Technology Acceptance Model

Get Complete Project Material File(s) Now! »

Findings and Analysis

The questionnaire used in the survey was developed based on the technology acceptance model (Davis, F. D. 1989) by adapting the questions from the original technology acceptance model questionnaire (Davis, F. D. 1989) to represent the variables of interest. In the following chapter, the results will be presented individually for each onboarding process, where each result is an average of the questions representing the variables investigated. The questions for each variable can be found in appendix C. The results were analyzed using a linear regression analysis and a multiple regression analysis to investigate each onboarding process effect on the factors of the technology acceptance model. Since the study was conducted within-subjects, a repeated measures ANOVA, analysis of variance, was executed to investigate if there was a difference between the onboarding processes. A total of 68 responses were collected in the study and any demographic information collected was voluntarily provided by the participants. The full raw data of the study can be found in appendix A and the analysis can be found in appendix D.

Demographics

The demographic information was collected through questions which were voluntary for participants to answer. Of the 68 participants in the study, 67 answered the question regarding their age, and all 68 participants answered the remaining demographic questions. The participants of the conducted study were made up of 56% male and 44% female, as visualized in figure 4.1. The participants were between the age range 19-57 years of age, where 19-26 represented the majority of participants, see figure 4.2. The participants travel habits as seen in figure 4.3 showed that 62% of the participants traveled several times a year, 24% traveled several times a month and less than 15% traveled less often. More than half of the participants sometimes or always booked their travels through a travel application as seen in figure 4.4. However, a large amount of 46% never booked their travels through a travel application.

Onboarding Process 1

According to the relations within the technology acceptance model, a multiple regression analysis was used to test the effects of perceived usefulness and perceived ease of use on attitude toward using for onboarding process 1. The results of the regression indicated that perceived usefulness and perceived ease of use explained 81% of the variance found in attitude (R²​=.81, F(2,66)=140.25, p<.001). It was found that perceived usefulness had a positive effect on attitude (β = .836, p < .001), meaning that attitude increases as perceived usefulness increases. Additionally, the results showed no effect of perceived ease of use attitude (β = .005, NS), meaning that the perceived ease of use had no influence on the attitude that participants had towards onboarding process 1 as seen in figure 4.5. To evaluate the remaining factors a linear regression analysis was performed. The results of the regression showed that perceived ease of use explained 38% of the variance found in perceived usefulness (R²​=.38, F(2,66)=10.77, p<.01) as well as significant effects of perceived ease of use on perceived usefulness (β = .421, p < .01). Furthermore, attitude explained 74% of the variance found in intention to use (R²​=.74, F(2,66)=84.34, p<.001) and attitude showed a positive effect on intention to use (β = .813, p < .001). Positive effect was also shown of perceived usefulness on intention to use (β .680, p < .001) and perceived usefulness explained 67% of the variance found in intention to use (R²​ =.67, F(2,66)=53.04, p<.001).

Onboarding Process 2

The multiple regression analysis was repeated for onboarding process 2 to test the effects of perceived usefulness and perceived ease of use on attitude. The results of the regression indicated that perceived usefulness and perceived ease of use explained 77% of the variance found in attitude (R²​=.81, F(2,66)=140.25, p<.001). The results showed a significant positive effect of perceived usefulness on attitude, meaning that the higher the perceived usefulness the higher the attitude (β = .836, p < .001). Furthermore, there was no effect of perceived ease of use on attitude (β = .073, NS), implicating that perceived ease of use had no impact on the attitude of participants towards onboarding process 2, as shown in figure 4.6. To investigate the remaining factors a linear regression analysis was run for onboarding process 2. The regression showed that perceived ease of use explains 36% of the variance in perceived usefulness (R²​=.36, F(2,66)=9.59, p<.01). There was a significant positive effect of perceived ease of use on perceived usefulness (β = .614, p < .01), meaning that a higher perceived ease of use contributes to a higher perceived usefulness. The results of the regression further indicated that attitude explains 75% of the variance in intention to use (R²​=.75, F(2,66)=84.34, p<.001) and there was a significant positive effect of attitude on Intention to use (β = .900, p < .001). Furthermore it was found that perceived usefulness explained 43% of the variance in intention to use (R²​=.43, F(2,66)=84.34, p<.001), which may overlap with the variance explained by attitude.​Significant positive effect of perceived usefulness on Intention to use (β = .765, p < .001) was found.

Comparison of Processes

A repeated measures ANOVA was conducted to investigate if there was a difference between the two onboarding processes. The results indicated that there were no significant effects for perceived usefulness (F(1, 76) = 2.711, NS), attitude (F(1, 76) = .227, NS) and intention to use (F(1, 76) = .226, NS). However, there was a significant difference for perceived ease of use (F(1, 76) = 3.341, p=.05) for which assumption of sphericity was violated and Greenhouse-Geisser correction was applied to correct the degrees of freedom of the F-distribution. Further investigation of the significant effect was done using a pairwise comparison which showed that perceived ease of use of onboarding process 1 was reduced by -0.25 points compared to onboarding process 2 (p= 0.05). This means that the perceived ease of use of onboarding process 1 was lower than for onboarding process 2. Furthermore, in common for both onboarding processes was that there was no significant effect of perceived ease of use on attitude, meaning that the perceived ease of use did not affect the attitude towards none of the processes. Thereby the lower perceived ease of use of onboarding process 1 was not significant. Ultimately the onboarding processes had an effect on all the factors of the technology acceptance model, however there was no significant difference between the two onboarding processes that indicated that one was better than the other.

Discussion and Conclusions

Discussion of Method

The purpose of the method was to investigate onboarding processes in mobile application and their effect on user attitude towards continued use of applications, as well as the difference between onboarding processes in mobile applications effect on user attitude towards continued use of applications. The method selected was based on the technology acceptance model (Davis, F. D. 1989). The model has been widely used in research on information technology and proven as a valid instrument to test technology acceptance and user behavior (Venkatesh, V. 2000). To test onboarding processes in mobile applications effect on the factors of the technology acceptance model, prototypes with user onboarding patterns were used. The prototypes used in the study represented real onboarding processes using onboarding patterns tailored to the application of choice. The use of prototypes over existing mobile applications with onboarding processes allowed for full control over the patterns, flow and content of the processes and eliminated external factors that could have influenced the results. The external factors eliminated included the need for participants to download existing applications on their device, and the inability to control that the participants would fully complete the onboarding without skipping the process. Other factors included the risk of participants having already used the application to be tested, as well as already having gone through the provided onboarding process. Finally, the inability to control the content of existing onboarding processes, as well as updates by those providing the application during the execution of the study was avoided. However, as it was apparent to the participants that the study was conducted with prototypes, it may have been a factor of influence on the experimental procedures. The testing of the prototypes yielded very similar results, which did not indicate any difference between the onboarding processes on the outcome of intended use. The onboarding processes were very similar in both flow and content, and it could be assumed that if the processes had larger differences, there would have been a more significant difference in the results. However, it is not certain whether this would have provided any meaningful insights rather than just generated larger differences in the result. Creating larger differences in the onboarding processes would also interfere with the representation of real onboarding processes. The study consisted of a survey which was divided into a series of steps, guiding the participant to first interact with the prototype in question and continue by filling out the corresponding questionnaire. The study was conducted within-subjects, meaning that participants tested both conditions in form of testing both of the prototypes. In order to counteract any impact of this, the order in which the prototypes were shown to the participant was randomized. The study was conducted online, which implied that participant interaction with the prototypes, as well as the context in which the prototypes were viewed, could not be controlled. It could not be ensured that an interaction took place or that the interaction was completed, meaning that the results rely on the sincerity of those partaking in the study. The questionnaire used in the study was developed based on the original technology acceptance model questionnaire and validated instruments for the technology acceptance model to ensure validity through construct representation, meaning that the questionnaire provided valid measurements for each construct. The questions were adapted to the purpose of the study to measure the variables of interest. In the original technology acceptance model questionnaire, the constructs are rated on a scale of 1-7 providing quantitative results. However, the scale was considered too large for the study in question, and a scale of 1-5 was used. To provide reliable results the survey contained control questions which allowed for removal of invalid responses. The study was conducted as a quantitative study since it aimed to investigate the effects of onboarding processes in mobile applications rather than to gain insight into the opinions of participants on the onboarding processes. The study allowed for testing of the specific hypothesis that onboarding processes in mobile applications affected attitude towards continued use of applications. The quantitative method generated results which provided statistics that could be used to draw generalizations on a population and could successfully fulfill the purpose of investigating the research questions.​Distributing the study online allowed for a broader reach, and a total of 68 respondents participated in the study. The sample size of 68 participants reflects a population of 80, with a confidence level of 95% and a margin of error of 5%. With a larger population the confidence level is reduced and as such it is recommended that further research could repeat the study on a larger sample size to ensure reliability on a larger scale. Due to limited resources and time constraints, the participants were reached through social media channels, which may have affected the demographic representation in the study. The age representation in the study was wide and covered ages ranging from 19-57 years of age. However, the age range 19-26 represented the majority of participants which may have introduced a bias. For future research it would be encouraged to test a wider age range to investigate whether different age groups would have an impact on the results. The study showed that only half of the participants sometimes or always used a travel application when booking their travels. The participants that were already familiar with using a travel application may have established a certain level of confidence, meaning that they may have found the introduction useful but had enough prior knowledge to figure out the functions and value of the application. As a result, these participants may not have experienced one onboarding screen to be significantly better than the other which may have influenced the results. Therefore, the selection of a mobile travel application may have had an influence on the results and may be subject of further investigation as only half of the participants represented this category of users. Further research may also investigate other demographic connections to the effects of onboarding.

Discussion of Findings

The theoretical background study showed that onboarding processes were argued to be beneficial for the first-time impression and are often implemented in mobile applications to help users find value of a product quickly. Since user onboarding is the first interaction a user has with a mobile application, it is worth investigating whether onboarding can affect user attitude towards continued use by enhancing this first-time interaction. In order to investigate this problem, this thesis has examined the effect of​ different onboarding processes in mobile applications on​ users attitude towards continued use of applications. By using common onboarding patterns to investigate this problem, it could determine the attitude of users on the onboarding methods used by developers today. In this chapter, the findings of the study will be discussed in order to answer the research questions on mobile application onboarding processes and their effect on user attitude towards continued use of applications.

1 Introduction
1.1 Background
1.2 Purpose and Research Questions
1.3 Delimitations
1.4 Limitations
1.5 Outline
2 Theoretical Background 
2.1 Mobile Devices
2.2 User Onboarding
2.3 Value Proposition
2.4 Customer Retention
2.5 Technology Acceptance Model
3 Method and Implementation 
3.1 Method
3.2 Implementation
4 Findings and Analysis 
4.1 Demographics
4.2 Onboarding Process 1
4.3 Onboarding Process 2
4.4 Comparison of Processes
5 Discussion and Conclusions
5.1 Discussion of Method
5.2 Discussion of Findings
6 Appendices
Appendix A Survey Results
Appendix B Prototypes
Appendix C Questionnaire
Appendix D Data Analysis

GET THE COMPLETE PROJECT
Mobile application onboarding processes effect on user attitude towards continued use of applications