The influence of motivation on dropout- and completion rates in distance education

Bachelor Thesis, 2018
51 Pages, Grade: 1.70


Table of Contents

List of tables

List of figures

List of abbreviations



2. Theoretical background
2.1 Dropout and completion rates
2.1.1 The ,student integration model’
2.1.2 Composite Persistence Model
2.2 Self-efficacy
2.2.1 The social cognitive theory and perceived self-efficacy
2.2.2 Contextuality of perceived self-efficacy
2.3 Defintion of distance and e-Learning
2.4 Issues and hypotheses

3. Method
3.1 Search terms and databases
3.2 selection pattern

4. Results
4.1 Group-comparing studies
4.2 Regressive Studies
4.3 Summary of the results

5. Discussion
5.1 Limitations
5.2 Resume and implications

6. References

List of tables

Table 1. Used search terms

Table 2. Result of literature research: group-comparing studies.

Table 3. Result of literature research: regressive studies.

Table 4. Results of group-comparing studies

Table 5. Results of regressive studies

List of figures

Figure 1. Student Integration Model; Tinto (1975), p

Figure 2. Composite Persistence Model; Rovai (2003), S

Figure 3. Difference between efficacy and outcome expectations; Bandura (1977), p

List of abbreviations

Abbildung in dieser Leseprobe nicht enthalten


Distance learning has changed massively over the last decade because of establishing new technical tools. Although the e-learning market is rapidly growing, completion rates remain lower than in traditional courses. To enhance retention rates many scientists think of improving motivation of the participants as a good approach. Following this recommendation, this paper aims to analyze quantitaive empirical studies from the last 15 years to work out a structural relationship of the construct of perceived self-efficacy and leaving behavior in e-Learning. Inconsistency of definitions were found regarding dropout, self- efficacy and e-Learning. Results suggest that there are direct and indirect effects of self- efficacy on completion rates. Further it was indicated, that there are interaction effects with other variables.

Keywords: Se lf-Efficacy, e-Learning, dropout, Academic Persistence


The Information age has made higher education more necessary than ever. Life- long learning processes and multiple further trainings and qualifications are very im- portant for modern global economy. Heavy loads of papers restricted flexibility of dis- tance education pre-millenial, but now technological inventions getting used in daily life has brought us a never seen before variety of possibilities for designing education pro- cesses. Internet and therefore mobility allow quick communication and enable the possi- bility of instructional designs more attractive than just interchanging via text messages. For this the combination with nearly no limitations in areal and chronological organiza- tion has made the market for digital education offers, commonly known as ‘e-Learning’, rapidly grow. Stratistics MRC (2017) reported a market value of over 165 billion US- Dollar in 2015. Although companies invest an enormous amount of money and time on improving the quality with the experience from the last 20 years, the completion rates remain low. Nistor and Neubauer (2010) for example found 23.9% of the participants in online courses at the Virtuelle Hochschule Bayern (VHB) not completing them. Park and Choi (2009) report even higher dropout rates of 54% on universities in the Midwest of the United States. Friðriksdóttir and Arnbjörnsdóttir (2017) attest completion rates under 5% for free online courses. Hart, Friedman and Hill (2018) brought up evidence for 6.8% to 8.9% lower success rates in online courses compared to traditional ones.

Former research identified different influencing factors for this phenome. The in- dividual character of adult education indicates a need for looking closer at psychological attributes as important factors. Parker (2003) for example states the influence of motiva- tional variables. Yukselturk, Ozekes and Turel (2014) brought evidence for perceived self-efficacy being a predictor for dropout-behaviour. In the following, this paper intends to work out the impact of perceived self-efficacy on dropout- and completion rates in e- Learning courses.

2. Theoretical background

2.1 Dropout and completion rates

Scientific research about dropout has the extensive problem of not being able providing a general thus a domain-specific definition. This leads to a difficult situation, keeping in mind the multidimensionality according to Grau-Valldosera and Minguillón (2014) that indicates a cross-linking of interaction effects of different influence factors. Tinto (1975) already called this situation one of the two big problems of previous research in this domain. He heavily criticized the absence of grading the various characteristics of leaving behavior. For this, the definition varies from leaving a course before completion to reaching a stated performance goal in graduation or sometimes the forced leaving of an institution and everything in between. The possibility that other factors than leaving the whole program can lead to absence from a course has to be considered, too, but often is not. The differences in research context complicate a clear differentiation and can increase the risk for producing varying results in similar researches. Lee and Choi (2011) for example found strong discrepancy in impact extent of demographical factors, what they ascribe to the definition problems. Thinking further there is the aspect of identifying leaving behavior to discuss. Schmidt (2011) suggests that a person leaving the research context voluntarily may not inform the researchers about this, especially in distance education settings with rare personal contact and commitment. On the other hand classifying students outrunning the standard period of study or having longer times of absence automatically as dropout is questionable. Grau- Valldosera and Minguillón (2014) state to keep in mind factors such as sickness or high stress level, especially when studying besides a full-time job, are maybe responsible for absence what does not make a graduation in the future impossible. For this reason, a highly common way is to look at graduation rates because of the better possibility to determine a dropout right. Other popular indicators are persistence, willing-to-complete and intent-to-persist. Until now there is no general definition established, what is still told to be a complicating circumstance for example by Levy (2007), but also by public institutions like OECD (2012). For this reason many models with different perspectives were developed. Sociologists, psychologists and criminologists tried to describe the process of dropout in various contexts. At this there can often be recognized a fluent passage between perspectives. Some of the theoretical work have a closer look to non- normative behavior, like Battin-Pearson et al. (2000), some other like Newmann (1992) to learner’s engagement and even some to institutional variables like Rumberger and Palardy (2004). The following paragraph describes two models that seem to be relevant for this paper because of its pedagogical perspective. First one is the ,student integration model’ by Vincent Tinto (1975), that is even now very popular in this research field, for example in research by Levy (2007) or Nistor and Neubauer (2010). This model was the base for many further developments, like the ,student attrition model’ by Bean and Metzner (1985). The second one is the ,composite persistence model’ by Rovai (2003), which was especially designed in terms of online education.

2.1.1 The ,student integration model’

The before mentioned claim for categorizing different grades of persistence or dropout behavior by Tinto (1975) could be only satisfied adequately by creating a longitudinal model for describing the dropout process. Tinto (1975) itself created one. The purpose was explaining processes in interaction between an individual person and an institution and point out their contribution to leaving behavior. He had the idea of trying to connect different manifestations of dropout behavior with the variations of process that lead to them. The result is no general definition, but a grid to look differentiated at various leaving behaviors, what can help understanding the inconsistent results of former research. His foundation for this institutional oriented model was Durkheim’s (1961) ,Theory of suicide’ and an adaption of cost-benefit-analysis from economic sciences.

Durkheim (1961) states that the decision to commit suicide is a result of poor integration in the social system. According to him, most important for integration are congruent ethical values and sense of community. Transferring this to dropout problematics Tinto (1975) was arguing with college social system having its very own ethical and social structures. If individual ethical values show strong discrepancy from those socially agreed would therefore decrease the interaction with other members and in the end leaving the system. This effect is strengthened by the missing commitment to the system, which arises from poor integration. For this the risk of looking for and taking alternatives, which leads to leaving the institution, increases. Tinto (1975) pointed out the specialty of educational institutions. To the social system of the respective institution, an academic system is inevitably added caused by nature of the educational sector. These are considered different, however communicating and interacting with each other and therefore have an impact on the decision of staying or leaving by their interaction. This characteristic offers multiple constellations. Poor normative and structural integration in the academic system could add to well integration to the social system of college as well as inversely. Therefore, a poor or well integration in both systems is possible. The role of these considerations gets clear in thinking about the impact of voluntary or forced leaving and the structural difference between. Forced leaving is commonly caused by normative non-matching with the academic system. Mostly grades are too low and lead to disqualification. A voluntary leaver may also suffer from bad grades and is afraid to fit the future expectations. A divergence with social standards, breaking social or academic rules, or a combination may also be reasons for leaving. The mentioned problem differentiating various dropout-behaviors and their impact factors is reflected by these considerations very well. This can reach a level of nearly perfect social integration combined with poor integration in the academic system or reversely. Durkheim (1961) stated the existence of a reciprocal functional relationship between integrations in the two systems. This follows the principle of forced disintegration in the academic system as the logical consequence from excessive social life at a certain point. Tinto (1975) criticized the model of Durkheim for its descriptive aim. He doubted the possibilities for prediction causes because of missing integration of individual psychological factors as impactors for the variance of individual dropout-behavior.

Tinto’s (1975) perspective is surely special because of inclusion of individuals and their special attributes. The goal of not only describing the process of dropout but also making the interaction of different variables and various dropout-behavior transparent lead him to look outside the educational institution system. The proceedings in this system isolated could not be determining because they do not observe the individual character of a person. So he tried to integrate factors from outside the system being assumed to influence the decision to stay or leave in his own model.

To identify relevant psychological factors contributing to educational persistence he transferred constitutions of cost-benefit-analysis to individual decision-making processes. The main idea was that every decision could summarized be the result of weighting of needed performance and expected outcome. He concludes that besides personal background, like socio-economic status and gender, motivational and personal attributes concerning education expectation play a significant role in dropout processes. The terminus ,individual educational goal commitment’ describes these thoughts. The commitment includes not only the level but also the intensity of commitment. For this, a person, expecting to reach a doctoral degree, would more likely complete a bachelor degree than a person expecting exactly this bachelor degree. Therefore, it seems to be logical, that manifested education expectations lead more commonly to success than nonspecific ones. This is what Tinto (1975) calls ,intensity’. To enroll in a study program because of lacking alternatives leads consequently to lower intensity than enrolling with the goal of a doctoral degree. Because an institution is needed for this, individual preferences could determine successful integration to the previously mentioned systems of the institution. For this reason Tinto (1975) states that also individual expectations about the institution, called ,individual’s institution commitment’, is a psychological factor. If one is enrolling in an university, which most family members graduated, this person is usually not likely to transfer voluntarily in another institution. Sometimes graduating in on a specific faculty can also be part of a long-term career plan, what may have the same effect mentioned above. Another factor could be the restriction of financial resources. The importance of this factors shows in their usability as predictors of the interaction of the individual with the education environment. These reflections are thought to remove the isolation of the process. All external influences and systems outside the institutional ones are manifesting in some forms on the commitments, following this logic circuit.

The excerpted factors were combined by Tinto (1975) to his ,student integration model’. He thinks of the dropout process as long-term interaction between individual and academic, and social system of the educational institution. The experiences taking place in this systems lead to continuous modification of goal commitment and institutional commitment by the person itself. This modification results in either persistence or some form of dropout behavior.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1. Student Integration Model; Tinto (1975), p.95.

As indicated in Figure 1 individual attributes, family and educational background of a person impact not only directly or indirectly the performance, but also the evolution of the excerpted above commitments. Goal and institutional commitment are reflections of positive and negative experiences and therefore are simultaneously possible predictors. The way to the decision of leaving a program proceeds through interaction of variables, the resulting commitments and the respective resulting behavior in the various systems, which interact with each other. The mentioned modification of commitments caused by level of reached integration to the respective systems is the final step. This process continues permanently and leads either to staying or to leaving the educational institution. The scheme points out Tinto’s (1975) thesis of individual integration to social and academic system holding the strongest relation to staying or leaving. He intends a direct impact of integration on goal and institutional commitment, even if all other variables stagnate. Carrying on these thoughts the interaction of them is a determinant for dropout, the manifestation of the behavior and at last the individual adaption of it. This leads to the presumption that both low goal commitment and low institutional commitment could cause dropout.

The manifestation of the behavior is also influenced by other factors such as characteristic of the institution. If someone’s educational expectations change negatively for any reason, this person would rather switch to an institution with adequate level of difficulty. Also, an internal switch is a possible option, for example if there is the opportunity to take a two-year degree program instead of the four-year program in the same institution. If the expectations got higher, the inverse situation can happen, of course. Surely, it has to be considered, that the level of institutional commitment has an impact on the decision of transferring. The key characteristics of this model is conclusively the integration of individual perception of reality, which explains various reactions on obviously equal situations. Summarizing Tinto (1975) believes that commitments are the result of individual cost-benefit weighing by a person, which reinforces the high influence of individual psychological attributes.

2.1.2 Composite Persistence Model

Tinto’s (1975) model addresses the typical environment of a traditional student. Because of the changed requirements by a shift of learners’ characteristics Rovai (2003) aimed to create a model, which especially addresses the environment of digital education. Bean and Metzner (1985) had already realized that Tinto’s (1975) model was not able to display the conditions of non-traditional studying adequately. They characterized a non- traditional student as being older than 24, living off-campus, studying part-time or a combination of these attributes. Furthermore, they enroll usually voluntarily, because there is no more legal duty for attendance. They concluded that influence of social structures of the institution is much lower than of external factors, compared to traditional students. In addition, they assumed diverging interests. Non-traditional learners would rather focus on academic instead of social offers. Various additional commitments produce structures, basically different from the ones a student immediately enrolling after high school has. Rovai (2003) argues there is necessity to integrate these parameters featured by adult education to his model. He had the idea of a synthesis between Tinto’s (1975) and Bean and Metzner’s (1985) models, expanded with the special needs of distance learners, including explicitly online learners. Rovais’s (2003) basic principle was the same as Tinto’s (1975). He claimed that persistence of learners would be achieved more likely if there were a good matching between student and institution. Analyzing various researches Rovai (2003) identified a range of influence factors.

The model states, depending on these reflections, demographical variables like age, gender, intellectual development and former academic performance to be predictors for the decision of participating in a distance-learning program. Some researches for example found an impact of former academic performance on the successful completion of online courses, what Rovai (2003) ascribes to expectations that are more realistic of course requirements and having practiced skills necessary for self-studying. Furthermore, existing computer-, time management, computer-based, interaction, read-write- and informational competencies provide more potential for successfully completing a course. The mentioned variables, regarding to Rovai (2003), not only have influence on initializing, but also shape the behavior and possibilities during the scheme. Poor computer skills can lead, for example, to increased amount of work, which may collide with other commitments, cause distress and decelerate social integration. At worst this leads to financial consequences because of the enormous amount of time needed. External influence factors during a program are financial circumstances, extent of employment, family commitment, support by external systems and life crisis, like divorce, sickness or loss of job. It has to be considered, that, unlike for traditional students, an education course is one of many commitments. For this, lacking commitment to academic and social system can lead quickly to dropout. Therefore, integration to these systems is very important, although it may have different characteristic. Besides the need for transparent and consistent structure of the courses and procedures, he states that identification with the institution and as a consequence thereof social integration are important internal factors. This includes easy access to support offers like library, information center or tutorials. The lack of personal contact in online courses makes it necessary for the participant to get quick access to appropriate contact points and seek for advice.

Non-transparent structures with undefined goals and elaborate contact support may be barriers lowering persistence. Furthermore, he claims compliance of teaching and learning strategies as an important factor. For this, it is necessary that learners are able to adapt adequate strategies. Especially under the circumstances of online learning and therefore high level of self-responsibility, focusing to learner autonomy and ability to self- regulated learning seems central. Regarding to Rovai (2003) it is recommended to support a broad range of learner characteristics, because studies provide contrary results in this topic. A non-matching of teaching and learning strategies could lead to higher amount of workload and difficulty, which could end up in dropout considering lacking amount of time. Another internal factor are the commitments from Tinto (1975), which were discussed above. Summed up the decision to persist is the result of the interactions between the variables, shown in Figure 2, at the stated moments.

At the end, Rovai (2003) also has to submit that establishing an easy formula is not possible because of the complex, manifold influences and not to underrate interaction effects between the factors. Monitoring critical variables and adjusting the institution to the special needs can help identifying at-risk students and decrease the dropout rate.

Abbildung in dieser Leseprobe nicht enthalten

Figure 2. Composite Persistence Model; Rovai (2003), S. 9.

2.2 Self-efficacy

Both Tinto (1975) and Rovai (2003) stated that individual characteristics have large influence on the process of dropout. Researchers, like Vallerand, Fortier and Guay (1997) and Parker (2003), found an impact of motivational variables on persistence. Schunk and Mullen (2012) for example claim learners’ engagement, defined as manifestation of motivational, affective and cognitive variables, to be a central element.

The variety of possible motivational variables makes it necessary to limit them for this paper. The introduced dropout models postulate the importance of integration to the existing systems. For this, one needs supporting skills such as social skills in a new environment or learning strategies to use in different settings. More important on the motivational section, thus, is to know the existence of them and how someone self- perceives them, because this will affect his/hers actions.

Therefore, this paper focusses on perceived self-efficacy. It was identified as relevant factor for dropout-determination from many researches, like Abdullah and Ward (2016). Their purpose was to extent the Technology Acceptance Model from Davis, Bagozzi and Warshaw (1989) to match the needs of e-Learning. Thereby they identified perceived self-efficacy as external factor, which influences the intention of using e- learning through indirect effects.

2.2.1 The social cognitive theory and perceived self-efficacy

The concept of perceived self-efficacy was founded by Canadian psychologist Albert Bandura. Although using motivational factors to explain behavior was no new idea, this concept, regarding to Zimmerman (2000), was different from similar ones by White (1959) or Maier and Seligman (1976).

Developing a new theory of learning, which is today very popular among experts in this domain, he tried to identify factors within the cognitive system of an individual, which influence the learning processes. Banduras (1986) social cognitive theory assumes that in a social context change of behavior is generated by imitating behavior of others. The basis hypothesis was an interaction of personal attributes, behavior and environment, which makes a human work. The key component is the cognitive manifestation of experiences leading to change. This was completely different from stimulus-response theories. This implicates the ability to plan and reflect behavior. Perceived self-efficacy, therefore, influences both behavior and the environment interacting with. Coincidently both environment and actions influence perceived self-efficacy again.

Abbildung in dieser Leseprobe nicht enthalten

Figure 3. Difference between efficacy and outcome expectations; Bandura (1977), p. 193.

As pictured in Figure 3, a person shows some sort of behavior expecting a certain outcome. Bandura (1977) takes a step back. He claims there is a variable with much impact on the decision which behavior to use, called efficacy expectations. Rotter (1966) already attested importance of this factor in learning processes. Assuming freedom of choice in facing or avoiding a critical situation, it seems to be logical, that preliminary considerations about the possible consequences are made.

Regarding to Bandura (1977) perceived self-efficacy, which is not genetically dispositioned but learnt, influences this process by providing information about the own skills. This includes a self-rating as well as the estimation of the scope of influence on the environment. If a person rates a situation as exceeding their own skills, a natural mechanism of avoiding them to protect itself is activated. The perception of a situation as manageable will on the other hand lead to active challenging the situation. This process respectively enhances itself. Being successful will reinforce challenging behavior. Having failures then again will enhance avoiding behavior. Summarized perceived self-efficacy can be seen as the product of all effects from experiences.

Bandura (1977) submits that perceived self-efficacy is not some information, which makes it possible for a person to classify itself generally. There may be a basic self- efficacy perception influencing one’s behavior. In most situations, the context is as much important as the subjective perception. Schunk (1995) assumes that information about the perceived self-efficacy includes besides the expectation of someone’s own skills also success experiences, perceived difficulty of the task, amount of effort and time, experienced support, similarity to models, reliability of convincers and the form and intensity of emotional symptoms. For this, one can have a definitely positive perceived self-efficacy in some situation and a negative one in another. The biological need for safety is some kind of explanation for Bandura (1977). Judging a situation properly considers not only the task but also other influencing and environmental factors. If situational components differ too much, a person can experience the same challenge one time as manageable, the other time as not. This happens amongst other things through categorizing components of the situation as secure or insecure environment. Insecure environments lead generally to low perceived self-efficacy. The influence starts on the initializing of a certain behavior to level of persistence and invested effort. If the person thinks, he/she can surely handle the situation, effort and persistence increase, Chemers, Hu and Garcia (2001) found out. Conversely the mechanism works regarding to Vuong, Brown-Welty and Tracz (2010) the same way.

Bandura (1977) explains this with three dimension of perceived self-efficacy. He assumes that deciding processes are influenced by magnitude, generality and strength. Magnitude describes the perceived level off difficulty of the situation, generality the difference between general and contextual perceived self-efficacy. Strength means that someone’s perceived self-efficacy is not to shake by failures until a certain level.

For a better understanding a differentiation of the experiences, that build efficacy beliefs, is necessary. The most direct and strongest source is the successful performance, regarding to Bandura (1977). It directly deposits in expectation of competencies and is mostly generalized. Especially for a general facing of stressing situations success experience is an important basis. Tinto (2013) compares this to Newton’s first law. If no force exists, self-efficacy stays the same. Success, in this depiction, is at the end the force to change the status, therefore positive performance reinforces itself. Schunk and Meece (2006) on the other hand submit that mistakes and failure are a part of our lifelong experiences. If someone has enough of the mentioned ,strength’, there will be no crucial impact by them. Kaplan, Peck and Kaplan (1995) state, if there is not enough ,strength’, failure can quite have negative effects. Regarding to Bandura (1977) vicarious experience can also be a source for rating. The underlying principle for this is social comparison, which regarding to Schunk and Meece (2006) works best, if the model is a person similar to the observer. Seeing success of subjective similar persons tempers fear and raises expectations for being successful itself. Because of the subjectivity of perceived self- efficacy Bandura (1977) assumes that verbal persuasion can also be a source for it. Making someone believe he can handle a challenge raises motivation, but the strength of this influence is very low. Even little mistakes can shake it. The last factor is emotional arousal. Excitement and stress leads for evolutionary reasons to changed performance, sometimes raised and sometimes even lowered, for example because of constant fear. For this reason, besides perceived controllability of a situation, thought control efficacy is important to have. Bandura (1997) states a suggestibility of negative thought until a certain point. If the person affected is not able to eliminate or relativize negative thoughts, distress can raise and influence efficacy beliefs.


Excerpt out of 51 pages


The influence of motivation on dropout- and completion rates in distance education
University of Regensburg
Catalog Number
ISBN (eBook)
Drop-Out, Academic Persistence, Self-efficay, e-Learning
Quote paper
Simon Dienst (Author), 2018, The influence of motivation on dropout- and completion rates in distance education, Munich, GRIN Verlag,


  • No comments yet.
Read the ebook
Title: The influence of motivation on dropout- and completion rates in distance education

Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free