Introduction

According to the Business Dictionary, digitalization is the “Integration of digital technologies into everyday life by the digitization of everything that can be digitized” (Business Dictionary, 2018). Therefore, in these times where digitalization has already integrated into many areas of our lives, it is obvious that traditional teaching methods at universities will not be unaffected. They will have to be rethought if indeed digitalization is going to improve traditional learning. Active learning, like using Business Simulation Games, is not a new method (Oblinger, 2004). Going back in history, Naylor found out that in the 1950s, the first management decision game has been invented aptly named “Top Management Decision Game” (Naylor, 1971). But of course, with the increase of knowledge in management areas and the rapid improvement of technologies over the last years, the business simulation games in recent times do not compare to the games 60 or 70 years ago (Chiu et al., 2005).

Apart from that, Faria and Wellington found out that about 98% per cent of universities in the United States are using Business Simulation Games in order to teach management skills (Faria & Wellington, 2004). Looking ahead, Vlachopoloulos thinks that Business Simulation Games are an indispensable method in educating managers. This is because they get the chance to link theory and team working without taking the risks that occur in the real world (Vlachopoulos & Makri, 2017).

Knowing all those facts, the question then arises, in which way do business simulation games contribute to the students learning experience? And what are the factors that influence them to use it? These are the questions that will be answered in this research paper. In the beginning, the author is going to give a background on established theories and models, followed by the data analysis of a conducted survey and finally, the recommendations and conclusion. The results of the analysis are based on the answers of students that took part in a management course in developed and developing countries, using the simulation game “Cesim”.

Problem Statement

Traditional teaching methods like theoretical learning do not provide the optimal platform for students to link abstract concepts and real-world problems (Prado et al, 2020). The research aims to find out whether Business Simulation Learning can be an effective tool to assist students in entrepreneurial decision making. The question then arises; what is the traditional method of teaching? It refers to the mode of teaching used back in the day and is regarded to be conventional and basic. It largely entails memorization by rote and recitation. There was minimal student engagement and the learner would remain relatively silent for the duration of the class unless prompted by the teacher to participate. It is grounded in the rationale that because the teacher knows and the learner does not, the learner has no business doing anything other than absorbing the content provided by the teacher (Aguilera-Castillo et al., 2020). The teacher was also responsible for enforcing the standards of behaviour. The student was subjected to periodic tests and assessments to gauge their ability to memorize and replicate what they had learned.

On the other hand, progressive modern teaching methods increase the participation of the student and use methods with a strong formative characteristic (Sirbu et al., 2015). Modern methods of teaching recognize and appreciate the fact that different learners have different learning needs and that they should be handled on an individual basis. Unlike traditional methods of learning that focus on learning by hearing, modern methods focus on learning by discovery and curiosity. They are not rigid but are flexible and relaxed, allowing co-operation between learners and teachers (Hallinger & Wang, 2020). While traditional methods seek to drill content into learners, modern methods are more interested in creating inner motivation within the learner.

Some of the different modern approaches to teaching are demonstrating, collaborating, questioning, modelling and explaining. Demonstrating means to offer a new perspective to the learner. Collaboration is where learners work together in teams which facilitate peer to peer learning. Questioning is a method of testing by the teacher to measure how much the learner has absorbed and retained. Modelling is using visual aids to support the learning process. Explaining is giving a talk or discourse on a specific topic with the aim of transferring understanding to the listener (Morin, 2020).

Zulfiqar, et al., (2019) explains that the business simulation games increase the student understanding about the different aspects of business and also increase their intention to start their business in Pakistan and China. Business simulation games incorporate elements of cooperation, modelling, and demonstration. They combine game elements within simulation programs to put the learner into a scenario they might find themselves in in real life. (Aguilera-Castillo et al., 2020) Simulations have a large array of benefits. They equip learners with key decision-making skills, they provide learners with a preview of what happens in real-life practice, it engages the students more and keeps them more interested and absorbed in the exercise thereby increasing comprehension and retention. There is also an element of building cooperation and synergy among learner groups and teams (Hallinger & Wang, 2020).

The efficiency of business simulation games depends largely on the different variables that are at play. The learner will desire to use the simulation if they perceive that it will add value to their lives. Moreover, if the learner believes the simulation to be easy enough to use, then they will desire to use and re-use it again. In the same vein, if the learner believes that there will be an adequate reward at the end for their troubles, they will be more invested in the simulation game (Morin et al., 2020). This paper seeks to understand how Business Simulation games can augment and integrate into mainstream learning. We also seek to find out what factors influence the learners desire to use the simulation again, and which variables and theories are attached to these factors.

Research Question

Q1: Are Business Simulation Games perceived as an effective teaching tool for entrepreneurial education?

Q2: What are the relevant variables and theories that have an impact on the intention, to reuse Business Simulation Games?

Q3: To what extent do the independent variables of the proposed Models influence the intention to continue using Business Simulation Games?

Research Objectives

Objective 1: To explore if the simulation helped students understand and integrate previous business course concepts in ways that enable them to apply the concepts in the future.

Objective 2: To analyze the relationship between the different factors impacting the intention to continue using Business simulation games

Objective 3: To measure how much the different variables influence the intention to continue

Literature Review

VOSviewer is the software used to get the first insight into the topic of Business Simulation learning. According to the Nees Jan van Eck & Ludo Waltman, who is the developer of the software, it can be used to create maps based on network data and for visualizing and exploring these maps (van Eck & Waltman, 2018). The main database which is used by the software for analyzing existing journals is called Web of Science. During the period of the research, the databases were extended by e.g. Springer Link and Search Engines like Google Scholar. The following picture gives a first impression of the software:

The research within the database can be conducted by different parameters like searching for a general topic, a particular title or a specific author. Moreover, the publishing year of the journal can be limited to a certain period. The author used keyword "Business Simulation Learning" and limited the period for the last 18 years, so in the result, no journals published earlier than in 2000 were considered. The first result was the following figure 2 :

Figure No. 2 shows the first outcome of the network visualization of VOSviewer. All in all, there were 364 journals found on the web of science, related to Business Simulation Learning. The circles in the Figure represent a term from the title and abstracts for a given topic of the publications in the data set. The terms are located based on the co-occurrences in the titles and abstracts. The higher the number of co-occurrence of two terms, the closer they are located on the map (van Eck & Waltman, 2018) The colours of the terms represent different categories related to a topic e.g. the red terms like "model", "technology", "simulation result", which could be summarized as technological science. The green labelled circles e.g. "study", "education", "university" can be summarized as education science. Furthermore, the blue labelled terms e.g. "success", "influence", "originality value" can be put into the category of personality science. The last category covered by the terms “Project”, “Business Process” can be generalized as Project Science. The table also shows how the different terms are linked with each other. The number of links between two terms indicates the number of publications, in which they occur together. (van Eck & Waltman, 2018) After having explored the general map in a more detailed way, by zooming closer into the different categories, the result was the following Figure 3:

The outcome of the deeper analysis shows, that there are many existing journals on the database Web of Science related to Business Simulation Learning, using the terms "study", "student" and "model". This leads the author to go for further research on these keywords to discover the gap that still needs to be filled in the existing literature. The outcome of the whole research process is, that there is already existing literature on Business Simulation Learning and its impacting factors, but only for the very small size of the target group, which will be expanded by the following work. To understand the research model, in the following different models and theories will be explained.

Technology Acceptance Model (TAM)

This is a theory developed by Davis, Bagozzi & Warshaw (1989), as an offshoot of the Theory of Reasoned Action by Fishbein & Ajzen (1975). TAM tries to measure to what extent a person will accept the new information technology presented to them. It can be used to eplain how willing a person is to adopt new technology, as well as exploring different factors that can impact the user's acceptance. Two variables form the backbone of TAM. The first is perceived usefulness (PU). This is how useful the user believes the new technology will be, and what kind of benefits will accrue to them if they adopt the technology. The second variable is the perceived ease of use (PEU). This is how easy the user believes the technology will be to use. The harder it is to use the less likely it is that the user will adopt the technology. In the research, TAM is an independent variable. It has a positive correlation with the dependent variable. It also has a significant impact on the dependent variable (Zulfiqar et al., 2019).

Constructivism Theory

This is a theory that was created by Jean Piaget. It states that teaching activities should be designed for teach Merrill (1991). The teacher is responsible for creating an ideal environment for learning. Shen (2008) posited that simulation games require environments that allow the learner to put what they have learned into practice. There are 3 variables that underpin this theory. The first is learning performance. This indicates how well the student can learn. The better the student can learn, the better the learning environment is. The second is the classroom climate. A better classroom climate is indicative of a better-constructed learning environment. The Constructivism theory is an independent variable that has a positive correlation with the dependent variable. It also has a significant impact on the dependent variable.

Motivational Theory

Motivation is why people do things. It is indispensable when it comes to impacting learning performances. This can further be divided into three, the theories of behaviour, humanism, and cognition. The motivation theory of behaviour states that learning is the source of motivation. The humanism theory of behaviour suggests that the teacher's roll over and above instilling knowledge is creating an adequate learning environment. Finally, the cognition principle says that behaviour is controlled by instincts, incentives as well as our beliefs, our thoughts and our expectations. The motivational theory is based on two variables. The first is perceived attractiveness. This indicates how attractive and appealing the learning experience is to the learner. Higher attractiveness would mean higher motivation. The second is perceived playfulness indicates how fun and easy it will be to play the simulation game. The better the game is perceived to be, the more motivation it is to play. The Motivational theory is an independent variable that has a positive correlation with the dependent variable. It also has a significant impact on the dependent variable.

Agency Theory (AT)

This theory was developed between 1960 and 1970 when economists linked human relationships with risk-sharing. Jensen & Meckling (1979) applied this idea to create AT. They describe agency as the relationship between a principal and an agent based on a contract.

It is based on 3 variables. The first is incentives. This refers to the incentive the principal gives the agent to keep them working in their best interest. The second is goal conflicts (Moon & Kim, 2001). This comes about when the aims of the agent tend towards self-gain rather than having the principal’s best interests at heart. This is handled through achieving goal congruence. The last variable is risk aversion. The risk appetites for the agent and the principle might differ, and this might create a difference in the opinion of which action is best (Tao, Cheng & Sun, 2009). The Agency theory is an independent variable that has a positive correlation with the dependent variable. It also has a significant impact on the dependent variable.

Expectation Confirmation Theory (ECT)

This is a theory developed by R.L Oliver in 1980 to be used in the field of marketing. The theory states that the initial tendency to purchase a commodity will affect the consumer’s behaviour to purchase it next time Oliver (1980)8. A higher level of satisfaction will translate into a greater willingness to purchase the product in future. This theory is underpinned by 3 variables. The first is the confirmation level. This indicates how close the experience was as compared to the user's expectation. The second is the satisfaction level. This refers to how much satisfaction the buyer derived from the purchase. The third is the continuance intention. This is the extent to which the buyer intends to buy the product again. The higher the intention the more likely the buyer will make a repeat purchase. In the research, ECT is our dependent variable and it is influenced by the other 4 theories.

Theoretical Framework

In this research, 5 theories of note were used. 4 theories were independent variables, with one being the dependent variable.See Figure 4:

Hypotheses

H1: TF has a positive relationship and has a significant impact on CF.

H2: EF has a positive relationship and has a significant impact on CF.

H3: LF has a positive relationship and has a significant impact on CF.

H4: AF has a positive relationship and has a significant impact on CF.

Methodology

The purpose of the study is hypothesis testing. We are interested in examining the relationship between different variables and theories with Business Simulation Learning. The unit of analysis is students from 3 universities around the world. We collected data from 103 students from Hochschule Bonn Rhein-Sieg in Germany, University of Nairobi in Kenya and GCUF University in Pakistan.

Our investigation was correlational. We sought to find out what factors influence a user to want to use and reuse a Business Simulation Learning module. We also wanted to establish whether a relationship existed or not between modern learning methods and Business Simulation Learning. Our sample n=103. We used a non-probability sampling method to create our sample. We collected data from students who participated in our business simulation game. The students who took part in the game belonged to classes that were not picked at random.

Our investigation was correlational. We sought to find out what factors influence a user to want to use and reuse a Business Simulation Learning module. We also wanted to establish whether a relationship existed or not between modern learning methods and Business Simulation Learning. Our sample n=103. We used a non-probability sampling method to create our sample. We collected data from students who participated in our business simulation game. The students who took part in the game belonged to classes that were not picked at random.

The research was cross-sectional. The questionnaire was only administered once per university and after that, there were no further interactions between the researcher and the subjects. We, therefore, viewed all the 103 students as a representative sample. We used an online questionnaire to collect data from the students. The questionnaire had a few nominal scales, but the overwhelming majority of the questions used an interval scale. The results of the questionnaire were then compiled and analyzed using SPSS. The reliability test needs to be done to check two important things: Firstly, to check if the questions of the survey are understood by the audience and secondly to see if the items of the questionnaire measure, what they should measure. In this work, multiple questions Likert scale was used. A ratio to check the reliability of a Likert scale is called “Cronbach’s Alpha”. Cronbach’s alpha is one way of measuring the strength of the consistency of the survey (Cronbach, 1951). The closer the result is to 1, the higher is the degree of covariance of the items. A high value for Cronbach's alpha means, that the items used, probably fit well to analyze a given topic. A score of more than 0.7 days, that the questions are understandable and can be used to analyze the topic. A value below 0.7, is going to be rejected. In the following table, the Cronbach's Alpha for the following five variables was calculated: Technological Factor, Emotion Factor, Learning Factor, Agency Factor. The result is the following Table 1:

Table 1 Analysis Of Cronbach’s Alpha
Variables	Cronbach's Alpha
Technological Factor	0.947
Emotion Factor	0.939
Learning Factor	0.938
Agency Factor	0.945
Confirmation Factor	0.933

Data Collection

The sample size was 103 respondents. Only one respondent was below the age of 17. 14 were between 18-20 while 65 were aged between 21 and 23. The remaining 23 were aged 24-26. 61% were male while 39% were female. We collected data exclusively from 3 universities, Hochschule Bonn Rhein-Sieg in Germany, University of Nairobi in Kenya and GCUF University in Pakistan. Observe below Tables

Table 2 Age Group Of All Respondents
Age	Frequency	Percent	Valid Percent	Cumulative Percent
Valid<17	1	1	1	1
18-20	14	13.6	13.6	14.6
21-23	65	63.1	63.1	77.7
24-26	23	22.3	22.3	100
Total	103	100	100

Table 3 Gender Of All Respondents
Gender	Frequency	Per cent	Valid Percent	Cumulative Percent
Valid Male	61	59.2	59.2	59.2
Female	42	40.8	40.8	100
Total	103	100	100

Table 4 Respondents From Different Universities
Name of the University	Frequency	Per cent	Valid Percent	Cumulative Percent
H-BRS (Germany)	17	16.5	16.5	16.5
UoN (Kenya)	53	51.5	51.5	68
GCUF (Pakistan)	33	32	32	100
Total	103	100	100

Results and Discussion

Descriptive Analysis

To get an overview of the whole scale, descriptive statistics is a useful tool. The results of the questionnaire are summed up in the following table 5:

Table 5 Analysis Of Descriptive Statistics
Descriptive Statistics	N	Minimum	Maximum	Mean	Std. Deviation
Technological Factor				3.128	0.8863
Emotion Factor				3.23	0.9483
Learning Factor	103	1	5	3.412	0.7878
Agency Factor				3.102	0.8104
Confirmation Factor				3.023	1.0794
Valid N (listwise)

In the first column of the table, you can see the variables:

• Technological Factor

• Emotion Factor

• Learning Factor

• Agency Factor

• Confirmation Factor

N describes the number of students that took part in the questionnaire. Furthermore, a Likert scale was used from 1 (Strongly Agree) up to 5 (Strongly Disagree). The mean for every variable is 3 or higher, which means that the students mostly agreed with the statements and questions which have been given. The standard derivation by how much the members of a group differs from the mean value for the group.

Measurement Model

Ringle & Sarstedt (2016) explain most of the recent used Partial Least Square (PLS) method Structure Equation Modelling (SEM). This study also used PLS-SEM for data analysis. Joe F Hair, Ringle & Sarstedt (2011) state that while using SEM the measurement model was evaluated by testing Internal Consistency Reliability (ICR), Convergent Validity (CV), and Discriminant Validity (DV). Internal Consistency Reliability (ICR) is a way to gauge how well a test or survey is measuring what we want to measure. It is also called the degree to which the item measures the latent variables (Farooq, 2018; Hair et al, 2011). Composite Reliability is used to measure the internal consistency reliability (Ringle & Sarstedt, 2016). The values of CR having greater than 0.70 shows that all variables are reliable (Joseph, 2019).

The results of composite reliability shows in that Technology Factor (TF) (0.899), Emotion Factor (EF) (0.927), Learning Factor (LF) (0.922), Agency Factor (AF) (0.849) and Confirmation Factor (CF) (0.965) measurements hold high internal consistency reliability.

As presented in Table number 6, the results of the second run indicated that all constructs have achieved a satisfactory level of AVE, Technology Factor (TF) (0.641), Emotion Factor (EF) (0.68), Learning Factor (LF) (0.598), Agency Factor (AF) (0.532) and Confirmation Factor (CF) (0.823) measurements thereby confirming the CV. Although other items of Learning Factors (V3_Learning_LM_2 0.636) have factor loadings less than the standard criterion (0.708), all items were maintained as other items of the same construct have achieved desired AVE scores (0.5) (Avkiran, 2017; Hair et al., 2017).

Table 6 Realibility And Validity
		Factor Loading	AVE	Cronbach Alpha	Composite Reliability
Technology Factor	TF		0.641	0.861	0.899
	V1_Techno_PE_1	0.746
	V1_Techno_PE_2	0.796
	V1_Techno_PE_3	0.821
	V1_Techno_PU_2	0.797
	V1_Techno_PU_3	0.841
Emotion Factor	EF		0.68	0.905	0.927
	V2_Emotion_PA_1	0.845
	V2_Emotion_PA_2	0.74
	V2_Emotion_PA_3	0.892
	V2_Emotion_PP_1	0.765
	V2_Emotion_PP_2	0.862
	V2_Emotion_PP_3	0.832
Learning Factor	LF		0.598	0.903	0.922
	V3_Learning_CC_1	0.737
	V3_Learning_CC_2	0.771
	V3_Learning_CC_3	0.825
	V3_Learning_CC_4	0.715
	V3_Learning_LM_2	0.636
	V3_Learning_LM_3	0.829
	V3_Learning_LP_1	0.84
	V3_Learning_LP_2	0.813
Agency Factor	AF		0.532	0.788	0.849
	V4_Agency_IN_1	0.794
	V4_Agency_IN_2	0.824
	V4_Agency_RA_1	0.713
	V4_Agency_RA_2	0.666
	V4_Agency_RA_3	0.631
Confirmation Factor	CF		0.823	0.957	0.965
	V5_Conf_CI_1	0.93
	V5_Conf_CI_2	0.919
	V5_Conf_CI_3	0.914
	V5_Conf_CL_1	0.877
	V5_Conf_CL_2	0.911
	V5_Conf_SL_1	0.891

Discriminant Validity

Fornell Larcker Criterion (see Table 7):

Table 7 Fornell Larcker Criterion
	CF	EF	AF	LF	TF
Confirmation Factor	0.907
Emotion Factor	0.854	0.824
Agency Factor	0.71	0.631	0.73
Learning Factor	0.845	0.807	0.639	0.774
Technology Factor	0.76	0.773	0.528	0.758	0.801

Another test for the discriminant validity of measurement models was performed by evaluating all cross-loading values of constructs' indicators. As a rule of thumb, indicators of measurement models should have the highest loading on their underlying latent construct, as compared to other constructs involved in the structural model (Farooq et al., 2017; Hair et al., 2017). Complete list of cross loading values of all indicators involved in the constructs of reflective measurement models is presented in Table 8. As per the findings presented in Table 7 all indicators (measurement scale items).

Table 8 Cross Loading
	TF	EF	LF	AF	CF
V1_Techno_PE_1	0.746	0.534	0.527	0.364	0.454
V1_Techno_PE_2	0.796	0.506	0.584	0.444	0.531
V1_Techno_PE_3	0.821	0.668	0.65	0.495	0.668
V1_Techno_PU_2	0.797	0.689	0.615	0.384	0.659
V1_Techno_PU_3	0.841	0.66	0.638	0.42	0.678
V2_Emotion_PA_1	0.646	0.845	0.61	0.527	0.661
V2_Emotion_PA_2	0.513	0.74	0.57	0.489	0.588
V2_Emotion_PA_3	0.734	0.892	0.728	0.588	0.795
V2_Emotion_PP_1	0.527	0.765	0.579	0.442	0.669
V2_Emotion_PP_2	0.696	0.862	0.751	0.491	0.747
V2_Emotion_PP_3	0.677	0.832	0.729	0.577	0.74
V3_Learning_CC_1	0.504	0.555	0.737	0.552	0.549
V3_Learning_CC_2	0.55	0.57	0.771	0.436	0.59
V3_Learning_CC_3	0.553	0.605	0.825	0.491	0.646
V3_Learning_CC_4	0.414	0.525	0.715	0.433	0.564
V3_Learning_LM_2	0.445	0.497	0.636	0.36	0.559
V3_Learning_LM_3	0.659	0.697	0.829	0.499	0.714
V3_Learning_LP_1	0.699	0.735	0.84	0.555	0.781
V3_Learning_LP_2	0.771	0.743	0.813	0.596	0.757
V4_Agency_IN_1	0.565	0.547	0.585	0.794	0.641
V4_Agency_IN_2	0.613	0.625	0.68	0.824	0.654
V4_Agency_RA_1	0.228	0.353	0.309	0.713	0.426
V4_Agency_RA_2	0.119	0.294	0.289	0.666	0.365
V4_Agency_RA_3	0.178	0.373	0.307	0.631	0.395
V5_Conf_CI_1	0.681	0.8	0.823	0.69	0.93
V5_Conf_CI_2	0.682	0.806	0.787	0.616	0.919
V5_Conf_CI_3	0.747	0.771	0.754	0.659	0.914
V5_Conf_CL_1	0.685	0.754	0.728	0.593	0.877
V5_Conf_CL_2	0.725	0.775	0.756	0.692	0.911
V5_Conf_SL_1	0.617	0.741	0.745	0.61	0.891

Structure Model

The structural model was assessed for the overall explanatory power of constructs through R square value, predictive relevance through Q square value and path coefficient β-values. Findings of the structural model are presented in Figure 2.

These results indicate that the proposed model has 52.1% of explanatory power for confirmation factor with R²=0.827. Moreover, it is found that the relationship between Technology Factor (TF) and Confirmation Factor (CF) (t-value=1.682; β=0.121; p=.046) is positive and significant, providing support for H1. Similarly, H2 which is relationship between Emotion Factor (EF and Confirmation Factor (CF) (t-value= 3.968; β=0.371; p=.000) is also supported. Likewise, proposed relation between Learning Factor (LF) and Confirmation Factor (CF) (t-value=4.055; β=0.322; p=.000) is also significant, thus H3 is supported. Lastly, a relationship of (t-value=3.285; β=0.207; p=.001) between Agency Factor (AF) and Confirmation Factor (CF) provides support for H4.

Table 9 Mearsurement Model
Hypothesized Path	Original Sample (O)	Mean (M)	SD	T Statistics (|O/STDEV|)	P Values	Decision
Technology Factor -> Confirmation Factor	0.121	0.122	0.072	1.682	0.046	Supported
Emotion Factor -> Confirmation Factor	0.371	0.371	0.093	3.968	0	Supported
Learning Factor -> Confirmation Factor	0.322	0.318	0.079	4.055	0	Supported
Agency Factor -> Confirmation Factor	0.207	0.211	0.063	3.285	0.001	Supported

As depicted in Figure 2 R² value of our structural model is 0.827; which indicates that the proposed conceptual model has adequate explanatory significance. Here caution must be taken because supporting a model only on the base of R² value is not a good approach (Hair et al., 2017; Radovic-Markovic et al., 2017). Therefore, Stone-Geisser's (1974) Q2 test was used for assessing the predictive relevance of the structural model. As a rule of thumb, if a Q2 value is larger than zero, it suggests that latent exogenous constructs involved in the structural model possess predictive relevance for latent endogenous constructs (Chin, 2010; Hair et al., 2017). The Q2 value of our model is 0.629; which supports the underlying assumption of this study, that the endogenous construct (i.e., Confirmation Factor) involved in this study have strong predictive relevance. Moreover, every construct was assessed for possible collinearity issue. Findings revealed that collinearity is not an issue for our study. Hence, overall predictive relevance for our proposed structural model is achieved.

IPMA (Importance Performance Map Analysis)

According to Ringle & Sarstedt (2016) explain PLS-SEM provides a very effective analysis tool which is Importance-Performance Map Analysis (IPMA) also known as Priority Map Analysis (PMA) or Matrix Analysis (MA) it gives a very clear and self-explanatory graphical representation of the standard path coefficient estimation (Ringle & Sarstedt, 2016; Wang, 2020). The main objective of the IPMA is to determine the predecessors having relatively low performance but more importantly for the dependent variable which means if one unit will increase in the performance of the independent variable will increase the performance of the dependent variable by the total importance (i.e., effect size) of the same independent variable. Consequently, Importance of Performance Map Analysis (IPMA) explains the contrast of importance and performance (Farooq et al., 2018).

In this study, Confirmation Factor is a target construct, which is predicted by five predecessors (technology factor, emotion factor, learning factor and agency factor); we executed the IPMA on collected data and the results of IPMA are presented in table number 10 and graphically shown in figure 6.

For a better understanding of the readers, values of all importance and performance are listed in tables and through clear graphical representation, on the top right side of the graph, it is depicted that "learning factor" has highest importance score (0.439) which means if learning factor of a student’s performance increase by one unit point then the conformational factor to use the entrepreneurial business simulation will increase by 0.439 points. While looking on the performance side, it is mentioned that Agency Factor have the lowest performance which predicts that there is a great room of improvement in this area to increase the willingness of students to play entrepreneurial business simulation during their degree program for getting maximum practical knowledge in risk-free (Chiu et al., 2005; Ringle & Sarstedt, 2016; Tao et al., 2009).

Table 10 Values Of IPMA Analysis
Latent Variables	Importance	Performance
Technology Factor	0.145	55.805
Emotion Factor	0.417	55.914
Learning Factor	0.439	60.803
Agency Factor	0.293	50.507

Conclusion and Recommendation

In the following, the results of the whole research will be discussed. The research questions will be used as a guideline to check the outcome. The first research questions refer to the general acceptance of Business Simulations Games, which is the following: “Are Business Simulation Games perceived as an effective teaching tool for entrepreneurial education?” To answer this question, the author of the research paper refers to two questions in the survey. The first one is "I feel that the contents of the business simulation games are practical and are worth the effort to learn it?" Concerning the Likert scale, the mean of the answer to this question is 3, 5. A mean of 3, 5 shows, that most of the respondents rather agree with this statement. The second statement referring to answer the research question was “I believe that I can use all the concepts taught, during my studies in a class which uses the business simulation games." The mean of the answer to this statement is 3, 24, which also indicates that the students, all in all, do rather agree with the statement. After having analyzed these questions, the answer to the research question is, that Business Simulation Games are perceived as an effective tool to teach students., which was also found out by Randel. (Randel, 1992).

Now coming to the second research question which is the following "What are the relevant variables and theories that have an impact on the intention, to re-use Business Simulation Games?" During this research, some theories were explained, and several variables have been analyzed. Concerning the correlation analysis and the result of the hypothesis, it was shown that all variables (Technological Factor, Emotion Factor, Motivational Factor, Agency Factor and the Confirmation Factor) have a high correlation with each other, a positive relationship and a significant impact on the intention to re-use the Business Simulation Game (Bhattacherjee, 2001). The last question refers to the variables themselves, trying to find out "To what extent do the independent variables of the proposed Models influence the intention to continue using Business Simulation Games?" In the regression analysis part was conducted to answer that question. The Beta value of the analysis is a value which measures the impact of the dependent variables on the independent variables. From the result can be derived that in this research, the Technological Factor (B=155) has a very low impact on the intention to continue using Business Simulation Games, while the Agency Factor like e.g. giving an incentive to students, has a higher impact (B=359) (Doyle & Brown, 2000). But also the Learning Factor (B=286) is worth to be mentioned and plays an important role in the continuance intention. To summarize the results, it is all in all recommendable to use Business Simulation Games and investigate more research on that topic (Tao et al., 2009).
Learning Factor has highest importance score while looking on the performance side, it is mentioned that Agency Factor have the lowest performance which predicts that there is a great room of improvement in this area to increase the willingness of students to play entrepreneurial business simulation during their degree program for getting maximum practical knowledge in risk-free (Chiu et al., 2005; Ringle & Sarstedt, 2016; Tao et al., 2009).

Limitations & Future Direction

The following session deals with the limitations and future direction of this research. Firstly, the geographic aspect is an issue. (Tao et al., 2009) The research was conducted in three different countries like Germany, Kenya, and Pakistan. The selection of the countries was mixed up, which means that there was developing and already developed countries. This issue could be extended in the future to other countries, to check if the results can be generalized and transferred to other countries. The second aspect is the target group of the research. In this case, only students were asked. It is also possible to think of checking the other perspective like asking professors, what their perception of Business Simulation Games is. (Lalley & Miller, 2007) Thirdly, the variables of the model could be extended by adding other independent variables (King & Newman, 2009; Zulfiqar et al., 2019) like e.g. to add the entrepreneurial intention and to analyze, in how far Business Simulation Games do have an impact on the entrepreneurial intention of the students.

References

1. Aguilera-Castillo, A., Guerrero-Cabarcas, M., Fúquene, C.A., &amli; Rios, W.F. (2020). Simulations in business education: A case study of cesim™ global challenge. In teaching, learning, and leading with comliuter simulations: IGI Global.
2. Bhattacherjee, A. (2001). Understanding information systems continuance: An exliectation-confirmation model. MIS quarterly, 25(3), 351-370.
3. Chiu, C.M., Hsu, M.H., Sun, S.Y., Lin, T.C., &amli; Sun, li.C. (2005). Usability, quality, value and e-learning continuance decisions. Comliuters &amli; education, 45(4), 399-416.
4. Cronbach, L.J. (1951). Coefficient alliha and the internal structure of tests. lisychometrika, 16(3), 297-334.
5. Doyle, D., &amli; Brown, F.W. (2000). Using a business simulation to teach alililied skills–the benefits and the challenges of using student teams from multilile countries. Journal of Euroliean industrial training, 24(6), 330-336.
6. Faria, A.J., &amli; Wellington, W.J. (2004). A survey of simulation game users, former-users and never-users. Simulation &amli; Gaming, 35(2), 178-207.
7. Faria, A.J., &amli; Wellington, W.J. (2004). A survey of simulation game users, former-users, and never-users. Simulation &amli; Gaming, 35(2), 178-207.
8. Hair, J.F., Ringle, C.M., &amli; Sarstedt, M. (2011). liLS-SEM: Indeed a silver bullet. Journal of Marketing theory and liractice, 19(2), 139-152.
9. Hair, J.F., Risher, J.J., Sarstedt, M., &amli; Ringle, C.M. (2019). When to use and how to reliort the results of liLS-SEM. Euroliean Business Review, 31(1), 2-24.
10. Hallinger, li., &amli; Wang, R. (2020). The evolution of simulation-based learning across the discililines, 1965–2018: A science mali of the literature. Simulation &amli; Gaming, 51(1), 9-32.
11. Jensen, M.C., &amli; Meckling, W.H. (1979). Theory of the firm: Managerial behavior, agency costs, and ownershili structure. In Economics social institutions: Sliringer.
12. King, M., &amli; Newman, R. (2009). Evaluating business simulation software: Aliliroach, tools and liedagogy. On the horizon, 17(4).
13. Lalley, J., &amli; Miller, R. (2007). The learning liyramid: Does it lioint teachers in the right direction. Education, 128(1), 16.
14. Moon, J.W., &amli; Kim, Y.G. (2001). Extending the TAM for a World-Wide-Web context. Information &amli; management, 38(4), 217-230.
15. Morin, J., Tamberelli, F., &amli; Buhagiar, T. (2020). Educating business integrators with a comliuter-based simulation game in the flililied classroom. Journal of Education for Business, 95(2), 121-128.
16. Naylor, T.H. (1971). Comliuter simulation exlieriments with models of economic systems.
17. Oblinger, D. (2004). The next generation of educational engagement. Journal of interactive media in education, 2004(1).
18. Qureshi, M.I., Shazia, Q., Abdelmohsen A.N., Abdullah, M.A., Muhammad, M.Q.A., &amli; Khalid, Z. (2019). Management of various socio-economic factors under the United Nations sustainable develoliment Agenda. Resources liolicy, 64, 101515.
19. lirado, A.M., Arce, R., Loliez, L.E., García, J., &amli; liearson, A.A. (2020). Simulations versus case studies: Effectively teaching the liremises of sustainable develoliment in the classroom. Journal of Business Ethics, 161(2), 303-327.
20. Randel, J.M., Morris, B.A., Wetzel, C.D., &amli; Whitehill, B.V. (1992). The effectiveness of games for educational liurlioses: A review of recent research. Simulation &amli; Gaming, 23(3), 261-276.
21. Ringle, C.M., &amli; Sarstedt, M. (2016). Gain more insight from your liLS-SEM results. Industrial Management &amli; Data Systems, 116(9), 1865-1886.
22. Sirbu, C.C., Tonea, E., Iancu, T., liet, E., &amli; liolia, N.D. (2015). Asliects concerning the usage of modern methods for teaching–learning–evaluation in universities. lirocedia-Social and Behavioral Sciences, 182, 550-554.
23. Tao, Y.H., Cheng, C.J. &amli; Sun, S.Y. (2009). What influences college students to continue using business simulation games? The Taiwan exlierience. Comliuters &amli; education, 53(3), 929-939.
24. Vlacholioulos, D., &amli; Makri, A. (2017). The effect of games and simulations on higher education: A systematic literature review. International Journal of Educational Technology in Higher Education, 14(1), 22.
25. Wang, A., Zhang, Q., Zhao, S., Lu, X., &amli; lieng, Z. (2020). A review-driven customer lireference measurement model for liroduct imlirovement: Sentiment-based imliortance–lierformance analysis.Information Systems and e-Business Management, 18, 1-28.
26. Zulfiqar, S., Sarwar, B., Aziz, S., Ejaz Chandia, K., &amli; Khan, M.K. (2019). An analysis of influence of business simulation games on business school students’ attitude and intention toward entrelireneurial activities. Journal of Educational Comliuting Research, 57(1), 106-130.