vulnerable to impact of criteria that compare with

other criteria.

We use DEMATEL to confirm the influence

relationship with the criteria, and expect to obtain

the most accurate influence weights. The purpose of

DANP is to solve the interdependence and feedback

problems of each criterion (Saaty, 1996). Therefore,

we structure the quality assessment model by

DEMATEL which combination with DANP model

to obtain the influential weight of each criterion as

show in Table 4.

Table 3: The total effect matrix of

for criteria.

0.773 0.848 0.901 0.738 0.768 0.978 0.836 0.790 0.919

0.868 0.827 0.920 0.805 0.816 1.011 0.865 0.804 0.921

0.802 0.812 0.695 0.656 0.687 0.858 0.734 0.680 0.801

0.886 0.940 0.911 0.662 0.784 0.979 0.833 0.774 0.902

0.787 0.857 0.833 0.690 0.617 0.872 0.743 0.704 0.815

0.920 0.954 0.902 0.778 0.795 0.879 0.859 0.815 0.956

0.882 0.882 0.863 0.702 0.737 0.941 0.729 0.800 0.925

0.690 0.698 0.673 0.550 0.580 0.731 0.680 0.541 0.740

0.885 0.894 0.882 0.728 0.748 0.954 0.871 0.809 0.805

Table 4: The gap evaluation of mobile learning by

VIKOR.

D/C

Local

Weight

Global weight

(DANP)

Mobile learning gap

()

0.348 0.197

0.329 0.115 0.113

0.339 0.118 0.213

0.332 0.116 0.266

0.322 0.296

0.300 0.097 0.228

0.310 0.100 0.366

0.389 0.125 0.294

0.331 0.295

0.331 0.109 0.266

0.310 0.102 0.338

0.359 0.119 0.284

Total gaps 0.261

In addition, we can find the critical criteria in

higher education of mobile learning user behaviour

are identified as technology facilitating conditions

()C , trust belief

()C and compatibility

()C .

Furthermore, the influence weights combine with the

DEMATEL technique to assess the priority of

EvaluatingMobileLearningAdoptioninHigherEducationbasedonNewHybridMCDMModels

269

problem-solving based on the gaps identified by

VIKOR method and the influence network

relationship map.

An empirical study involving mobile learning

user behaviour in the multiple stages (intention stage

and adoption stage) are used to evaluate and

improve the total accreditation gaps using the

VIKOR method, as listed in Table 4. Decision

makers can identify problem-solving issues

according to this integrated index, either from the

perspective of the criteria as a whole or from that of

an individual dimension.

Using the overall/dimension criteria, the gap

values can be determined by the priority sequence

improvement for reaching the desired level. In the

intention stage, resource facilitating conditions

()C

with a higher gap value of 0.366, are the first

criterion to be improved.

Improvement priority can also be applied to the

individual dimension. In the attitude-related

behaviours

()D

of the intention stage, for instance,

the priority gap values are ordered as follows:

complexity

()C , compatibility

()C , relative

advantage

()C

. In the perceived behavioural control

()D of the intention stage, the priority gap values

are ordered as follows: resource facilitating

conditions

()C , technology facilitating conditions

()C , self-efficacy

()C . In the trust-related

behaviours

()D of the intention stage, the

improvement priorities can be sequenced as follows:

structural assurance

()C , trust belief

()C ,

disposition to trust

()C . Using the gap values

provided by the panel experts above, improvement

priority schemes are unique and comprehensive,

both from the separate dimensions and from the

overall points of view, as shown in Table 4.

For decision makers, understanding

improvement priorities of mobile learning user

behavior for client must be easier to understand than

the gaps in higher education.

3.2 Discussions and Implications

The empirical results are discussed as follows. First,

according to the DEMATEL model, we could

recognize the interrelationship of each dimension

and criterion the influential relationship network

map for each dimension and criterion (as Fig. 1

shows). In Fig. 1, the perceived behavioral control

()D is affecting other dimensions- attitude-related

behaviors

()D , and, trust-related behaviors

()D ;

visibly perceived behavioral control

()D

plays an

important role and it has the highest and intensity

influence in its relationship to other dimensions.

Thus, higher education leader should first improve

it, then, followed by attitude-related behaviors

()D ,

trust-related behaviors

()D for evaluating and

improving the mobile learning user behaviors in the

higher education.

Second, after analyzing the dimensions, we

would illustrate the considered-criteria in each

dimension. According to the results, we illustrate the

influence relationship-digraph-map of criteria in Fig.

1. Hence, for the influence relationship of these

criteria, in the attitude-related behaviors dimension

()D

: compatibility

()C

was the most influence

criterion and should be improved first, followed by

relative advantage

()C and complexity (see Fig. 1

for more details on the causal relationship in

D ,and

D ). Each of the evaluation dimensions and

criteria creates the necessary behaviors for inducing

mobile learning user behaviors in the higher

education. Therefore, high education leader should

evaluate all of the dimensions and criteria for the

mobile learning user behavior in accordance with

Fig. 1. This evaluation method can be used in most

of the higher education. However, school leader

should keep in mind that, when applying this model,

some differences exist. The level of importance for

the 9 criteria may vary according to the particulars

of each high education, and the school leader should

compare the evaluation methods for each mobile

learning user behavior model before making

deciding upon the optimal using adoption method.

Fig. 1: The influential network relationship map of each

dimension and criterion

Finally, the overall gap values (i.e., the distance

to 0) shown in Table 4 that indicate room for

CSEDU2013-5thInternationalConferenceonComputerSupportedEducation

270

improvement are 0.261 in the intention stage, and

0.235 in the adoption stage. In the multiple-stage

perspective, the perceived behavioral control

()D ,

featuring the largest gap value of 0.296 in intention

stage, and the trust-related behaviors

()D featuring

the largest gap value of 0.295 in adoption stage,

which should be the first priority for improvement if

decision makers wish to achieve the desired level.

For long-term improvement, the decision makers

should manage internal motivation carefully, as

mentioned above. Given these empirical findings,

our results, as holistically formulated in Table 5,

fulfil the purpose of this research. Evaluating the

mobile learning user behaviour model provided by

this study can extend to most higher education using

mobile learning user behaviour. However, school

administrators should be cautious when applying

this model. The importance of the 9 criteria may

vary according to the situation, and administrators

should compare the mobile learning user behavior

and define the gap in that stage before making

decision on optimal technology use.

Table 5: Sequence of improvement priority for mobile

learning user behaviour.

Formula Sequence of

improvement priority

F1:Influential network of

dimensions

()

()D

()C

()D

()C

F2:Influential network of

criteria within individual

dimensions

()

F3:Sequence of

dimension to rise to

aspired/desired level (by

gap value, from high to

low)

()D

()C

()D

()C

()D

()C

F1:Influential network of

dimensions

()D

()C

()D

()C

4 CONCLUSIONS

Mobile learning service has an important role in the

training of higher education. Its decisions are

complicated by the fact that various criteria are

uncertainty and may vary across the different

product categories and use situations. Based on the

export and literature review, we developed the three

dimensions and 9 criteria that align with the mobile

learning service of environment. So we applied the

methodology of hybrid MCDM model combining

DANP with VIKOR in empirical case. The main

reason is among the numerous approaches that are

available for conflict management, hybrid MCDM is

one of the most prevalent. VIKOR is a method

within MCDM; it is based on an aggregating

function representing closeness to the ideal

(aspiration level), which can be viewed as a

derivative of compromise programming for avoiding

“choose the best among inferior alternatives (i.e.,

pick the best apple among a barrel of rotten apples)”.

In a decision-making process, we used the global

and local weights into alternatives performance,

such as that in Table 5, to allow school leader to

select the best mobile learning adoption factor. We

haven't only selected the best factor, but also found

how to improve the gaps to achieve the aspiration

level in mobile learning service performances.

REFERENCES

Liu, J. J. H., Tzeng, G, H., 2012. Comments on “Multiple

criteria decision making (MCDM) methods in

economics: an overview, Technological and Economic

Development of Economy, 18(4), 672-695.

Opricovic, S., Tzeng, G. H., 2004. Compromise solution

by MCDM methods: A comparative analysis of

VIKOR and TOPSIS, European Journal of

Operational Research, 156(2), 445-455.

Opricovic, S., Tzeng, G. H., 2007. Extended VIKOR

method in comparison with outranking methods,

European Journal of Research, 178(2), 514-529.

Saaty, T. L., 1996. Decision Making with Dependence and

Feedback: Analytic Network Process, RWS

Publications, Pittsburgh, New York.

Taylor, S., Todd, P., 1995. Decomposition and crossover

effects in the theory of planned behavior: a study of

consumer adoption intentions, International Journal of

Research in Marketing, 12(2), 137-155.

Tzeng, G. H., Tsaur, S. H., Laiw, Y. D., Opricovic, S.,

2002a. Multicriteria analysis of environmental quality

in Taipei: Public preferences and improvement

strategies. Journal of Environmental Management,

65(2), 109-120.

Tzeng, G. H., Teng, M. H., Chen, J. J., Opricovic, S.,

2002b. Multicriteria selection for a restaurant location

in Taipei. International Journal of Hospitality

Management, 21(2), 171-187.

Tzeng, G. H., Lin, C. W., Opricovic S., 2005. Multi-

criteria analysis of alternative-fuel buses for public

transportation. Energy Policy, 33(11), 1373-1383.

Yu, P. L., 1973. A class of solutions for group decision

problems. Management Science, 19(8), 936-946.

EvaluatingMobileLearningAdoptioninHigherEducationbasedonNewHybridMCDMModels

271