人們眼中的創新事物：以Uber台灣為例 - 政大學術集成

全文

(1)國立政治大學國際傳播英語碩士學位學程 International Master’s Program in International Communication Studies College of Communication National Chengchi University. 碩士論文 Master’s Thesis 政治. 大. 立. ‧ 國. 學 ‧. 人們眼中的創新事物：以 Uber 台灣為例 Innovation as in the eyes of its beholders: A case study on Uber adoption in Taiwan. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. Student: Joy Chia-ho Wan Advisor: Dr. Cheng Kuo. 中華民國 108 年 7 月 July, 2019. DOI:10.6814/NCCU201900593.

(2) Acknowledgment. If there is anyone that I would want to owe it all to, it has to be God. I would never have the intelligence to write up even the tiniest part of this thesis in English if it was not God leading me and accompanying me all the way through. From major tasks like collecting data, which I gathered 337 responses. 政治大 producing tables, God had heard my pleas and helped me time and again. It is 立 in mere two weeks, to minor problems like getting rid of unwanted lines while. ‧ 國. 學. not a perfect research as there were insignificant results which met me by surprise, but God knows that I have come up with my best with His help. I now. sit. Nat. innovation adoption for further researches.. y. ‧. leave the rest to whoever wants to investigate people’s relationship with. io. al. er. In terms of people, I give my greatest thanks to Dr. Cheng Kuo, a. v i n C h when the time was believe that I can finish the thesis e n g c h i U running short. As an adult n. distinguished scholar in the academic field of communications who chose to. student, I enjoyed the freedom to do academic research as I was being supervised by Dr. Kuo. She would guide the directions and advise on major issues, leaving the rest for me to test and try. There were moments that I believed that doing research was fun. And the research and writing of this thesis had not been a torment as some graduate students had claimed. Last but not least, I would like to thank my mother, who took care of my children, feeding them and tending their daily needs as I spent time away racing to finish the degree paper. I also thank my husband for his tolerance when I. DOI:10.6814/NCCU201900593.

(3) was sticking to my computer writing and unable to do household chores. There were days when he did the laundry, but I did not manage to fold the clothes and pants, leaving them scattered around. The house was messy, and I am thankful that he was able to live with it. The accomplishment of one thing is the fruit of many helping hands. My thanks go to all the people who had lent that helping hand to make the thesis a mission that became possible.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. DOI:10.6814/NCCU201900593.

(4) Abstract. There were over 170K startup companies introducing innovative products and services worldwide as of April 2019. Not all of them can survive long enough to carve their names, and many failed. Statistics showed that over 70 percent of new companies failed because they ignored people in their product designs or marketing strategies. Therefore, the study aims to find out the. 治政 personality and perceived attributes of innovation, 大 one of the major concepts of 立 relationship between people and innovation adoption. Five Factor Model of. Diffusion of Innovation, were employed to examine if there are links between. ‧ 國. 學. the different personality traits and how people perceive the attributes of an. ‧. innovation. Using Uber as a case study, the research used mixed methods. y. Nat. combining quantitative and qualitative tools for data collection. The results. er. io. sit. concluded that an innovation’s attributes, including compatibility, trialability, and complexity, would influence innovation adoption. Also, different personality. al. n. v i n C h attributes differently. traits would perceive innovation e n g c h i U Finally, people’s need should be addressed if an innovative product or service seeks to survive in the market.. DOI:10.6814/NCCU201900593.

(5) 摘要. 截至 2019 年 4 月為止，超過 17 萬家的新創公司分散在全球各地，販售他們的創新產品與服務。其中許多還未能讓人所熟知，就關門大吉。數據顯示，新創公司活不下去，百分之 70 的原因在於忽略人的因素。該研究以 Uber 為個案，企圖找出人與創新採納之間的關聯。研究從五大人格特質（Five Factor. 立. 政治大. Model of personality）出發，檢視不同個性的人對一件創新事物是否有不同的. ‧ 國. 學. 觀感（perception），以及人格特質是否影響他們對於新事物的採用。研究以. ‧. 質、量化方式並行。分析後發現，創新事物的屬性如相容性、可試用性、複雜. y. Nat. n. al. er. io. sit. 性能影響新事物的採用與否。不同的人格則有可能使人們對創新事物的觀感不. i n U. v. 同。研究也發現人們的需求對於創新事物的採納似乎也具影響力。. Ch. engchi. DOI:10.6814/NCCU201900593.

(6) Table of Content. Chapter 1 Introduction .................................................................................. 3 1.1 Businesses fail because “people” are missing ........................................... 3 1.2 Uber as an innovation: its initial success and challenges ......................... 4 1.3 Research base and framework ................................................................. 7 Chapter 2 Literature Review ......................................................................... 9 2.1.1 Perceived attributes of innovation ........................................................... 9. 政治大 2.2.1 Personality traits ................................................................................... 12 立 2.1.2 PAI and innovation adoption ............................................................... 11. 2.2.2 Personality traits and innovation adoption ............................................ 14. ‧ 國. 學. 2.2.3 Personality traits and PAI ...................................................................... 16. ‧. 2.3 Demographics and innovation adoption: H25 & 26 .................................. 22. sit. y. Nat. Chapter 3 Methodology ............................................................................... 26. io. er. 3.1 Data collection and sample .................................................................... 26 3.2 Measures ................................................................................................. 27. al. n. v i n Ch 3.2.1 Personality traits ............................................................................ 27 engchi U 3.2.2 PAI ................................................................................................. 28. Chapter 4 Results ....................................................................................... 30 4.1 Demographic profile ............................................................................... 30 4.2 Measurement Validation .......................................................................... 32 4.3 Data analysis ........................................................................................... 34 4.3.1 PAI, personality traits, and Uber usage: H1—H8 ........................... 36 4.3.2 Personality traits and relative advantage: H9—H12 ...................... 38 4.3.3 Personality traits and compatibility: H13—H16 .............................. 39 1. DOI:10.6814/NCCU201900593.

(7) 4.3.4 Personality traits and trialability: H17—H20 ................................... 41 4.3.5 Personality traits and complexity: H21—H24 ................................. 42 4.3.6 Demographics and Uber usage: H25 & H26 .................................. 43 4.4 Qualitative results .................................................................................... 45 Chapter 5 Discussion .................................................................................. 47 5.1 Summary of findings ................................................................................ 47 5.1.1 PAI and innovation adoption........................................................... 48 5.1.2 Personality traits and innovation adoption ..................................... 48. 政治大 5.1.4 Demographics and Uber usage ..................................................... 51 立 5.1.3 Personality traits and PAI............................................................... 49. 5.2 Other findings........................................................................................... 52. ‧ 國. 學. 5.2.1 Hierarchical regression analysis results ......................................... 52. ‧. 5.2.2 Implications of Q29 ........................................................................ 55. sit. y. Nat. 5.2 Limitations and further research............................................................... 56. io. er. Chapter 6 conclusion .................................................................................. 60 6.1 Different personality traits tend to view innovations in different ways ...... 60. al. n. v i n C hcomplexity matter........................................ 6.2 Compatibility, trialability and 61 engchi U 6.3 To need, or not to need, that is the question ............................................ 62 References ................................................................................................... 63. 2. DOI:10.6814/NCCU201900593.

(8) Chapter 1 Introduction. According to PitchBook, a financial data and software company, there are currently over 170K new startup companies, backed by venture capitals, worldwide as of April 2019 (What you can research, 2019). All of them provide innovative products and services that stand a chance to change people’s way. 政治大. of life. Fundraising platform Kickstarter, online course provider Udemy, and. 立. ‧ 國. 學. sharing economy champions Uber and Airbnb are some of the well-known brands. However, startups face a high-level of uncertainty (Schmitt, Rosing,. ‧. Zhang, & Leatherbee, 2017), and their innovative products do have high rates. sit. y. Nat. io. n. al. er. of failure (Griffith, 2014).. Ch. engchi. i n U. v. 1.1 Businesses fail because “people” are missing. CB Insights, a US-based consulting firm, analyzed 101 startup companies on why their businesses failed. The results showed that 42 percent of the respondents said their products “have no market need,” 17 percent viewed their products as “user un-friendly,” while another 14 percent admitted that they “ignored customers” (The Top 20 Reasons, 2018). In conclusion, up to 73 percent of the reasons why a startup company fails can be attributed to people, 3. DOI:10.6814/NCCU201900593.

(9) be it consumers or clients. If people perceive a product in a positive way and choose to adopt it, there is a market. If not, there is no market. Therefore, to focus on product innovation itself without considering the people factor would not solve the problem of why an innovation does not spread well. And Diffusion of Innovations (DOI), as proposed by Rogers in 1962 (Rogers, 1983), has been commented as “heavily pro-innovation (Larsen, 2001).”. 立. 政治大. Based on the above observation, the thesis argues that to know why an. ‧ 國. 學. innovation is eventually adopted, one has to look into people’s differences, such. ‧. as personality traits, and whether they perceive an innovation in a favorable. Nat. io. sit. y. light. This research selects Uber as a case study because it is an internationally. n. al. er. well-known brand and has a presence in Taiwan, which makes it accessible in collecting local data.. Ch. engchi. i n U. v. 1.2 Uber as an innovation: its initial success and challenges Uber Technologies Inc. was established in 2009 in San Francisco, with the notion of sharing rides that connects independent drivers and customers (Min, So and Jeong, 2018). Uber, initially called UberCab (Wirtz &Tang, 2016), taps into a void in the city where street taxis are difficult to hail, fail to arrive on time, 4. DOI:10.6814/NCCU201900593.

(10) and are poorly maintained. But with Uber, all passengers need to do is download the APP, register, enter a credit card number, then all set. After that, passengers can get their Uber services just with a push of a button. Locations of the cabs can be tracked on GPS, so passengers know exactly where they are. People who take Uber cars reportedly have an enjoyable experience because it arrives on time and offers a quicker payment transfer (Ng, 2016). As. 政治大. of 2016, Uber had accumulated 40 million monthly active riders worldwide. 立. (Kokalitcheva, 2016).. ‧ 國. 學. Nevertheless, Uber was not short of problems at home in the US. Uber. ‧. drivers-- individuals that offer “shared rides” -- do not need to acquire taxi. Nat. io. sit. y. licenses, and this incurred hostility from local taxi drivers, who comply with. er. every government rule. The company then changed its name from UberCab to. al. n. v i n C h some of the regulations. Uber (Wirts et al., 2016) to bypass engchi U. Uber’s success led to its expansion overseas, including the UK, France, Germany, South Africa, India, China, and Taiwan. While growth can be seen in those markets, Uber is confronted with the same challenges that it faces in the US: antagonism from local taxi drivers. Protests have been seen in France, Germany, and India (Wirts et al., 2016).. 5. DOI:10.6814/NCCU201900593.

(11) In Taiwan, Uber’s most successful North Asian market (Li & Cheng, 2019), the same old problems occurred. To ease animosity, Uber Taiwan launched cooperation with local taxi companies and initiated a campaign called “Win Together #No one should lose” (Uber Taiwan, 2019), making it clear that Uber works hand-in-hand with Taiwan’s transportation industries. However, the strong opposition from local taxi drivers still caused the Taiwan government to. 政治大. propose amending Article No. 130-1 of the Transportation Management. 立. Regulations. The amendment, which had been dubbed the “Uber clause,”. ‧ 國. 學. required Uber to charge passengers by the hour, instead of by meters. Uber,. ‧. on the other hand, refused to accept the amendment, saying it deprived drivers. Nat. io. sit. y. and rental companies of business opportunities (Shan, 2019). The controversy. er. later drew reactions from the American Institute in Taiwan, the de facto. al. n. v i n embassy of the US to the C country, that Taiwan was “unfriendly for h e nsaying gchi U innovators (Lee, 2019).”. After rounds of negotiations, the government introduced a multipurpose taxi service program that would issue commercial licenses for non-traditional, ridehailing service drivers and encouraged Uber drivers to join in. A multipurpose taxi service is defined as one in which the driver is not legally required to use a yellow taxi. The fares are metered. However, passengers must contact the 6. DOI:10.6814/NCCU201900593.

(12) drivers through apps. At the time of this writing, over 800 Uber drivers had obtained their commercial licenses to drive non-traditional types of taxis (Hundreds in Taipei, 2019).. 1.3 Research base and framework Uber has provided a good many materials for research in the academic world.. 立. 政治大. Some scholars discussed Uber as an economic topic for its role in the sharing. ‧ 國. 學. economy. Others viewed it as a government regulation issue (Min et al., 2018). ‧. because just like Airbnb and Amazon Go, a cashless store introduced by. Nat. io. sit. y. Amazon, innovations often find themselves way ahead of government. er. regulations. And when there is a law, it does not necessarily work in favor of. al. n. v i n the innovation. Airbnb is nowC subject that regulates short-term rental in h e ntoga law chi U the US (Daniels, 2018) that could hamper its development, whereas a San Francisco official is seeking to ban Amazon Go from operating locally (Ioannou, 2019). Still other researchers recognized individuals as pivotal to Uber usage, not the innovation itself. Min et al., (2018) argued that however innovative such as Uber is, its adoption hinges on people and their perception towards it. They 7. DOI:10.6814/NCCU201900593.

(13) examined how respondents. perceive. relative. advantage,. complexity,. compatibility, and observability, four of the innovation characteristics proposed by Rogers (1995) through perceived ease of use (PEOU) and perceived usefulness (PU), the two major constructs of Technology Acceptance Model (Davis, 1986). The result is that PEOU and PU would affect how people see those attributes of innovations.. 政治大. Following this line of thought, the thesis proposes that people are the. 立. determinant on whether an innovation can survive, adoption-wise. This study. ‧ 國. 學. proposes the following questions:. ‧. RQ 1. How would the PAI of Uber influence its usage in Taiwan?. Nat. io. sit. y. RQ 2. How would personality traits affect their Uber usage?. n. al. er. RQ 3. How would personality traits affect people to perceive Uber’s attributes?. Ch. engchi. i n U. v. This study first examines the interplay among personality traits, perceived attributes of innovations (PAI), and innovation adoption, hypothesizing that personality alone can influence innovation adoption and PAI. Then the research discusses the methodology used to test the hypotheses, evaluates the results and points out limitations and further discussion concerning the topic.. 8. DOI:10.6814/NCCU201900593.

(14) Chapter 2 Literature Review. In investigating the personality factor in the innovation adoption as in the case of Uber usage Taiwan, the thesis employs the Five Factor Model (FFM) of personality (McCrae & Costa, 1999) and PAI, firstly introduced by Rogers in. 政治大. 1962 (Rogers, 1983). The purpose of this research is to provide new startups. 立. ‧ 國. 學. with an idea about what influence people’s decision to use an innovative product? Is this adoption behavior based on the perceived attributes of a new. ‧. product? Or is the behavior the result of potential consumers’ personality. sit. y. Nat. io. n. al. er. difference? By choosing Uber as a case on the innovation adoption in Taiwan,. i n U. v. the research seeks to find out the answers to those questions.. Ch. engchi. 2.1.1 Perceived attributes of innovation DOI has been one of the major theories in examining the spreading process of how an innovation finds its way from early adopters to the majority of people. Rogers (1995) suggested that PAI, or perceived characteristics of innovation that he had used interchangeably in his discussion of the diffusion theory, are important in explaining the diffusion of an innovation. 9. DOI:10.6814/NCCU201900593.

(15) In Diffusion of Innovations Fourth Edition (1995), Rogers maintained that the five PAI are relative advantage, compatibility, complexity, trialability, and observability, and each is associated with the rest four and yet are conceptually different. Relative advantage is the degree to which an innovation is perceived as being better than the idea it supersedes (p. 212). Compatibility is the degree to which an innovation is perceived as consistent with the existing values, past. 政治大. experiences, and needs of potential adopters. This includes compatible or. 立. incompatible (1) with sociocultural values and beliefs, (2) with previously. ‧ 國. 學. introduced ideas, or (3) with client needs for the innovation (p. 224). Complexity. ‧. is the degree to which an innovation is perceived as relatively difficult to. Nat. io. sit. y. understand and use (p. 242). Trialability is the degree to which an innovation. er. may be experimented with on a limited basis (p. 243). Observability is the. al. n. v i n degree to which the results ofCan U to others (p. 244). hinnovation e n g c harei visible. The classic five PAI have later been extended and combined with PEOU and PU, evolving into eight perceived characteristics of innovation that would include voluntariness, image, relative advantage, compatibility, ease of use, trialability, result demonstrability, and visibility (Moore & Benbasat,1991).. 10. DOI:10.6814/NCCU201900593.

(16) 2.1.2 PAI and innovation adoption Although the five classic PAI captured the essence of how innovation can be viewed. Not all five elements have an equal influence in terms of innovation adoption. In a research on the use of e-appointment scheduling (EAS) services in a primary health care clinic in Australia, it was found out that only perceived relative advantages, complexity, compatibility, and trialability were related to. 政治大. the adoption of the EAS services (Zhang, Yu, Yan, & Spil, 2015).. 立. While studying user adoption of Amazon Go, Kras (2018) identified relative. ‧ 國. 學. advantage, compatibility, and complexity as having direct relationships with. ‧. innovation adoption, concluding that compatibility and relative advantage were. Nat. sit. n. al. Van Slyke, Belanger,. er. io. adoption.. y. positively related to adoption whereas complexity was negatively related to. v i n C h Comunale (2004) and e n g c h i U also. recognized relative. advantage, compatibility, and complexity as the most relevant variables in terms of innovation adoption in the diffusion researches. Besides, Yi, Fiedler, and Park (2006) indicated that relative advantage, complexity, and compatibility are “the only innovation characteristics consistently related to innovation adoption and implementation.”. 11. DOI:10.6814/NCCU201900593.

(17) In addition to the three key variables concerning innovation adoptions, trialability is considered appropriate when non-adopters of innovation are considered (Min et al., 2018). Since the thesis aims to research both users and non-users of an innovation, trialability is added in as a variable as a result. Because observability has been broken into two concepts, its discussion would involve implications on result demonstrability and visibility, making the. 政治大. study on observability alone complicated. The thesis excludes observability for. 立. clarity reason.. ‧ 國. 學. Hence, the research focuses on relative advantage, compatibility,. ‧. complexity, and trialability, all of which are as perceived, and proposes that. Nat. io. sit. y. H1: The perceived relative advantage of Uber is positively related to its usage.. er. H2: The perceived compatibility of Uber is positively related to its usage.. al. n. v i n H3: The perceived trialability C of h Uber is positively e n g c h i Urelated to its usage.. H4: The perceived complexity of Uber is negatively related to its usage.. 2.2.1 Personality traits An innovation is an idea, practice, or object that is perceived to be new by an individual (Flight, D’Souza, & Allaway, 2011). In other words, the perception of innovations stems from the eyes and minds of their beholders. To develop 12. DOI:10.6814/NCCU201900593.

(18) an innovative product without investigating people’s mind would leave a crucial part of innovation adoption unattended. The FFM of personality has been known as offering a parsimonious taxonomy, while many other studies in applied psychology tend to provide exhaustive examinations that led to hundreds of personality traits (Barnett, Pearson, Pearson, & Kellermanns, 2015). The FFM labels neuroticism,. 政治大. extraversion, openness to experience, agreeableness, and conscientiousness. 立. as major personality traits of individuals. A brief view of the five personality traits. ‧. ‧ 國. 學. is as below:. y. Nat. Table 1. FFM personality traits & characteristics Extraversion. • • • •. Energeticg Risk-takinga Assertivee Confidentd. •. •. Altruistice Cooperativee Tend to truste Modeste. Conscientiousness. • • • •. Deliberateb Diligentg Perseveringc Achievement-orientedg. Neuroticism. • • • •. Moodyh Fearfula Anxiousb Stressedi. Agreeableness. Ch. er. n. al. engchi • •. sit. Characteristics. io. Personality traits. i n U. v. 13. DOI:10.6814/NCCU201900593.

(19) (Continued from Table 1) Openness to experience. • • • •. Imaginativeg Curiousg Willing to learnf Variety-seekinga. aAnic. (2007). bAslan & Cheung-Blunden (2012). cCiavarella, Buchholtz, Riordan, Gatewood, & Stokes (2003). dKolb & Griffith (2009). eLahti, Räikkönen, Lemola, Lahti, Heinonen, Kajantie, Pesonen, Osmond, Barker, & Eriksson (2013). fMatzler & Mueller (2011). gMcCrae et al. (1999). hMichikyan, Subrahmanyam, & Dennis (2014). iMoutafi, Furnham, & Tsaousis (2005).. 政治大. 2.2.2 Personality traits and innovation adoption. 立. Studies on early adopters suggest that personality traits might explain the. ‧ 國. 學. reason why some people adopt innovations earlier than others. For example,. than. later. adopters,. which. can. be. related. to. io. sit. rationality. y. Nat. greater. ‧. those who adopt early tend to have higher aspirations for education and have. er. conscientiousness and openness to experience (Cisternas-Godoy, 2016).. al. n. v i n C h direct relationships Barnett et al., (2015) tested e n g c h i U between the Big Five personality traits and technology use of a web-based classroom technological system and discovered that conscientiousness demonstrated a positive association with the actual use of technology, while agreeableness had “no relationship” with the actual use of technology. Their results on neuroticism showed that the more emotional people are, the less likely they are to adopt new technology. Another research on the adoption behavior of computer-based 14. DOI:10.6814/NCCU201900593.

(20) learning. identified. openness. to. experience,. conscientiousness,. and. extraversion as having strong significance over adoption (Khan, Iahad and Miskon, 2014). On the contrary to the three personality traits – openness to experience, conscientiousness and extraversion – that were often associated positively with innovation adoption, neuroticism showed a negative relationship with adoption.. 政治大. Aside from the research by Barnett et al. (2015), a study on e-book adoption. 立. discovered that people with the trait of neuroticism, which is linked with fearful. ‧ 國. 學. and anxious towards innovations, seemed to avoid using the electronic format. ‧. of publications. Rather, they were more comfortable with traditional printed. Nat. io. sit. y. books (Bansal, 2011). Another study on the adoption of Google glasses also. er. showed that neurotic people were more nervous and worrying than those. al. n. v i n C hstable. In general, whose emotions were relatively e n g c h i U neurotic individuals were less likely to adopt smart glasses (Rauschnabel, Brem, & Ivens, 2015).. Following this line of discussion, the thesis chooses to include conscientiousness, openness to experience, extraversion, and neuroticism as key constructs in measuring personality traits’ relationships with Uber usage. Hypotheses concerning personality and the usage are as below: H5: The more conscientious people are, the more they are going to use Uber. 15. DOI:10.6814/NCCU201900593.

(21) H6: The more open to experience people are, the more they are going to use Uber. H7: The more extraverted people are, the more they are going to use Uber. H8: The more neurotic people are, the less they are going to use Uber.. 2.2.3 Personality traits and PAI. 政治大. The thesis argues that people are the true determinant of innovation. 立. adoption, as opposed to the innovation-oriented diffusion theories, and seeks. ‧ 國. 學. to explore further whether personality differences would lead to divisive. ‧. perceptions of those PAI.. Nat. io. sit. y. Relative advantage. er. Relative advantage is the extent to which an innovation is perceived as. al. n. v i n C(Linton being better than its precursor 2013) and is referred to as the h e n&gWalsh, chi U benefit of adopting the new technology when compared to its cost (Mohr, Sengupta, & Slater, 2009). It takes diligent efforts and deliberation to compare and decide whether an innovation is indeed better than the previous ones. Therefore, the thesis predicts that. 16. DOI:10.6814/NCCU201900593.

(22) H9: The more conscientious people are, the more they find out the relative advantage of Uber. People who are open to experience tend to seek variety and are intellectually curious. They are willing to think about new ideas and unconventional values. Open people also tend to hold a positive attitude towards learning new things (Matzler et al., 2011). Relative advantage is gained. 政治大. through the comparison between at least two products, whereas variety-. 立. seeking means to prefer having more than one choice. Therefore, it can be. ‧ 國. 學. predicted that it takes less effort for variety-seeking people to find out and. ‧. distinguish a product’s relative advantage among an array of competing ones.. Nat. io. sit. y. The thesis suggests that. n. al. er. H10: The more open to experience people are, the more they find out the relative advantage of Uber.. Ch. engchi. i n U. v. Extraversion is linked to being active and risk-taking (Anic, 2007). Extraverts work quickly, tend to lack patience, and have relatively short attention spans (Lynch & Chernatony, 2007). Relative advantage is usually found in an innovation that improves the efficiency of work or reduces the operational cost (Chen & Zhang, 2016). Therefore, it can be inferred that extraverts tend to recognize the relative advantage more easily than other 17. DOI:10.6814/NCCU201900593.

(23) people. Their lack-of-patience nature will propel them to constantly on the lookout for more efficient, cost- and time-saving options. The research hypothesizes that H11: The more extraverted people are, the more they find out the relative advantage of Uber. Neuroticism, anxiety, and fearfulness have been linked together (Panitz,. 政治大. Sperl, Hennig, Klucken, Hermann, & Mueller, 2018) in the context of innovation. 立. literature. Previous studies mentioned in section 2.2.2 have indicated a. ‧ 國. 學. negative link between neuroticism and innovation adoption. Hence, the. ‧. research predicts a negative relationship between neuroticism and of relative. Nat. io. sit. y. advantage. The thesis hypothesizes that. n. al. er. H12: The more neurotic people are, the less they find out the relative advantage of Uber.. Ch. engchi. i n U. v. Compatibility Compatibility has been defined as closely linked to relative advantage in innovation studies. While relative advantage is the incremental benefit to be gained by using one innovation over its alternatives, compatibility is the extent to which an innovation is compatible with the user's prior experiences (Davis, 2003). 18. DOI:10.6814/NCCU201900593.

(24) Because compatibility is a closely-knit construct with relative advantage, the thesis proposes that relationships between the four personality traits and relative advantage might also exist in the case of compatibility. The study hypothesizes that H13: The more conscientious people are, the more they perceive Uber as compatible with existing products.. 政治大. H14: The more open to experience people are, the more they perceive Uber as. 立. compatible with existing products.. ‧ 國. 學. H15: The more extraverted people are, the more they perceive Uber as. ‧. compatible with existing products.. Nat. sit. n. al. er. io. with existing products.. y. H16: The more neurotic people are, the less they perceive Uber as compatible. Trialability. Ch. engchi. i n U. v. Trialability is whether an innovation offers the possibility to be trialed and experimented before adoption (Sanson-Fisher, 2004). In the context of Uber usage, which combines APP and riding experience, conscientious consumers are the group of people who are more likely to go to the extra mile to install the APP, register an account and then get experimental rides. The diligent nature of conscientious people might help them perceive Uber as more trialable than 19. DOI:10.6814/NCCU201900593.

(25) those who don’t bother to research the product at all. Hence, the thesis suggests that H17: The more conscientious people are, the more they perceive Uber as trialable. Trialability involves experiment with new things. Openness to experience has curiosity and broad-minded as two of its major characteristics (Ciavarella. 政治大. et al., 2003). People who are curious and broad-minded are more open to. 立. experimenting with innovations. The thesis argues that. ‧ 國. 學. H18: The more open to experience people are, the more they perceive Uber as. ‧. trialable.. Nat. io. sit. y. Trialability, which is associated with the experiment, can also be attractive. er. to extraverted potential consumers as they are active (Ciavarella et al., 2003).. al. n. v i n C h with an innovation Willing to take actions to experiment e n g c h i U may help them find out the trialability more easily. The thesis proposes that H19: The more extraverted people are, the more they perceive Uber as trialable. Neuroticism is characterized as anxious, having phobias, worrisome, and stressful (Ahmad, Ganaie, & Suhial, 2015), it is not difficult to figure out that neurotic people, due to their phobias towards new inventions, may not be bold enough to try out new products, thus not being able to find out whether a new 20. DOI:10.6814/NCCU201900593.

(26) product is trialable. The thesis suggests a negative relationship between neuroticism and trialability, indicating that H20: The more neurotic people are, the less they perceive Uber as trialable. Complexity Complexity is the perceived difficulties that people may encounter when seeking to understand and use technology (Shihab, Meilatinova, Hidayanto, & Herkules, 2017).. 立. 政治大. While conscientious people may also find an innovation complex, they are. ‧ 國. 學. less likely to become troubled by it. Rather, due to their persevering nature. ‧. (Ciavarella et al., 2003) they are likely to wade through difficulties and. Nat. io. sit. y. complexity as they investigate the new product. In light of this, the thesis. n. al. er. suggests a negative relationship between conscientiousness and complexity and hypothesizes that. Ch. engchi. i n U. v. H21: The more conscientious people are, the less they are daunted by the complexity of Uber. Complexity of an innovative product can be perceived as less of an issue for people who are curious and have an appetite to learn, which are important properties of openness to experience. Despite they may also perceive an innovation as complex, their positive attitude towards learning (Matzler et al., 21. DOI:10.6814/NCCU201900593.

(27) 2011) would not allow them to become intimidated by the complexity. Therefore, the thesis hypothesizes that H22: The more open to experience people are, the less they are daunted by the complexity of Uber. Extraversion is linked to assertiveness (Ciavarella et al., 2003). Assertive people are confident (Kolb et al., 2009). Though they may see an innovation as. 政治大. complex, they are less likely to feel intimidated by it. The study proposes that. 立. H23: The more extraverted people are, the less they are daunted by the. ‧ 國. 學. complexity of Uber.. ‧. Anxious, neurotic people (Aslan et al., 2012) and easily stressed (Moutafi. Nat. io. sit. y. et al., 2005) may find investigating, trialing or using an innovation a stressful. er. thing to do, thus perceiving it as more complex than people who are. al. n. v i n C h the thesis proposes comparatively less neurotic. Hence, e n g c h i U that. H24: The more neurotic people are, the more they are daunted by the complexity of Uber.. Demographics and innovation adoption: H25 & 26 In the innovation adoption case of Uber, gender difference emerged to be a factor predicting adoption behavior. A study on Uber adoption in Bangladesh 22. DOI:10.6814/NCCU201900593.

(28) and Pakistan discovered that women were more likely to take Uber than men because Uber showed a better safety record than other local taxi services. Uber also allowed passengers to designate their preferred drivers, so women can choose female drivers as they want, thus leading more women to be willing to use Uber compared with men (Zafar & Rahman, 2018). Therefore, the thesis would like to find out whether gender also plays a role. 政治大. in Uber usage in Taiwan and hypothesizes that. 立. H25: Gender difference will influence Uber usage.. ‧ 國. 學. Also, a study on ride-hailing services (such as Uber) in California showed. ‧. that people with higher income are more likely to use the services than. Nat. sit er. io. Mokhtarian, 2018). y. individuals with lower or medium-income level (Alemi, Circella, Handy, &. al. n. v i n The thesis wants to testC if Taiwan h e n gis cthehsame i U in terms of the relationship. between Uber usage and income level and predicts that H26: The higher the monthly disposable income, the more they will use Uber.. 23. DOI:10.6814/NCCU201900593.

(29) The research structure is shown below, and a full list of hypotheses is in Table 2.. Demographics • Gender • income Uber usage. 政治大. 立. 學. ‧ 國. Personality traits • Conscientiousness • Openness • Extraversion • Neuroticism. PAI • Relative advantage • Compatibility • Trialability • Complexity. Table 2. List of hypotheses PAI and innovation adoption. ‧. H1: The perceived relative advantage of Uber is positively related to its adoption. H2: The perceived compatibility of Uber is positively related to its adoption.. io. Personality traits and innovation adoption. n. al. er. sit. y. Nat. H3: The perceived trialability of Uber is positively related to its adoption. H4: The perceived complexity of Uber is negatively related to its adoption.. i n U. v. H5: The more conscientious people are, the more they are going to use Uber. H6: The more open to experience people are, the more they are going to use Uber. H7: The more extraverted people are, the more they are going to use Uber. H8: The more neurotic people are, the less they are going to use Uber.. Ch. engchi. Personality traits and PAI-- Relative advantage H9: The more conscientious people are, the more they find out the relative advantage of Uber. H10: The more open to experience people are, the more they find out the relative advantage of Uber. H11: The more extraverted people are, the more they find out the relative advantage of Uber. H12: The more neurotic people are, the less they find out the relative advantage of Uber.. 24. DOI:10.6814/NCCU201900593.

(30) (Continued from Table 2) Compatibility H13: The more conscientious people are, the more they perceive Uber as compatible with existing products. H14: The more open to experience people are, the more they perceive Uber as compatible with existing products. H15: The more extraverted people are, the more they perceive Uber as compatible with existing products. H16: The more neurotic people are, the less they perceive Uber as compatible with existing products. Trialability H17: The more conscientious people are, the more they perceive Uber as trialable. H18: The more open to experience people are, the more they perceive Uber as trialable. H19: The more extraverted people are, the more they perceive Uber as trialable. H20: The more neurotic people are, the less they perceive Uber as trialable.. 立. ‧ 國. 學. Neuroticism. 政治大. ‧. H21: The more conscientious people are, the less they are daunted by the complexity of Uber. H22: The more open to experience people are, the less they are daunted by the complexity of Uber.. y. Nat. sit. n. al. er. io. H23: The more extraverted people are, the less they are daunted by the complexity of Uber. H24: The more neurotic people are, the more they are daunted by the complexity of Uber.. Ch. engchi. Demographics and Uber usage. i n U. v. H25: Gender difference will influence Uber usage. H26: The higher the monthly disposable income, the higher the chances they will use Uber.. 25. DOI:10.6814/NCCU201900593.

(31) Chapter 3 Methodology. This research was conducted through mixed methods, combining a closedended quantitative survey and open-ended, structured qualitative interviews.. 政治大. This allowed the study to gain a better understanding of the research topic, as. 立. personality, or the human mind, is a complex issue to comprehend.. ‧. ‧ 國. 學. 3.1 Data collection and sample. sit. y. Nat. io. al. er. In the quantitative part, an online survey was conducted using snowball. v. n. sampling, meaning that the questionnaires were first distributed to the author’s. Ch. engchi. i n U. contacts then spread out to the networks of those contacts. In addition to snowballing, the survey was also posted on social media, such as online graduate student communities on Facebook. A total of 337 questionnaires were collected. All questionnaires were filled out completely, and no missing data was found. In the qualitative part, the author selected four participants, two had used Uber before the in-depth interviews, and two had never tried. Two of them were 26. DOI:10.6814/NCCU201900593.

(32) graduate students at National Chengchi University, both of whom heard about Uber and are Uber users. The third interviewee had a research background in transportation issues. The fourth one was a cram school teacher. They were given open-ended questions corresponding to research hypotheses. Questions are listed in appendix 1.. 3.2 Measures. 立. 政治大. This research drew on three ready-made measurement scales, built for. ‧ 國. 學. personality traits and PAI, respectively. Modification of those scales was being. er. io. sit. y. Nat 3.2.1 Personality traits. ‧. made to fit the research purpose of the thesis.. al. n. v i n C h for itself theU“Big Five Inventory,” whose The FFM of personality developed engchi. copyright was held by Berkeley Personality Lab (The Big Five Inventory, 2007). The Lab offered several language versions of the personality test in the form of the 5-point Likert scale and allowed academic use for free. The thesis took advantage of the Chinese version of the scale then altered it into a measurement tool to evaluate the personality traits of Uber users, including conscientiousness, openness to experience, extraversion, and neuroticism. 27. DOI:10.6814/NCCU201900593.

(33) 3.2.2 PAI Two existing scales previously developed for PAI-related researches were employed to form the thesis’ measurement part that tested respondents’ perceptions towards the attributes of Uber. To obtain better knowledge about consumer product and marketing management, Flight et al. (2011) developed a comprehensive 43-item scale. 政治大. that measured 15 unique innovation characteristics, with the help of data. 立. collected from 628 respondents.. ‧ 國. 學. Moore et al. (1991), on the other hand, investigated users’ adoption of. ‧. Personal Work Stations and created a 38-item measurement instrument that. Nat. io. sit. y. comprised eight PAI, based on Rogers’ classic five (1995). They added. n. al. er. voluntariness, image, and ease of use while replacing observability with. Ch. engchi. visibility and result demonstrability.. i n U. v. Based on the three measurement scales, this research formed a 39-item scale that measures Uber usage behavior, personality traits, and PAI. And since Uber usage is a continuous behavior, it was measured by frequency of use over the past year. The response category included “never,” “1-5 times,” “6-10 times,” “11-15 times,” “more than 15 times.” Personality traits and PAI were measured by the 5-point Likert scale, ranging from 1 (strongly disagree) 28. DOI:10.6814/NCCU201900593.

(34) to 5 (strongly agree). Each construct of personality trait and PAI was measured with four questions.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 29. DOI:10.6814/NCCU201900593.

(35) Chapter 4 Results. This chapter details the quantitative and qualitative results of the research. The first section provides a demographic overview of the respondents. The second section presents the measurement validation detailing the reliability of. 政治大. constructs used to test the proposed relationships. The third section examines. 立. ‧ 國. 學. hypotheses employing multiple linear regression analysis, with the support of qualitative interviews.. ‧ sit. y. Nat. io. n. al. er. 4.1 Demographic profile. i n U. v. Of all 337 responses collected, non-Uber users accounted for 43.3 percent. Ch. engchi. of the sampled population, whereas Uber users took up the rest 56.7 percent. The majority of users reported that they took Uber 1-5 times over the past year (31.5 percent). Most of the Uber adopters spent less than NT$500 using the service over the same period (19 percent). Female respondents constituted 72.1 percent of the sample, while male respondents accounted for 27.9 percent. Some 49 percent of the total respondents were aged 21-39 years old, although people between 40-59 years old also took up 43.3 percent of the total sample. 30. DOI:10.6814/NCCU201900593.

(36) Up to 73.9 percent of the respondents said they live in Taipei. About half of the people (50.4 percent) owned a bachelor’s degree or equivalent, while those who owned a master’s degree or above also accounted for 31.2 percent of the total population. Almost half of the respondents (49.6 percent) reported that over NT$30,000 is at their disposal per month. Another 22.6 percent said NT$10,001- NT$20,000, 16.3 percent of them said NT$10,000 and below, and. 政治大. 11.6 percent estimated their monthly disposable income to be NT$20,001-. 立. 學. ‧ 國. NT$29,999. Table 3 details the demographic results. Table 3. Demographic profile Respondents 94 243. n. e n g2c h i. sit. 8 165 146 18. er. io. Ch. 27.9 72.1. y. Nat. al. ‧. Gender Male Female Age 20 and below 21-39 40-59 60 and above Educational Level Junior high school and below Senior high school or equivalent University or equivalent Graduate school Others Monthly disposable income NT$10,000 and below NT$10,001- NT$20,000 NT$20,001- NT$29,999 NT$30,000 and above Residence Taipei Taichung Kaohsiung Others. Percentage. i n U. v. 2.4 49 43.3 5.3. 59 170 105 1. 0.6 17.5 50.4 31.2 0.3. 55 76 39 167. 16.3 22.6 11.6 49.6. 249 20 26 42. 73.9 5.9 7.7 12.5. Note: The total percentage of monthly disposable income became 101 percent due to rounding. 31. DOI:10.6814/NCCU201900593.

(37) 4.2 Measurement Validation Validation of the measurements was evaluated based on individual indicator reliability and construct reliability. Factor analysis was used to check the loading value of each indicator. An indicator can be viewed as part of a construct when its loading value is ≥.707, but values between .4 and .7 are allowed if it helps improve content validity (Sánchez & Sahuquillo, 2016). In this. 政治大. research of Uber usage, three items concerning openness to experience,. 立. compatibility, and trialability had loading values below .5 and were removed to. ‧ 國. 學. enhance indicators’ validity.. ‧. Construct reliability was assessed employing Cronbach’s Alpha. All. Nat. io. sit. y. Cronbach coefficient alpha values, except conscientiousness, were higher than. n. al. er. the threshold of .7. Konduri, Gupchup, Borrego, and Worley-Louis (2006). Ch. engchi. pointed out that a reliability coefficient of .7. iv n isUconsidered acceptable, but. reliabilities of .5 to .7 can also be sufficiently reliable.. Table 4. Measurement Validation Constructs and indicators. Factor Loading. Conscientiousness Deliberate Diligent Persevering Achievement-oriented. Cronbach’s Alpha .641. .529 .774 .779 .680 32. DOI:10.6814/NCCU201900593.

(38) (Continued from Table 4). Openness to experience Imaginative Curious Willing to learn Extraversion Energetic Risk-taking Assertive Confident Neuroticism. .763 .695 .674 .831 .855 .793 .787 .592 .886. 政治大 .868 .780 .797 .740 .877 .689. n. Ch. engchi. y. .782 .887. al. sit. Requires previous knowledge to learn Difficult in design. io. .785 .945 .931. .865. er. Nat. Acceptable in society Compatible with existing lifestyle Compatible with the lifestyle longed for Trialability Can be trialed for free Offers lots of opportunities to try Can be fully trialed Complexity Difficult to use Takes a long time to learn. ‧. ‧ 國. 立. .769 .856 .805. 學. Moody Fearful Anxious Stressed Relative Advantage Time-saving Cost-saving Efficiency-improving Self-image enhancing Compatibility. .738. i n U. v. .876. .875 .895 .915. .877 .862 .886. 33. DOI:10.6814/NCCU201900593.

(39) 4.3 Data analysis Multiple linear regression analysis was employed to analyze the data for two main reasons. Firstly, personality trait is not mutually exclusive as everybody owns more than one trait. Secondly, the decision of Uber usage is more of an aggregate result after all viewing and reviewing all attributes of an innovation. An adoption behavior is hardly the consequence of evaluating only. 政治大. one PAI. Therefore, to investigate whether personality differences and PAI lead. 立. to the usage of Uber and whether these personality traits lead to divisive. ‧ 國. 學. perceptions towards the attributes of Uber, multiple regression serves as an. ‧. appropriate analyzing tool.. Nat. io. sit. y. The 26 proposed hypotheses were divided and tested in groups, as shown. er. in this section. Bivariate correlations were first conducted to each group before. al. n. v i n C h The results were the multiple regression analysis. e n g c h i U shown as below. To save space, in Table 5 Uber usage is abbreviated as Uber, conscientiousness as CN, extraversion as EV, openness to experience as OP, neuroticism as NR, relative advantage as RA, compatibility as CP, trialability as TR, and complexity as CX.. 34. DOI:10.6814/NCCU201900593.

(40) Table 5. PAI, personality traits and Uber usage correlations Uber. CP. TR. .441**. .616**. .588** -.468**. .000. .002. .000. .000. .000. .000. 337. 337. 337. 337. 337. 337. 337. 337. 337. .172**. 1. .444**. .534**. -.025. .285**. .149**. .000. .000. .643. .000. .006. .000. .001. CX. 337. 337. .293** -.187**. 337. 337. 337. 337. 337. 337. 337. .120*. .444**. 1. .433**. .011. .195**. .208**. .027. .000. .000. .838. .000. .000. .000. .019. 337. 337. 337. 337 337 337 337 治政 .256 .534 .433 1 -.281 大 .207 立 .000 .000 .000 .000 .000 337. 337. **. **. **. 337. **. 337. 337. 337. 337. -.168**. -.025. .011 -.281**. 1. .002. .643. .838. .000. 337. 337. 337. 337. .441**. .285**. .195**. .000. **. .232**. .291** -.127*. .340** -.206**. .000. .000. .000. 337. 337. 337. 337. .054. -.064. -.072. .189**. ‧. CX. .027. .326. .239. .189. .000. 337. 337. 337. 337. 337. .207**. .054. 1. .657**. .000. .000. .326. .000. .000. .000. 337. 337. 337. 337. 337. 337. **. **. **. .000. .006. .000. .000. .239. .000. 337. 337. 337. 337. 337. 337. 337. .588**. .293**. .291**. .340**. -.072. .624**. .774**. 1 -.518**. .000. .000. .000. .000. .189. .000. .000. .000. 337. 337. 337. 337. 337. 337. 337. .000. al. n. TR. .002. io. CP. .256** -.168**. Nat. RA. .120*. .002. RA. 學. NR. .172**. 1. NR. 337. .616**. C .208 .232 .149 h e ngchi. -.468** -.187** -.127* -.206**. y. OP. OP. sit. EV. EV. er. CN. Pearson Correlation Sig. (2tailed) N Pearson Correlation Sig. (2tailed) N Pearson Correlation Sig. (2tailed) N Pearson Correlation Sig. (2tailed) N Pearson Correlation Sig. (2tailed) N Pearson Correlation Sig. (2tailed) N Pearson Correlation Sig. (2tailed) N Pearson Correlation Sig. (2tailed) N Pearson Correlation Sig. (2tailed) N. ‧ 國. Uber. CN. i v337 n U .657 -.064 337. **. 1. .624** -.351**. .774** -.527** .000. .000. 337. 337. 337. 337. .189** -.351** -.527** -.518**. 1. .000. .001. .019. .000. .000. .000. .000. .000. 337. 337. 337. 337. 337. 337. 337. 337. 337. **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). 35. DOI:10.6814/NCCU201900593.

(41) 4.3.1 PAI, personality traits, and Uber usage: H1—H8 Hypotheses 1 to 8 tested about PAI and personality traits’ relationships with Uber usage. Together, the eight variables of PAI and personality explained 43.2 percent of the variance in Uber usage, with an adjusted R Square value at .432. Three of the four variables, namely compatibility, trialability, and. 政治大. complexity, were found to be significant in predicting the usage of Uber in. 立. Taiwan. Compatibility was positively related to Uber usage (β = .349, p < .001).. ‧ 國. 學. Trialability was positively related to its usage, with β=.216, p <.05. Complexity. ‧. was negatively related to its usage, with β = -.138, p < .05.. Nat. io. sit. y. The statistically supported hypotheses meant that perceptions towards the. er. perceived compatibility and trialability of Uber were positively related to its. al. n. v i n C hnegatively relatedUto Uber usage. usage, whereas complexity was engchi. Concerning the relationship between compatibility and Uber usage, D.L., a researcher on Taiwan’s transportation regulations and a non-Uber user, commented: People who are already accustomed to dialing 55688 (a ride-hailing service provided by Taiwan Taxi) might think that Uber is as convenient as 55688, and therefore, are willing to use Uber. 36. DOI:10.6814/NCCU201900593.

(42) The statement of “as convenient as” can be seen as a claim that supported the perceived compatibility of Uber. Personality-wise, quantitative, and qualitative analyses produced mixed results. Quantitative results showed that no relationships were found between the four personality traits and the usage behavior of Uber (Table 6). However, in-depth interview suggested that extraversion is likely to be linked positively with Uber usage.. 政治大. Rendering her thoughts as to the kind of people who would use Uber, L.K.,. 立. an Uber user and a graduate student, said:. ‧ 國. 學. I think it must be extraverts who are more likely to use Uber. Since there are. ‧. still some disputes about Uber, if you are not comfortable enough, you. Nat. io. sit. y. wouldn’t use it. There are conflicts about licenses and with other taxi drivers.. er. Also, extraverted people are those who are comfortable with sharing. al. n. v i n C hUber and other taxiUcompanies. You have to share personal information with engchi. with them your information like credit card numbers. So, it has to be risktaking people or something like that.. 37. DOI:10.6814/NCCU201900593.

(43) Table 6. PAI, personality traits, and Uber usage: H1—H8. 立. 政治大. ‧ 國. 學. 4.3.2 Personality traits and relative advantage: H9—H12. Hypotheses 9, 10, 11, and 12 propose possible relationships between the. ‧. four featured personality traits and the perceived relative advantage of Uber.. sit. y. Nat. io. n. al. er. In total, the four personality traits (conscientiousness, openness to. i n U. v. experience, extraversion, and neuroticism) explained 8.5 percent of the. Ch. engchi. variance in how people perceived the relative advantage of Uber, with an adjusted R Square value of .085. Of the four traits tested, conscientiousness was found to be significantly related to relative advantage, with β =.211, p < .05. Therefore, hypothesis 9 was statistically proven valid: The more conscientious people are, the more they find out the relative advantage of Uber. Table 7 details the results.. 38. DOI:10.6814/NCCU201900593.

(44) Table 7. Personality traits and relative advantage: H9—H12 Hypotheses. Coeff. (β). t-value Sig.. Supported. H9. Conscientiousness. .211. 3.259. .001. Yes. H10. Openness to experience. .060. .986. .325. No. H11. Extraversion. .092. 1.326. .174. No. H12. Neuroticism. .084. 1.521. .129. No. a. Adjusted R square = .085 b. Dependent variable: relative advantage. Speaking of why conscientious people can recognize that there are relative advantages about Uber, D.L. commented:. 政治大 People who are fussy and attend to details are more likely to find out the 立. ‧ 國. 學. relative advantage of Uber when compared with local taxis.. ‧. Attention to details, as noted by the interviewee, is a common property. sit. y. Nat. shared by conscientious people. This means that conscientious people are. n. al. er. io. more likely to discover the relative advantage of Uber.. Ch. engchi. i n U. v. 4.3.3 Personality traits and compatibility: H13—H16 Hypotheses 13, 14, 15, and 16 predict possible relationships between personality traits and the perceived compatibility of Uber. The explanatory power of personality traits to predict the variance in the compatibility of Uber was 5.7 percent, as the adjusted R Square value was at .057. Two of the four personality traits were significantly related to the compatibility of Uber.. 39. DOI:10.6814/NCCU201900593.

(45) Openness to experience was positively related to the compatibility of Uber (β = .136, p < .05), and extraversion was also positively related to compatibility, with β=.170, p <.05. Hypotheses 14 and 15 were supported, showing that the more open to experience or extraverted people are, the more they can perceive Uber as compatible with existing products in the market, for instance, local taxis. Table 8 details the results of the hypotheses.. 政治大 Coeff. (β) t-value. Table 8: Personality traits and compatibility: H13—H16. 立 Conscientiousness. H13. -.003. -.042. Openness to experience. .136. 2.201. H15. Extraversion. .170. 2.469. H16. Neuroticism. -.018. -.322. .967. No. .028. Yes. .014. Yes. .748. No. er. io. sit. y. Nat. a. Adjusted R square = .057 b. Dependent variable: compatibility. Supported. ‧. H14. Sig.. 學. ‧ 國. Hypotheses. A positive relationship was also found between openness to experience and. n. al. Ch. engchi. compatibility as Uber user, L.K, pointed out:. i n U. v. Tech-savvy people will find about the compatibility of Uber. Because they are willing to learn about technology, feel comfortable using it and thus can tell about the compatibility of Uber more easily than other people.. A willingness to learn new stuff has been associated as one of the important characteristics of openness to experience (Matzler e. al., 2011).. 40. DOI:10.6814/NCCU201900593.

(46) 4.3.4 Personality traits and trialability: H17—H20 Hypotheses 17, 18, 19, and 20 suggest personality traits’ possible relationships with the perceived trialability of Uber. Personality traits accounted for 13.9 percent of the variance in the trialability of Uber, with an adjusted R Square value at .139. Two of the four personality traits were shown to be. 政治大. significantly related to the PAI. Openness to experience was positively related. 立. to the trialability of Uber (β = .148, p < .05). Extraversion was also positively. ‧ 國. 學. related to trialability, with β=.212, p <.05.. ‧. Thus, Hypotheses 18 and 19 were quantitively supported, meaning the. Nat. io. sit. y. more open to experience or extraverted people are, the more they can view. n. al. er. Uber as trialable as an innovation. Table 9 details the results of the hypotheses.. Ch. i n U. v. Table 9: Personality traits and trialability: H17—H20 Hypotheses Coeff. (β) t-value H17. Conscientiousness. H18. engchi. Sig.. Supported. .114. 1.812. .071. No. Openness to experience. .148. 2.516. .012. Yes. H19. Extraversion. .212. 3.228. .001. Yes. H20. Neuroticism. -.011. -.202. .840. No. a. Adjusted R square = .139 b. Dependent variable: trialability. Openness to experience and perceived trialability was also found to be positively related as T.T., a graduate student and an Uber user, commented:. 41. DOI:10.6814/NCCU201900593.

(47) The first time I tried Uber was not out of necessity. I was just interested in it. Just curious. I later found out that it was kind of convenient, and then I grew to have a preference of it over other taxi services.. Curiosity is linked to the personality trait of openness to experience (McCrae et al., 1999).. 政治大. 4.3.5 Personality traits and complexity: H21—H24. 立. Hypotheses 21, 22, 23, and 24 suggest the possible relationships between. ‧ 國. 學. personality traits and the perceived complexity of Uber. Personality traits. ‧. explained 6.4 percent of the variance in the complexity of Uber, with an adjusted. Nat. io. sit. y. R Square at .064.. er. Of all the personality traits, neuroticism was found to be significantly related. al. n. v i n -.165,Cph < .05. Hypothesis e n g c h i U 24,. to the PAI, with β =. which predicts that. neuroticism is positively related to the complexity of Uber, was proven valid. The result showed that the more neurotic people are, the more they perceive the complexity of Uber as too daunting to cope with. It meant that neurotic people were more likely to see installing mobile APPs, registering an account, getting acquaintances with the GPS, then finally hailing a car as too complicated to adopt. The statistical result was also consistent with a qualitative 42. DOI:10.6814/NCCU201900593.

(48) interview, in which an interviewee mentioned that being timid can hold people back from using Uber. K.J., a non-Uber user, commented: I think timid people may hold a reserved attitude toward Uber adoption.. Timid is also one of the characteristics of neuroticism (Hurley, 2010). Table 10 details the results of the hypotheses. Table 10: Personality traits and complexity: H21—H24 Hypotheses Coeff. (β) tSig. value. 政-.124治 -1.897 大. Conscientiousness. H22. Openness to experience. -.041. -.675. H23. Extraversion. -.075. -1.097. H24. Neuroticism. .165. 2.951. No. .500. No. .273. No. .003. Yes. ‧. ‧ 國. 立. .059. 學. H21. Supported. a. Adjusted R square = .064 b. Dependent variable: complexity. n. Ch. engchi. er. io. sit. y. Nat. al. i n U. 4.3.6 Demographics and Uber usage: H25 & H26. v. Hypotheses 25 and 26 suggest that gender and monthly disposable income may also influence people’s decision to use Uber. Demographics explained 4.6 percent of the variance in Uber usage, with an adjusted R Square value at .046. Both gender and monthly disposable income were shown to be significantly related to Uber usage. Gender difference influenced Uber usage (β = -.194, p < .001), with male respondents (coded as 1) being the majority of. 43. DOI:10.6814/NCCU201900593.

(49) users. Monthly disposable income was also positively related to Uber usage, with β=.116, p <.05. Table 11 details the results of the hypotheses. Table 11: Demographics and Uber usage: H25 & H26 Hypotheses Coeff. (β) t-value. Sig.. Supported. H25. Gender. -.194. -3.633. .000. Yes. H26. Monthly disposable income. .116. 2.179. .030. Yes. a. Adjusted R square = .046 b. Dependent variable: Uber usage. A qualitative interview with a female non-Uber user explained the gender. 治政 difference in Uber usage. K.J., a cram school teacher 大and Non-Uber user, said: 立 ‧ 國. 學. I would consider credibility and safety issues. Whether the use of Uber is protected by the government is very important. For women, safety is the. ‧. primary concern. In the early days, it was unsafe for females to take taxis,. sit. y. Nat. io. n. al. er. especially after Peng Wan-ru (彭婉如, a prominent women movement figure. i n U. v. in Taiwan) was killed while taking a taxi ride. It was terrifying. It is better for. Ch. engchi. women to choose a taxi company that has earned government approval. But as to Uber, I would ask: what it is? With doubts in my mind, I would not take risks and use Uber.. An interview also supports a positive relationship between monthly disposable income and Uber usage. D.L., the non-Uber user, commented:. 44. DOI:10.6814/NCCU201900593.

(50) People who are highly paid may use Uber more. Uber drivers would wait for you like private car chauffeurs, and not turn away and just leave if you are late.. 4.4 Qualitative results Most of the qualitative results were consistent with quantitative analysis, and there are two things worth mentioning. Firstly, while multiple regression. 政治大. analysis did not show relationships between personality traits and Uber usage,. 立. a qualitative interview suggested that extraverted people were more likely to. ‧ 國. 學. take Uber. Secondly, albeit not addressed in hypotheses, interviewees pointed. ‧. out the needs of various kinds as a motivation to use Uber.. y. Nat. er. io. sit. K.J., a non-Uber user, said:. It’s not necessarily about personality traits; it’s more about necessity. Let’s. n. al. Ch. engchi. i n U. v. say people who live in a rural area and have a certain need for transportation. It might not be possible for them to grab a taxi on the roadside and go, and this is where Uber comes to mind.. She also mentioned the need for privacy as a motivation to use Uber: People with a greater need for privacy might find out the differences (the relative advantage and compatibility) between an Uber and existing products since they are always searching for transportation tools that better protect 45. DOI:10.6814/NCCU201900593.

(51) their privacy. Take Han Bing, daughter of Kaohsiung mayor. One day she was caught off guard and taken pictures by a stranger as she got off a taxi. She might want to find an alternative way to transport, rather than taking the local taxis.. T.T., an Uber adopter, commented: Those people who have a constant need to travel by taxi or the alike would. 政治大. notice the relative advantage/compatibility of Uber. People who only take a. 立. taxi once in a while would not recognize the differences between the two. My. ‧ 國. 學. mother, for example, is now an Uber user as she grew older and no longer. ‧. rides a motorcycle. She noticed that Uber offers a lot of discounts, and she. Nat. n. al. er. io. sit. y. can tell which one saves more money.. Ch. engchi. i n U. v. 46. DOI:10.6814/NCCU201900593.

(52) Chapter 5 Discussion. The purpose of this study is to investigate the roles of personality traits in Uber usage and the perception of innovation attributes. The thesis. 政治大 linear regression analysis and qualitative interviews were used to test the 26 立 hypothesizes that people have the final say in using an innovation. Multiple. ‧ 國. 學. proposed relationships, and results were detailed in Chapter 4.. ‧ sit. y. Nat. 5.1 Summary of findings. n. al. er. io. The first finding is that the adjusted R-square values explaining the. Ch. i n U. v. variance in Uber usage and PAI were low as presented in Chapter 4. Chatterjee,. engchi. Singh, Goyal, and Gupta (2015) explained the low R-squared values and adjusted R-squared values, arguing that human behaviors are hard to predict. They pointed out that if the R-square values are low but have statistically significant predictors, one can still draw important conclusions based on the data results.. 47. DOI:10.6814/NCCU201900593.

(53) 5.1.1 PAI and innovation adoption When all four PAI (relative advantage, compatibility, trialability, and complexity) were counted in as independent variables in multiple regression analysis, all were predictors to the dependent variable of Uber usage, except relative advantage (β =.029, p > .05). The reason that relative advantage failed to predict a relationship with. 政治大. innovation adoption can be explained by Tornatzky and Klein (1982), who. 立. claimed that relative advantage is a reduplicative notion to compatibility (Lin &. ‧ 國. 學. Li, 2014). This could mean that for those people who have a habit of using. ‧. mobile phones for various functions, Uber is just another taxi service with APPs. Nat. io. sit. y. and is no different from other products, for example, Taiwan Taxi, one of the. n. al. er. leading cab services in Taiwan. As a result, the relative advantage of Uber was. Ch. engchi. not being distinguished as an attribute.. i n U. v. 5.1.2 Personality traits and innovation adoption The results in Chapter 4 showed mixed results as qualitative analysis disproved all hypotheses concerning personality traits and Uber usage, while qualitative interview said that extraversion might be linked to innovation adoption. Hence, there is no concrete conclusion as to whether personality 48. DOI:10.6814/NCCU201900593.

(54) traits have, or do not have an influence on innovation adoption, and this requires more researches.. 5.1.3 Personality traits and PAI Relative Advantage Of all personality traits, only conscientiousness was found to be positively. 政治大. associated with the perceived relative advantage of Uber (β =.211, p < .05),. 立. meaning that the more conscientious people are, the more they can find out the. ‧ 國. 學. relative advantage of Uber.. ‧. Therefore, conscientiousness may predict a positive relationship with. Nat. io. sit. y. relative advantage. Open people’s ability to find out the relative advantage of. er. an innovation as cited in previous literature could be enlarged. The same thing. al. n. v i n Ccounterparts. happened to their extraverted h e n g c hAlso, i Uneurotic people’s inability to. find out relative advantage was over-emphasized as the negative relationship between neuroticism and relative advantage was not significant (β =084, p > .05). Compatibility & Trialability As expected, both openness to experience and extraversion were positively associated with the trialability of an innovation, meaning that people 49. DOI:10.6814/NCCU201900593.

(55) who are curious, broad-minded, assertive and confident are more likely to perceive Uber as compatible with existing taxi services. They are also more likely to view an innovation as trialable in the case of an innovation. There was no relationship found between conscientiousness and relative advantage (β = -.003, p > .05). The “fussy and attend to details” comment provided by non-Uber user D.L. in Chapter 4 may help explain the result. It is. 政治大. possible that conscientious people are so good at attending to details that they. 立. the two as compatible with each other.. 學. ‧ 國. found only what is different (such as relative advantage) and failed to perceive. ‧. Neuroticism’s negative relationships with compatibility and trialability were. Nat. io. sit. y. not significant. Standardized beta coefficient between neuroticism and. er. compatibility has a value of -.018, p > .05, while the value between neuroticism. al. n. v i n CThe and trialability is -.011, p > .05. h efindings h i U that just because people n g csuggested are fearful, phobias, and easily worried does not mean they lack the abilities to perceive an innovation as something compatible with previous products. Also, being neurotic does not mean they cannot see a new product as trialable.. 50. DOI:10.6814/NCCU201900593.

(56) Complexity Of the four personality traits tested, only neuroticism is significantly related to complexity (β =.165, p < .05), suggesting that the more neurotic people are, the more they perceive the complexity of an innovation as daunting. Although conscientiousness, openness to experience, and extraversion did show negative standardized beta coefficient values, none of these was. 政治大. significant enough to prove a negative relationship with the complexity (p > .05).. 立. This meant that those personality traits’ capabilities to overcome the perceived. ‧ 國. 學. complexity might be highly stressed in previous studies, and people do feel. ‧. daunted sometimes no matter how conscientious, open, or extraverted they are.. al. er. io. sit. y. Nat 5.1.4 Demographics and Uber usage. n. v i n C h of gender andUmonthly disposable income Both demographic variables engchi. were proven to influence Uber usage. But, contradicting to the literature in Chapter 2 where Zafar et al., (2018) discovered that more women adopted Uber due to the company’s better safety records in Bangladesh and Pakistan, this study in Taiwan showed male respondents used more Uber than women did. Therefore, the gender issue of Uber adoption should be taken into the social context, and not to be generalized universally. 51. DOI:10.6814/NCCU201900593.

(57) 5.2 Other findings 5.2.1 Hierarchical regression analysis results To gain a comprehensive view of the variables’ influences on Uber usage, a hierarchical regression analysis was conducted, and the results were displayed in Table 12. Table 12: Hierarchical multiple regression results on Uber usage Coeff. (β). Neuroticism Model 2. Nat. Openness to experience (Continued from Table 12) Extraversion. io. Neuroticism. al. n. Relative advantage Compatibility Trialability Complexity. .183. 2.670. .008. -.116. -2.068. .039. .018. .343. .732. -.073. -1.518. .064. 1.185. -.087. -1.939. .053. v i n 4.781 C h .349 U i e h n .216 g c 3.038. .614. ‧. Conscientiousness. .827. 學. ‧ 國. Extraversion. .320. .130. y. 立. Openness to experience. 治 .996 政 .065 大 .013 .218. .029. sit. Conscientiousness. Sig.. er. Model 1. t-value. .505. .237. .000 .003. -.138. -2.724. .007. Conscientiousness. .018. .337. .736. Openness to experience. -.073. -1.498. .135. Extraversion. .057. 1.037. .300. Neuroticism. -.082. -1.824. .069. Relative advantage. .036. .622. .534. Compatibility. .349. 4.744. .000. Trialability. .216. 3.026. .003. Complexity. -.139. -2.741. .006. Gender. .018. .420. .675. Model 3. 52. DOI:10.6814/NCCU201900593.

(58) (Continued from Table 12). Monthly disposable income. .047. 1.118. .264. a. (Model 1) Adjusted R square= .068 b. (Model 2) Adjusted R square= .432 c. (Model 3) Adjusted R square= .431 d. Dependent variable: Uber usage. Model 1 in Table 12 showed that when the four personality traits were put together to test their relationship with Uber usage, personality in total explained 6.8 percent of the variance in Uber usage (adjusted R square = .068). Only. 政治大. extraversion (β =.183, p < .05) and neuroticism (β = -.116, p < .05) were. 立. significantly related to Uber usage.. ‧ 國. 學. Model 2 showed that when PAI were added in, the explanatory power was. ‧. enhanced to 43.2 percent (adjusted R square = .432). The model indicated that. sit. y. Nat. io. al. er. not a single relationship was found between personalities and Uber usage. The. v. n. only factors that predicted Uber usage were compatibility (β = .349, p < .001),. Ch. engchi. i n U. trialability (β = .216, p < .05) and complexity (β = -.138, p < .05). Model 3 showed that when demographic factors (gender and monthly disposable income) were also counted in, the explanatory power in variance of Uber usage was slightly lowered to 43.1 percent (adjusted R square = .431), weakened by the two demographic variables. In model 3, gender and monthly disposable income became non-significant using hierarchical regression. Compatibility (β = .349, p < .001), trialability (β = .216, p < .05) and complexity 53. DOI:10.6814/NCCU201900593.