INCREASING TRUST IN HIGH-RISK DIGITAL MARKETS:
A CLOSER LOOK AT THE SERVICES INDUSTRY
Master thesis, MScBA, specialization Strategic Innovation Management University of Groningen, Faculty of Economics and Business
Abstract
The purpose of this study is to investigate which signals customers use in online high-risk service markets - compared to online low-risk service markets - to determine trust and risk and form purchasing intentions. Although much is known for trust solutions in products, much less literature is present on the services markets and on the effects of trust mechanisms in markets with different risk-levels. The study is performed by means of an online test environment and survey among 509 consumers. The results prove the main dynamics between perceived trust, perceived risk and purchasing intention in service markets to be comparable with those in product markets, across both high-risk and low-risk market situations. Furthermore, this study shows lower perceived trust and (surprisingly) lower perceived risk towards suppliers in high-risk service markets, of which the latter is possibly explained by increased search behavior. In high-risk service markets customers prefer technical trust solutions, while in low-risk service markets customers prefer social trust solutions. Academic and managerial implications as well as future research directions are discussed.
Keywords; trust solutions, online purchasing, perceived risk, perceived trust, purchasing intentions, services industry, high-risk markets, supplier preference
Acknowledgements
TABLE OF CONTENTS
1. INTRODUCTION ... 3
2. THEORETICAL FRAMEWORK ... 5
2.1 Causes and drivers of trust problems in electronic markets ... 6
2.2 Trust solutions ... 7
2.3 Conceptual framework ... 9
2.4 High-risk vs. low-risk markets ... 11
3 METHODOLOGY ... 14
3.1 Test environment ... 14
3.2 Sample selection ... 14
3.3 Procedure and measurement ... 15
4. RESULTS ... 18 4.1 Descriptive statistics ... 18 4.2 Manipulation check ... 19 4.3 Hypotheses testing ... 20 5. DISCUSSION ... 27 5.1 Theoretical implications ... 27 5.2 Conclusion ... 29 5.3 Managerial implications ... 30
5.4 Limitations and directions for further research ... 31
6. REFERENCES ... 33
7. APPENDICES ... 36
Appendix I: Questionnaire ... 36
1. INTRODUCTION
The evolution of e-commerce is causing radical changes in the business world, as network-based, global online business models develop rapidly and increasingly dominate and develop advantages over traditional business models across markets and industries (Wymbs, 2000). Between 2007 and 2015, the proportion of individuals between 16 and 74 years of age in the European Union who have purchased products or services online has risen from 30% to 53%, while at the other hand, sales revenues through traditional channels continue to drop, forcing offline businesses to adapt to the new market dynamics (Eurostat, 2015). In the group of internet users in the European union that do not yet make online purchases 75% prefers shopping in physical stores in order to better assess the desirability, fit and quality of the offered products and 27% are concerned with payment security and privacy (Eurostat 2015).
secure payment systems and review mechanisms (Ba et al., 2003; Babic et al., 2015; Dholakia, 2005; Komiak & Benbasat, 2004).
Initially, trust problems in digital markets originated in the anonymity and lack of face-to-face contact between the trading parties, the lack of hands-on assessment of the product’s quality and the absence of enforceable legal systems. Although legal systems have caught up to cover online markets and trust solutions have become more advanced and more commonly used, there is still much room for improvement as illustrated by the statistics of Eurostat (2015). The market areas in the e-commerce field that are especially challenged, are market areas where purchases have a high price, are less frequently purchased, have high involvement and quality can only be assessed after purchase (Ba and Pavlou, 2002; Bart et al., 2005).
This research paper analyzes the current literature on the impact of consumer trust problems in online markets and the currently available trust solutions. In particular it will study the three main categories of trust solutions - social, economic and technological - on the extent to which they increase perceived trust and reduce perceived risk in consumers and as such positively impact purchasing intentions. This paper will research the effectiveness of each of these trust solutions compared independently and consecutively it will study the effectiveness of joint presence of multiple trust solutions per supplier on consumers’ purchasing intentions.
2005; Lee et al., 2000), this study will focus on the less studied online service markets. Additionally, this study aims to identify and study those areas in e-commerce that experience lowest trust and highest risk perception of consumers. As each market field has different dynamics, high-risk market areas can be expected to have their own dynamics in trust and risk perception. As such; this study will examine the dynamics of trust perception, risk perception and purchasing intentions of customers, as well as the effect of trust signaling solutions in high-risk service markets compared to low-risk service markets.
Which signals do customers use in online high-risk service markets compared to online low-risk service markets, to determine trust and risk and form purchasing intentions?
Insights on how these most challenged e-commerce markets can increase perceived trust, decrease perceived risk and increase purchasing intention can help firms within these areas to develop, optimize, innovate and grow their business to maturity in current market fields as well as when venturing into new fields of the e-commerce market. As this research question revolves around trust, risk and purchasing intentions of customers, this study chooses to focus on the consumer perspective.
This research paper is structured starting with a theoretical framework containing a literature review and the introduction of the conceptual model and the hypotheses of this study. The theoretical framework is followed by a description of the chosen methods and an overview of the test results. The test results will be interpreted in the discussion section, where there will be distinction between theoretical implications and practical managerial implications. The paper concludes with stating the limitations of this study and proposed directions for future research.
2. THEORETICAL FRAMEWORK
model will be introduced with its hypotheses on the general dynamics between the concepts perceived risk, perceived trust and purchasing intentions and the effects of trust solutions. After that, the literature on high-risk market areas (as compared to low risk market areas) will be discussed, followed by the introduction of hypotheses regarding high-risk market areas.
2.1 Causes and drivers of trust problems in electronic markets
Causes of trust problemsTrust problems in electronic markets have several origins in the dynamics of trading compared to traditional, physical trading. Information asymmetry, a situation where both trading parties do not possess the same information, is one of the major problems in electronic markets according to Ba, Whinston & Zhang (2002). Among the different characteristics of asymmetric information, two are strongly connected with trust in consumers; the anonymous identities of online trading parties due to the lack of personal face-to-face interaction - which in traditional business settings provide basic trust between suppliers and their customers - and the uncertain quality of the offered products and services as online markets do not provide the chance to physically assess the quality of products and services by looking, touching, and feeling (Ba, Whinston & Zhang, 2002).
Drivers of trust and risk perceptions
e-commerce and the Internet in general by means of the customer’s level of experience and familiarity. Personality-oriented trust is seen as the total of a customer’s shopping style characteristics and dispositions to trust online markets. Since experience-based and personality-oriented trust determinants lie more stable within the customer’s dispositions, character and the customers’ lifetime experiences, this study focuses on cognition-based and affect-based trust, as only those can be directly or indirectly managed by a supplier’s operation and methods (Kim et al., 2008) making insights in these areas more useful and valuable for e-commerce businesses and suppliers.
2.2 Trust solutions
In most markets and industries online suppliers and resellers choose to combine several trust signaling solutions. Although there are a great amount of different trust solutions available they can be grouped in the following three types; social trust solutions, economic trust solutions and technical trust solutions. As mentioned in 2.1 suppliers can only affect cognition-based and affect-based trust. Table 1 provides an overview of the drivers of perceived risk and trust and their trust solutions as found in the combined literature.
Social trust solutions
Economic trust solutions
An Economic trust solution is present when a supplier offers - paid or unpaid - insurance or guarantee to protect its customers financially if they might suffer fraud or unexpected damages. Another form of economic trust solutions are escrow services. Escrow services act as a trusted third party in a transaction, providing safe methods to transfer items and payments to both parties – the most commonly used escrow services are Escrow.com and Transpact.com. Escrow services are mostly used in cases of higher order value because of their added cost to a transaction, but since the Internet allows these services to be performed in an automated manner, the costs of these services decrease and they are increasingly used in lower value orders also. An Escrow service collects the agreed payment amount for the purchased product or service from the customer, of which the supplier is notified as a green light to start delivery of its product or service. After delivery by the supplier, the customer confirms the escrow service the product or service is received according to expectations, so that the escrow service can release the payment amount to the supplier (Ba, Whinston & Zhang, 2002).
Technical trust solutions
have found quality assessment of physical goods to be mostly helped by third party mechanisms as compared to intangible goods and services.
Problem Origin Available Solution Solution Type Trust Type Addressed Anonymous
Identities
Trader Reviews Social Solution Affect Based Trust Third Party Seals Technical Solution Affect Based Trust Insurance & Guarantee Economic Solution Cognition Based Trust Escrow Services Economic Solution Cognition Based Trust Digital Verification Technical Solution Cognition Based Trust Uncertain Quality
Product
Product Reviews Social Solution Affect Based Trust Insurance & Guarantee Economic Solution Cognition Based Trust Absence of Legal
Systems
Insurance & Guarantee Economic Solution Cognition Based Trust Escrow Services Economic Solution Cognition Based Trust Table 1. Drivers of Perceived Risk and their Trust Solutions
2.3 Conceptual framework
This study formulates a conceptual model that builds upon the core model of Kim et al. (2008) who show the relationships between perceived trust, perceived risk, purchasing intention and their antecedents, but introduces the presence of social, economic and technical trust solutions as well as the influence of a high-risk market situation compared to a low-risk market situation. Finally, while the study of Kim et al. (2008) is performed in a product market, this study aims to study the relationships between each variable in a services market in order to test the applicability of the model for services.
Perceived risk, trust and purchasing intention
actual behavior or purchasing decision.
Mayer et al. (1995) state that when a customer perceives a supplier’s ability, benevolence and integrity to be sufficient, the customer will develop perceived trust towards this supplier. When perceived risk is present, trust is a key determinant of action, as stated by Luhmann (1988) and there are two ways in which trust mitigates risk effects in online purchasing decisions, a direct effect of trust on purchasing intention and an effect of perceived trust on purchasing intention that is mediated by perceived risk. Regarding the relationship of perceived trust to purchasing intention; perceived risk is proven to influence online purchasing decisions (Antony, Lin & Xu, 2006), for example; in an online market compared to offline a customer has no ability to touch, feel or try a product to assess its quality, where in an offline market this is possible. As this increases the perceived risk directly, it is clear why customers are more hesitant in engaging in online transactions and as such purchasing intentions. These reasons lead to the following hypothesis:
Hypothesis 1: Perceived trust (a) decreases perceived risk and (b) increases purchasing intention, while (c) perceived risk decreases purchasing intention.
Multiple trust solutions per supplier
as much as the trust antecedents in Kim et al. (2008) are of distinct nature, this study expects the different trust solutions to not overlap entirely and as such; that the effect of multiple trust solutions combined will be larger than the effect of a single trust solution alone. This leads to the following hypothesis:
Hypothesis 2: The availability of more trust solutions per supplier will result in (a) higher perceived trust, (b) lower perceived risk and (c) higher purchasing intention towards this supplier.
Figure 1. The Conceptual Model
2.4 High-risk vs. low-risk markets
While much research is done on the dynamics between trust, risk and behavioral intentions (e.g. Kim et al., 2008), less research is done on how these mechanisms differ across market categories. This seems interesting, as studies of Ba and Pavlou (2002), Bart et al. (2005) and Lee et al. (2000) have shown that risk, trust and the effect of trust solutions differs across online market areas. For this reason, this study proposes to investigate the effect of different trust solutions on perceived trust level, perceived risk level and purchasing intention in high-risk service markets compared to low-risk service markets. H3b (+) H3a (-) H3c (-) H2b (-) H1b (+) H1a (-) H1c (-) H2c (+) H2a (+)
High-Risk Market Situation
Identification of high-risk markets
For the identification of high-risk markets, risk is again defined following Kim et al. (2008), focusing on the for e-commerce predominant financial transaction risk types, financial risk, product risk and information risk. The study of Bart et al. (2005) shows that high financial risk coincides directly with high price levels. Ba and Pavlou (2002) explain this by stating that more expensive products and services are believed to pose higher risks on customers because the seller has higher incentives to cheat, and the customer puts more of his resources at stake. Bart et al. (2005) continue stating that the market areas with infrequently purchased, high priced, high-involvement services and products, such as computers and automobiles are perceived to have the relatively high risk levels for customers, while market areas with frequently purchased, low priced, low-involvement items such as books and magazines have relatively low risk levels for consumers. Finally, Lee et al. (2000) have shown that in transaction of products or services where quality can only be assessed after purchase (so called experience goods) or where quality is hard to assess even after several purchase experiences (also known as credence goods), risk to customers is higher. Since services are by definition different from products in the sense that they cannot be physically examined, services are by definition experience goods.
Effects of High-risk markets on purchasing intentions
Concerning customer perception and behavioral intentions in high-risk markets, the study of Bart et al. (2005) shows that high-risk markets coincide with high involvement and high engagement of customers as well as more search behavior of customers as a result to that. Search behavior, according to Kumar, Lang and Peng (2005) represents the customers’ level of search into information and trust signals typically required to make a purchase in a certain market area, service type or product type. In other words, when risk is higher, customers tend to invest more time and energy to ensure them of making the right purchasing decision. This implies that more attention is given to trust signals in high-risk markets.
significant results in customers’ purchasing intentions in high-risk markets due to expensive products being involved in the transactions, while Lee et al. (2002) also found negative ratings to weigh significantly heavier in high-risk markets with experience goods where quality can only be assessed after purchase. The findings that customers show more search behavior in high-risk markets and are also more attentive to trust solutions and more affected by trust solutions, suggests that in high-risk markets perceived trust is lower, which customers compensate by search and trust solutions. If perceived trust is lower, the model of Kim et al. (2008) leads us to the following hypothesis on high-risk market areas:
Hypothesis 3: In high-risk service markets compared to low-risk service markets customers will show (a) lower perceived trust, (b) higher perceived risk and (c) lower
purchasing intention towards suppliers.
that subjective measures do not possess.
Hypothesis 4: In high-risk service markets consumers will (a) have a higher preference for technical and economical trust solutions and (b) a lower preference for social trust solutions than in low-risk service markets.
3 METHODOLOGY
3.1 Test environment
As the goal of this study is to compare the effects of different trust solutions on consumer perception and behavioral intentions this study chooses a quantitative approach. Since merely asking customers their opinion in a survey on different trust increasing solutions across different markets might not give the most realistic representation of their choices in reality, this study makes use of an online test environment exposing each participant to different suppliers to choose from in different market situations, only to be followed by questionnaire items afterwards. All test environments and questionnaire items are run on the participant’s Internet device of choice, mobile, tablet or desktop, at a time and place of their own convenience, in order to keep the environment as natural as possible, or in other words; as close as possible to the participant’s environment in which online purchases would be made.
3.2 Sample selection
larger majority of the population, the insights in this group may be useful to predict how the majority of the population may later react once they get increasingly experienced in online purchasing. This sample choice also prevents having too much influence of variation in demographics like age and geographic location. However, one downside of this sample group is that the study may have limited generalizability to populations of other age groups and countries.
3.3 Procedure and measurement
Research conditionsAt the beginning of the test, all participants were divided randomly over the three research conditions, in which each participant would stay for both the low-risk and the high-risk market situation. In each condition the participants were exposed to three suppliers with respectively; one of the three different trust solutions per supplier (condition 1), a combination of two of the three different trust solutions per supplier (condition 2) and a combination of all three of the three different trust solutions per supplier (condition 3). As such; in research condition 3, with all three of the trust solutions per offer, all three suppliers are technically using the same mechanism. As in condition 3 all three suppliers have the exact same three trust solutions present; to reduce participants’ confusion, each sellers’ trust solutions were listed in a different order to make the similarity less obvious. The use of these research conditions enables this study to investigate the effect of and relative preference for each type of trust solution individually, as well the effect of joint presence of two and three trust solutions at the same time.
Low Risk Market High Risk Market
Procedure and manipulations
After the random selection over the three research conditions, all participants are presented a situational description of a low-risk service market (a low-cost, afternoon activity with friends) and a high-risk service market (a fully organized, costly, 12-month long world trip for 2 persons). After each of the two situational descriptions, the participants are asked to select their preferred option out of three suppliers, where dependent of the research condition, each supplier contains its own (combination of) one, two or three different trust solutions; social trust solution (review of 4/5 stars), economic trust solution (money-back-guarantee) and technical trust solution (verified supplier). The values within each specific trust solution were kept constant and positive; e.g., whenever social trust solution was present, it was shown as a positive social rating of 4/5 stars and never differed in rating score. To avoid any potential undesired influences of name or picture preferences among suppliers, no name, picture or advertisements are shown, but the suppliers are represented by only the icons of it’s available trust solution(s) and nothing more. As such, as the research manipulates only the presence of trust solutions per supplier, this study infers which trust solutions customers use in each market situation, in order to study the influence of- and relative preference of participants for each (combination of) trust solution(s). For example; if a supplier with a social trust solution is preferred over a supplier with a technical trust solution in a certain market situation, this study infers that social trust solutions are more relevant to customers in that market situation.
Type of trust solution Representation in study Social trust solution Review of 4/5 stars Economic trust solution Money-back-guarantee Technical trust solution Verified supplier
Table 3. Representation of trust solutions in research environment Measurement
questionnaire items used to measure perceived trust, perceived risk and purchasing intention with respectively four, four and three items based on Kim et al. (2008) and Paolucci et al. (2009). Reliability analysis shows that each item set has a score on Cronbach’s alpha above the acceptance threshold of 0.7 in both low-risk and high-risk situation’s measurements, respectively 0.75 and 0.83 for perceived trust, 0.77 and 0.81 for perceived risk and 0.77 and 0.77 for purchasing intention. As such it can be concluded that each variable is reliably measured by its questionnaire items.
To make sure the difference in suppliers based on only the trust solutions is noticeable, four questionnaire items for perceived supplier variation (as used by Paolucci et al., 2009) are also implemented. The reliability of these items is also acceptable with a Cronbach’s Alpha of 0.83. In appendix I all questionnaire items are shown as well as an example of the test environment.
Figure 2. Flow Chart of Research Methodology
Control variables
To provide a stronger test of the hypotheses this study included control variables for customer variation; respectively age and experience level with online purchasing. Both variables are based on the research of Kim et al. (2008), where older and less experienced online shoppers were found to have lower trust perceptions and higher risk perceptions in
2. Random selection of participants over research conditions 1,2 and 3 3. Supplier selection in low-risk service market situation: afternoon activity 4. Assessment of chosen supplier for (in)dependent variables
5. Selection of supplier in high-risk service market: organized world trip 6. Assessment of chosen supplier for (in)dependent variables
7. Control variables
online purchasing decisions than their younger, more experienced counterparts. The age of the participant was measured with a single open question, while the participant’s experience level with online purchasing was measured with three 5-point Likert-scale questionnaire items as used in the study of Kim et al. (2008).
4. RESULTS
This chapter elaborates on the results of the study, starting with the descriptive statistics and manipulation checks and followed by the tests of each hypothesis. It gives an overview of the tested hypotheses.
4.1 Descriptive statistics
In total 509 participants completed the questionnaire, distributed across the three research conditions with 201 (condition 1), 174 (condition 2) and 134 (condition 3) participants. The average age of the participants was 23.67 (SD=3.86). Of the participants 352 (69.2%) were female and 157 (26.3%) were men. The average reported number of hours per day spent online was high with 5.38 (SD=3.12), while the average frequency of online purchases over the last 3 months was also relatively high with 6.58 (SD=7.75) product purchases and 3.74 (SD=7.21) service purchases, adding up to a number of 10.31 (SD=11.69) online purchases for both product and services combined. Compared to the frequency of online purchases in Europe (Eurostat, 2017), these numbers are relatively high, as the highest proportion of the population of Europe made only one or two online purchases in the last three months. The frequencies of online purchases per online shopper over the past three months in the populations of respectively Europe and the Netherlands specifically were 16% and 24% 1-3 times, 16% and 23% 3-5 times, 7% and 10% 6-10 times and only 6% and 6% more than 10 times. As was assumed in the sample selection (see methods section) the studied sample is relatively more familiar with the internet and making online purchases (Eurostat, 2017). Finally, the average reported joy in making online purchases in general is 3.85 (SD=0.98) on a 5-point Likert scale with minimum 1 (completely disagree) and maximum 5 (completely agree).
research condition were 5:17 minutes and 60% in condition 1, 5:11 minutes and 61% in condition 2 and a remarkable 6:54 minutes and 41% in condition 3. This is the condition in which in both market situations each supplier had all three of the three different trust solutions, and as such offer the exact same trust mechanisms. Of all 509 participants, 280 (55%) used their mobile device, 216 (42%) used their pc/laptop and 13 (3%) used their tablet to complete the questionnaire. A complete overview of all statistics can be found in Appendix II.
4.2 Manipulation check
Low-risk and high-risk services markets
To check if the intended manipulation of the low-risk versus high-risk service market situation was effective, both market situations are compared across all conditions in a paired samples t-tests (see appendix II). The results show a significant difference between both market situations, however the difference is the opposite direction than the expectation of the manipulation; a decrease in perceived risk level of -0.15 (p > 0.001). Specifying this comparison between market situations for each research condition shows a significant decrease in perceived risk of -0.16 (p > 0.01) in condition 1 (one trust solution per supplier), an insignificant increase of 0.05 in condition 2 (two trust solutions per supplier and a significant decrease of -0.05 in condition 3 (p > 0.05). These results indicate that the manipulation on risk-level does not have the intended effect, however several possible explanations are proposed in the discussion chapter.
Perceived supplier variation
variation in condition 1 and 2, where the suppliers differed in trust solutions, and not in condition 3 where each supplier offers the same trust solutions.
A one-sample t-test on the item perceived supplier variation shows significant difference from test value 1 (no perceived variation) on the 5-point Likert-scale items (M=3.65, SD=0.82, p < 0.001) across research conditions and across low-risk and high-risk service markets. Specifying for the different research conditions the results show the highest perceived supplier variation in condition 1 (M=3.76, SD=0.69) and condition 2 (M=3.66, SD=0.71), while showing slightly lower, but surprisingly still significant supplier variation from test value 1 in condition 3 (M=3.13, SD=0.98) against expectation. These results show that the participants have noticed the difference in trust solutions between suppliers in the conditions where there were differences, but also in the condition where there were no differences between suppliers (each supplier offered all three of the three different trust solutions, due to which the offered the exact same although being listed in a different order).
4.3 Hypotheses testing
Hypothesis 1 assumes that perceived trust decreases perceived risk and increases purchasing intention, while perceived risk decreases purchasing intention. The hypothesis is tested using a regression analysis among the dependent variables in both the low-risk and the high-risk market situation specifically and in both market situations combined. In the data of the low-risk and high-risk service market situations combined, perceived trust decreases perceived risk by .56 and increases purchasing intention by .64, while perceived risk decreases purchasing intention by .27, all significant at a <.001 level.
Specified for the low-risk service market situation perceived trust decreases perceived risk by .50 and increases purchasing intention by .59, while perceived risk decreases purchasing intention by .23, all significant at a <.001 level.
Concluding; the hypothesis is accepted for a, b and c in both high-risk and low-risk market situation. As such the dynamics between the dependent variables prove to be quite consistent across markets.
Variables B SE Standardized
Coefficients Beta
t Sig. R2 H1a Perceived Trust >
Perceived Risk -.56*** .05 -.43 -10.74 .000 .43 H1b Perceived Trust > Purchasing Intention .64*** .04 .55 14.87 .000 .55 H1c Perceived Risk > Purchasing Intention -.27*** .04 -.30 -8.32 .000 .30
Table 4. Regression analysis across risk situations; N=1018; *** = significant at the 0.001 level
As there is significant correlation and regression between the variables perceived trust, perceived risk and purchasing intention a mediation effect can be assumed. Based on the model of Kim et al. (2008) it is expected that perceived risk is the mediator in this relationship, meaning perceived trust influences purchasing intention, both directly and indirectly through perceived risk.
However; repeating this test for the low-risk market situation and high-risk market situation separately, the results show that the indirect effect of the independent variable perceived risk on the dependent variable purchasing intention through the mediator variable perceived risk is significant in both risk situations separately. The condition of significant regression of the independent variable (perceived trust) on the dependent variable (purchasing intention) is met in the low-risk situation and in the high-risk situation, with 0.59 (p<0.001) and 0.54 (p<0.001) respectively. The second condition of significant regression of the independent variable (perceived trust) on the mediator (perceived risk) is met in the low-risk situation and in the high-risk situation, with -0.50 (p<0.001) and -0.54 (p<0.001) respectively. Consecutively, the regression of both the mediator (perceived risk) and the independent variable (perceived trust) together on the dependent variable (purchasing intention), shows that the effect of the independent variable (perceived trust) on the dependent variable purchasing intention decreases to 0.56 (p<0.001) in the low-risk condition and 0.46 (p<0.001) in the high-risk condition, when the mediator variable (perceived risk) is added to the regression model.
Test Variables B SE Standardized Coefficients Beta
t Sig. R2 Step
1 Perceived Trust (IV) > Purchasing Intention (DV) .64*** .04 .55 14.87 .000 .55 Step
2
Perceived Trust (IV) > Perceived Risk (MV)
-.56*** .05 -.43 -10.74 .000 .43
Step 3
Perceived Trust (IV) > Purchasing Intention (DV)
.61*** .05 .52 12.68 .000 .56 Perceived Risk (MV) >
Purchasing Intention (DV)
-.07 .04 -.07 -1.78 .076
Table 5. Mediation analysis across risk situations; N=1018; *** = significant at the 0.001 level
Hypothesis 2 assumes that more trust solutions per supplier will result in higher perceived trust, lower perceived risk and higher purchasing intention in consumers. The hypothesis is tested with a one-way ANOVA, which results show there are significant mean differences between the conditions on each dependent variable. The specific differences are investigated with the Bonferroni Post-Hoc test. This test shows that perceived trust in condition 1 is 0.34 higher than in condition 3 (SE=0.06, p<0.001) and 0.25 higher in condition 2 than in condition 3 (SE=0.06, p=0.001). Perceived risk proves to be 0.22 lower in condition 1 than in condition 3 (SE=0.06, p=0.027) and 0.26 lower in condition 2 than in condition 3 (SE=0.09, p=0.009). Purchasing intention is 0.26 higher in condition 1 than in condition 3 (SE=0.07, p=0.001) and 0.24 higher in condition 2 than in condition 3 (SE=0.07, p=0.004). This means the hypothesis is rejected, as the
hypothesis is significant in the inverse direction on each dependent variable. Condition 1 Condition 2 Condition 3
Perceived Trust 4.083 3.993 3.741,2 Perceived Risk 2.363 2.333 2.591,2 Purchasing Intention 3.643 3.633 3.391,2
Table 6. Means of dependent variables per research condition
Note: Superscripts indicate to which other condition (1,2 or 3) the mean differs based on Bonferroni test.
lower perceived trust and (remarkably) 0.13 lower perceived risk than in low-risk conditions (both significant at the 0.001 level), while purchasing intention does not significantly differ. All reported effects are absolute differences on the 5-point Likert scale. Additionally comparing low-risk and high-risk service market situations for each separate condition shows stronger, more significant results in the case of one trust solution provided than in the conditions with two and three trust solutions provided; -0.21 lower perceived trust (p < 0.001), 0.16 lower perceived risk (p < 0.01) and 0.12 lower purchasing intention (p < 0.05). Across conditions hypothesis 3 is accepted for perceived trust but rejected for perceived risk (in opposite direction) and for purchasing intention (no significant overall result).
Δ Dependent var. All conditions Condition 1 Condition 2 Condition 3 Δ Perceived trust -.15*** -.21*** .04** -.14 Δ Perceived risk -.13*** -.16** .05 -.05* Δ Purchasing intention -.05 -.12* .05 -.03
Table 7. Paired sample T-test of high-risk value minus low-risk value in dependent variables per research condition
Note: * significant at the 0.05, ** significant at the 0.01, *** significant at the 0.001 level
Hypothesis 4 assumes that customers in high-risk service markets will (a) have a higher preference for technical and economical trust solutions and (b) a lower preference for social trust solutions than in low-risk service markets. This hypothesis is checked per research condition 1, 2 and 3.
In the condition with one trust solution per supplier, the high-risk market shows (a) higher preference for technical trust solutions and economical trust solutions and (b) lower preference for social trust solutions (respectively 53.2%, 24.2% and 22.4%) compared to the low-risk service market situation (respectively 20.4%, 6.0% and 73.6%). Social trust solutions are the most frequently preferred trust solutions in the low-risk service market situation, whereas technical trust solutions are most preferred trust solution in the high-risk service market situation. When testing the significance of the difference in preference between both market situations in a in a Chi-square test (see appendix II) a significant difference is found (p<0.001).
than in the low-risk situation (19.5%), confirming part a of the hypothesis. Officially confirming part b of the hypothesis is not possible for this condition, as the social trust solution which this part of the hypothesis addresses appears only in joint presence with technical or economical trust solutions. However it can be observed that the combination of social and technical trust solutions that is most preferred in the low-risk situation (65.6%) is much less preferred in the high-risk situation (28.2%), suggesting that part b of the hypothesis is true here also. When testing the significance of the difference in preference between both market situations in a in a Chi-square test a significant difference is found (p<0.001).
In the condition with all three trust solutions per supplier – where all suppliers as such offer the same trust solutions – surprisingly there is a relative preference nonetheless. There is a relative preference of 38.1% for the supplier showing social trust solution listed first in the low-risk market situation, and a relative preference of 49.3% for the supplier showing the technical trust solution listed first in the high-risk market situation. When testing the significance of the difference in preference between both market situations in a in a Chi-square test (see appendix II) the difference in relative preference prove to be non-significant (p=0.06). This is not surprising, judging that each supplier offers the same combination of trust solutions.
Table 8. Preferred trust solutions per condition in low-risk and high-risk situation
Note 1: numbers in brackets after percentages are the absolute number of the frequencies Note 2: Marked percentages show most preferred supplier per condition and risk level Note 3: marked numbers show highest perceived trust, lowest perceived risk and highest purchasing intention per condition and risk level
Hypothesis Result
1: Perceived trust (a) decreases perceived risk and (b) increases buying intention, while (c) perceived risk decreases purchasing intention.
Accepted for a, b and c and in both the high-risk and low-risk market situation. Perceived risk mediates between perceived trust and purchasing intention.
2: The availability of more trust solutions per supplier will result in (a) higher perceived trust, (b) lower perceived risk and (c) higher purchasing intention towards this supplier.
Rejected; opposite direction for each dependent variable; more trust solutions lead to lower perceived trust, higher perceived risk and lower purchasing intention.
3: In high-risk service markets compared to low-risk service markets customers will show (a) lower perceived trust, (b) higher perceived risk and (c) lower purchasing
intention towards suppliers.
Accepted for perceived trust and rejected for perceived risk (opposite direction) and purchasing intention (not
significant).
4: In high-risk service markets customers will (a) have a higher preference for
technical and economical trust solutions and (b) a lower preference for social trust solutions than in low-risk service markets.
Accepted for a and b in condition 1 and 2, where the
suppliers differed in trust solutions, and rejected for condition 3, where suppliers offered the same trust solutions. Technical trust solutions are most preferred in high-risk markets, while social trust solutions are most preferred in low-risk markets. Table 9. Overview of hypotheses and their results
Low Risk Service High Risk Service Trust Solutions %
5. DISCUSSION
To conclude this study, the results of the four formulated hypotheses are discussed in regard to their theoretical implications with which the research question will be answered. Based on the theoretical implications a set of managerial implications are formulated and finally this chapter will discuss this study’s research limitations and directions for further research.
5.1 Theoretical implications
This study confirms earlier findings of Kim et al. (2008) that perceived trust (a) decreases perceived risk and (b) increases purchasing intention, while (c) perceived risk decreases purchasing intention. The results are significant for a, b and c in both the high-risk and the low-risk market situation. The results also confirm that perceived trust positively affects purchasing intention both directly and indirectly - by reducing perceived risk - proving that perceived risk is a partial mediator of the relationship between perceived trust and purchase intention. This proves that the dynamics as seen in the model of Kim et al. (2008), on which this study’s conceptual model is based, are present in similar fashion in services markets as they are in product markets. Second, the dynamics between the variables prove to be quite consistent across high-risk and low-risk market situations.
increasing similarities increase decision stress in customers, which in turn can cause lower perceived trust, higher perceived risk and lower purchasing intention. This assumption is supported by the amount of time needed to complete the questionnaire, which was longer in condition 3 (6:54 minutes) compared to condition 1 (5:17 minutes) and condition 2 (5:11 minutes). Lee et al. (2002) also provide a possible explanation as they prove that in case of perceived risk customers show more involvement and more (information) searching behavior. While the results of this searching behavior should increase confidence in making the right decision and therefore increase trust, in this instance the only information available showed no differences between suppliers and as such, no information could be found to increase confidence and through that increase trust to lower risk.
mechanisms (e.g. industry specific legal regulations in the travel industry, travel insurances and cancelation insurances or perhaps the firm size of operators in the industry) to limit perceived risk, that are beyond the three online trust solutions as included in this study. The fact that a trust solution could potentially lower perceived risk even when the risk-characteristic of the market situation increases and perceived trust decreases is very encouraging for further investigation on this topic.
Finally, this study proves that in high-risk service markets, customers have a higher preference for technical and economical trust solutions markets and a lower preference for social trust solutions than in low-risk service markets. The most preferred trust solutions in low-risk market situations are social trust solutions, while technical trust solutions are most preferred in high-risk service markets. In other words; in low-risk situations customers tend to trust their peers, while in high-risk situations customers tend to put their trust on authorities and regulations. The initial assumption of this study has been that in high-risk markets objective and institutional instruments (technical and economical trust solutions, e.g. verified supplier or money-back guarantee) are of higher value than subjective and peer-to peer instruments (social trust solutions, e.g. feedback mechanisms and trader reviews), since they possess protective capabilities that subjective measures do not have (they are subjective opinions). The statement of Ba and Pavlou (2002) that in market situations where prices are higher (and therefore customers’ risk is higher) suppliers have more incentive to cheat, supports the idea that trust solutions with actual protective capabilities are preferred over merely subjective and describing trust solutions in high-risk market situations. As there is no specific previous literature to be found that explains the relative preference and importance of different trust solutions across different trust situations, and therefore no specific literature that can support this assumption further, this topic is highly interesting for further research.
5.2 Conclusion
are present in service markets in the same fashion as they are in product markets (as shown by Kim et al., 2008) and in high-risk markets, the same way as in low-risk markets. According to the conceptual model of this study, perceived trust decreases perceived risk and increases buying intention both directly and indirectly (by decreasing perceived risk), while perceived risk decreases purchasing intention.
Furthermore this study has proven that in high-risk market situations technical and economical trust solutions are more preferred, while social trust solutions are less preferred compared to in low-risk market situations. In absolute terms; technical trust solutions are most preferred in high-risk markets, while social trust solutions are most preferred in low-risk markets. The presence of more trust solutions per supplier led to lower perceived trust, higher perceived risk and lower purchasing intention. This result is surprising, however it may be explained as more trust solutions per supplier went
combined with increasing similarities between suppliers, which may have trust aversive nature as it increases decision stress in consumers and makes it harder for customers to use search behavior in order to increase trust. Finally, this study shows high-risk service markets find lower perceived trust in customers as well as (surprisingly) lower perceived risk towards suppliers in, of which the latter is possibly explained by increased search behavior.
5.3 Managerial implications
The findings of this study may assist e-commerce firms and professionals to better incorporate trust solutions in their channels in order to improve sales performance, venture into new markets and grow transaction volume.
the positive effect of perceived trust on purchasing intention works both directly and indirectly (by decreasing perceived risk), this is the most important explanatory variable influencing purchasing intention, for which reason it is very much worthwhile and highly recommendable to invest in increasing perceived trust in customers.
High-risk markets areas are characterized by high prices, high-involvement, high complexity, information asymmetry and experience- or credence attributes. In markets with these characteristics it is recommended to focus most on technical trust solutions - such as authentication and verification of the supplier and third-party seals – and economical trust solutions – such as financial guarantees or secured escrow services - whereas in low-risk service markets (without these characteristics) focus on social trust solutions - such as reviews and ratings from other consumers on the supplier’s reputation and performance - is expected to be most useful.
Finally, as increased similarities between suppliers may cause increased decision stress and increased difficulty to perform information searching behavior, it is evident that a supplier needs to keep innovating to stand out from its competitors in order to facilitate decision-making and be successful. Although this is a challenging mission, as each competitor will increasingly copy and adopt each innovation, solution and measure of competitors that proves to be beneficial, standing out is of critical importance to each firm’s competitive advantage and performance.
5.4 Limitations and directions for further research
intention in order to compare the ratings between the selected and not-selected suppliers. In addition, as there is no specific previous literature that specifically explains the relative preference and importance of different trust solutions across markets with different risk-levels, this general topic is highly interesting for further research.
Also, this research setup did not include a matter of degree within its trust solutions. Each trust solution type (social, economic and technical) was used in a fixed and non-varying manner. It could also be interesting to study the effect of differences in scores for each trust solution. For example, assuming a supplier provides all three trust solutions, social, economic and technical; in which of these trust solutions would variation in score be of most effect on customers’ purchasing intentions (e.g. a lower rating, a less extensive economic trust solution or a less qualitative technical solution)?
Although this study has used a simulated online test-environment to study preference for suppliers based on trust solutions, the participants in this study did not make real purchases. Naturally testing with real-life purchasing environments has more practical challenges, but the generalizability to the real world can be improved even further by testing in situations with actual real purchases.
This study has explored the role and impact of trust solutions in online purchasing decisions in two specific low-risk and high-risk service markets, it would be interesting to study generalizability over several kinds of service markets in low-and high-risk situations. Especially interesting markets to study in further research, are the markets that are most challenged based on currently known literature, which would be high-priced, high-involvement, complex services in which information asymmetry and experience or credence attributes are present. Examples of markets within these characteristics are financial services and advisory services, which quality can only be assessed after purchase or - in case of credence services – are even hard for customers to assess after multiple purchases and experiences due to information asymmetry.
purchasing decisions as it is currently, the most important question is how a supplier can achieve to stand out and maintain competitive advantage in an environment where all competing suppliers can also offer every same trust solution and achieve equally high trust scores.
6. REFERENCES
Antony, S., Lin, Z., & Xu, B. (2006). Determinants of escrow service adoption in consumer-to-consumer online auction market: an experimental study. Decision Support Systems, 42(3), 1889-1900.
Ba, Sulin and Paul A. Pavlou (2002), "Evidence of the Effect of Trust Building Technology in Electronic Markets: Price Premiums and Buyer Behavior," MIS Quarterly, 26(3), 243-268.
Ba, S., Whinston, A. B., & Zhang, H. (2003). Building trust in online auction markets through an economic incentive mechanism. Decision Support Systems, 35(3), 273– 286.
Babić, A., Sotgiu, F., de Valck, K., & Bijmolt, T. H. A. (2015). The Effect of Electronic Word of Mouth on Sales: A Meta-Analytic Review of Platform, Product, and Metric Factors. Journal of Marketing Research, LIII(June), 150817081817000.
Baron, R. M. and Kenny, D. A. (1986) The Moderator-Mediator Variable Distinction in Social Psychological Research – Conceptual, Strategic, and Statistical
Considerations, Journal of Personality and Social Psychology, Vol. 51(6), pp. 1173–1182.
Bart, Y., Shankar, V., Sultan, F., & Urban, G. L. (2005). Are the Drivers and Role of Online Trust the Same for All Web Sites and Consumers? A Large-Scale Exploratory Empirical Study. Journal of Marketing, 69(4), 133–152.
Connelly, B. L., Certo, S. T., Ireland, D., & Reutzel, C. R. (2010). Signaling theory: A review and assessment. Journal of Management, 37(1), 39–67.
Eurostat. (2015) Eurostat News Release: E-Commerce by Individuals, 1 out of 2 persons in the EU purchased online in 2015. Eurostat Press Office, (December), 6.
Eurostat (2017). E-Commerce statistics for Individuals. Eurostat Press Office, Data extracted in February 2017 from http://appsso.eurostat.ec.europa.eu/nui/show.do. Fang, Y., Qureshi, I., Sun, H., McCole, P., Ramsey, E., & Lim, K. H. (2014). Trust,
Satisfaction, And Online Repurchase Intention: The Moderating Role of Perceived Effectiveness of E-commerce Institutional. Misq, 38(2), 407–427.
Fitzsimmons, J. A., & Fitzsimmons, M. J. (2000). New service development: creating memorable experiences. Thousand Oaks: Sage.
Floyd, K., Freling, R., Alhoqail, S., Cho, H. Y., & Freling, T. (2014). How online product reviews affect retail sales: A meta-analysis. Journal of Retailing, 90(2), 217–232.
Kim, D. J., Ferrin, D. L., & Rao, H. R. (2008). A trust-based consumer decision-making model in electronic commerce: The role of trust, perceived risk, and their
antecedents. Decision Support Systems, 44(2), 544–564.
King, R. A., Racherla, P., & Bush, V. D. (2014). What we know and don’t know about online word-of-mouth: A review and synthesis of the literature. Journal of
Interactive Marketing, 28(3), 167–183.
Kollock, P. (1999). The production of trust in online markets. Advances in Group Processes, 16.
Komiak, S. X., & Benbasat, I. (2004). Understanding Customer Trust in Agent-Mediated Electronic Commerce, Web-Mediated Electronic Commerce, and Traditional Commerce. Information and Technology Management, 5(1/2), 181–207. Kumar, N., Lang, K. R., & Peng, Q. (2005). Consumer Search Behavior in Online
Shopping Environments. 38th Hawaii International Conference on System Sciences - 2005, 0(C), 1–10.
Lee, E. J., & Shin, S. Y. (2014). When do consumers buy online product reviews? Effects of review quality, product type, and reviewer’s photo. Computers in Human
Behavior, 31(1), 356–366.
Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of management review, 20(3), 709-734. Mcknight, D. H., & Chervany, N. L. (2014). What Trust Means in E- Commerce
Customer Relationships : An Interdisciplinary Conceptual Typology. International Journal of Electronic Commerce. (February 2015), 37–41.
Nijssen, E. J., Hillebrand, B., Vermeulen, P. A. M., & Kemp, R. G. M. (2006). Exploring product and service innovation similarities and differences. International Journal of Research in Marketing, 23(3), 241–251.
Paolucci, M., Conte, R., Marmo, S., Picascia, S., Quattrociocchi, W., Eymann, T., ... Brito, I. (2009). Social knowledge for eGovernance: Theory and technology of reputation. In eRep booklet, EU sixth framework programme, Priority 7, Citizens and governance in the knowledge based society.
Zhu, F., & Zhang, X. (Michael). (2010). Impact of Online Consumer Reviews on Sales: The Moderating Role of Product and Consumer Characteristics. Journal of
7. APPENDICES
Appendix I: Questionnaire
Step 1: Random selection in research condition
Step 2: Low-risk market situation: Fun Afternoon (Introduction to low-risk service) You are planning a fun afternoon activity with 3 friends. You see your friends every week and want to do something fun. Price: €30/p. Which supplier would you choose? Step 3: Questionnaire A, B, C
Step 4: High-risk market situation: World Trip (Introduction to high-risk service)
You and your partner/best friend are interested in an organized world-trip for 12 months. This is the first time you are making a long trip, and you have saved up for it over the last 3 years. Price: €15.000/p. Which supplier would you choose?
Step 5: Questionnaire A, B, C Step 6: Questionnaire D, E
Questionnaire A. Perceived Trust
Items used: a combination of Kim et al. (2008) and Paolucci et al. (2009) 1. This supplier is trustworthy.
2. This supplier gives the impression that it keeps promises and commitments. 3. I believe that this supplier has my best interests in mind.
4. I believe this supplier is honest. For reference: questions (Kim et al. 2008)
1. This site is trustworthy.
2. This Website vendor gives the impression that it keeps promises and commitments.
3. I believe that this Website vendor has my best interests in mind. For reference: questions (Paolucci et al. 2009)
1. In hoeverre vertrouw je de verkoper van het product?
2. In hoeverre denk je dat de verkoper van de dienst zijn of haar afspraken na zal komen?
Questionnaire B. Perceived Risk
Items used: a combination of Kim et al. (2008) and Paolucci et al. (2009) 1. Purchasing from this supplier has an above average risk level 2. Purchasing from this supplier would involve more financial risk. 3. How would you rate your overall perception of risk from this supplier? 4. How do you rate the risk this supplier will not deliver?
For reference: questions (Kim et al. 2008)
1. Purchasing from this Website would involve more product risk (i.e. not working, defective product) when compared with more traditional ways of shopping. 2. Purchasing from this Website would involve more financial risk (i.e. fraud, hard
to return) when compared with more traditional ways of shopping. 3. How would you rate your overall perception of risk from this site? For reference: questions (Paolucci et al. 2009)
1. In hoeverre denk je dat de verkoper het specifieke product dat op de website was afgebeeld zal leveren (en niet een ander product)?
2. In hoeverre denk je dat het product op tijd geleverd zal worden?
3. In hoeverre denk je dat het product helemaal niet geleverd zal worden? 4. Hoe schat je de kans in dat de kwaliteit van het product tegenvalt? 5. Hoe schat je het risico in van je aankoop?
Questionnaire C. Purchasing intentions
Items used: adaptation to the items used in Kim et al. (2008) (only adaptation: vendor=supplier)
1. I am likely to purchase the services of this supplier. 2. I am likely to recommend this supplier to my friends.
3. I am likely to make another purchase from this supplier if I need the services (replaced: products) that I will buy.
Questionnaire D. Perceived supplier variation
Items used: adaptation to the items as used in Paolucci et al. 2009
1. To what extent did you feel that the suppliers (replaced: sellers) varied in the products they offered?
2. To what extent did you feel that the suppliers (replaced: sellers) varied in their reputation?
3. To what extent did you feel that the suppliers (replaced: sellers) varied in the quality of their descriptions?
4. To what extent did you feel that the suppliers (replaced: sellers) varied in the quality of their product?
Questionnaire E. Control variables 1. Age: ... years
Questionnaire F. Perceived supplier variation
Items used: adaptation to the items as used in Paolucci et al. 2009
(adaptation: “on bidding websites like Ebay” replaced by “on the internet”) 1. How much time do you spend on the internet on an average day? … minutes 2. How often in the past three (replaced: 12) months have you booked services on the
internet? … times
3. How often in the past three (replaced: 12) months have you ordered products on the internet ? … times
4. To what extent do you enjoy purchasing online? 1-5
Example Questionnaire items
Appendix II: Statistics
DESCRIPTIVE STATISTICSTestCondition
Frequency Percent Valid Percent
Cumulative Percent Valid 1 201 39.5 39.5 39.5 2 174 34.2 34.2 73.7 3 134 26.3 26.3 100.0 Total 509 100.0 100.0 DSex
Frequency Percent Valid Percent
Cumulative Percent Valid Man 157 30.8 30.8 30.8 Vrouw 352 69.2 69.2 100.0 Total 509 100.0 100.0 Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
DAge 509 17 55 23.67 3.857
Valid N (listwise) 509
Descriptive Statistics
N Minimum Maximum Mean
Questionnaire statistics Condition 1
Questionnaire statistics Condition 3
MANIPULATION CHECKS
One-Sample Statistics – all conditions
N Mean Std. Deviation
Std. Error Mean
TPVariation 509 3.5614 .82188 .03643
One-Sample Test – all conditions
Test Value = 1 t df Sig. (2-tailed) Mean Difference 95% Confidence Interval of the Difference Lower Upper TPVariation 70.312 508 .000 2.56139 2.4898 2.6330
One-Sample Statistics – per condition
TestCondition N Mean Std. Deviation
Std. Error Mean
2 TPVariation 174 3.6580 .70773 .05365
3 TPVariation 134 3.1269 .97610 .08432
One-Sample Test – per condition
TestCondition Test Value = 1 t df Sig. (2-tailed) Mean Difference 95% Confidence Interval of the Difference Lower Upper 1 TPVariation 57.135 200 .000 2.76741 2.6719 2.8629 2 TPVariation 49.541 173 .000 2.65805 2.5521 2.7639 3 TPVariation 25.223 133 .000 2.12687 1.9601 2.2937
Paired Samples Test – all conditions overview
All conditions Condition 1 Condition 2 Condition 3
Pair 1 TLowTrust - THighTrust -.15029*** -.21020*** .03667** -.13789 Pair 2 TLowRisk - THighRisk -.13016*** -.15672** .05277 -.05266* Pair 3 TLowBuyInt - THighBuyInt -.04781 -.11774* .04601 -.02702 Note: * Significant at the 0.05, ** Significant at the 0.01, *** significant at the 0.001
Paired Samples Statistics – all conditions
Paired Samples Test – all conditions Paired Differences t df Sig. (2-tailed) Mean Std. Deviati on Std. Error Mean 95% Confidence Interval of the Difference Lower Upper Pair 1 TLowTrust - THighTrust -.15029 .54294 .02407 -.19758 -.10301 -6.245 508 .000 Pair 2 TLowRisk - THighRisk -.13016 .73182 .03244 -.19389 -.06643 -4.013 508 .000 Pair 3 TLowBuyInt - THighBuyInt -.04781 .65250 .02892 -.10463 .00901 -1.653 508 .099
Paired Samples Statistics – condition 1
Mean N Std. Deviation Std. Error Mean Pair 1 TLowTrust 3.8706 201 .46509 .03280 THighTrust 4.0808 201 .51901 .03661 Pair 2 TLowRisk 2.2077 201 .65771 .04639 THighRisk 2.3644 201 .77012 .05432 Pair 3 TLowBuyInt 3.5307 201 .65002 .04585 THighBuyInt 3.6484 201 .63613 .04487
Paired Samples Test – condition 1
Pair 3 TLowBuyInt - THighBuyInt
-.11774 .65231 .04601 -.20847 -.02702 -2.559 200 .011
Paired Samples Statistics – condition 2
Mean N Std. Deviation Std. Error Mean Pair 1 TLowTrust 3.8635 174 .45899 .03480 THighTrust 3.9899 174 .57423 .04353 Pair 2 TLowRisk 2.2457 174 .68556 .05197 THighRisk 2.3261 174 .73242 .05552 Pair 3 TLowBuyInt 3.6398 174 .58611 .04443 THighBuyInt 3.6284 174 .63797 .04836
Paired Samples Test – condition 2
Paired Differences t df Sig. (2-tailed) Mean Std. Deviati on Std. Error Mean 95% Confidence Interval of the Difference Lower Upper Pair 1 TLowTrust - THighTrust -.12644 .52451 .03976 -.20492 -.04795 -3.180 173 .002 Pair 2 TLowRisk - THighRisk -.08046 .69369 .05259 -.18426 .02334 -1.530 173 .128 Pair 3 TLowBuyInt - THighBuyInt .01149 .66753 .05061 -.08839 .11138 .227 173 .821
Paired Samples Statistics – condition 3
THighRisk 2.5858 134 .76710 .06627
Pair 3 TLowBuyInt 3.3657 134 .68260 .05897
THighBuyInt 3.3856 134 .67520 .05833
Paired Samples Test – condition 3
HYPOTHESES TESTING HYPOTHESIS 1
Regression with purchasing intention and perceived risk across risk situations
Variables B SE Standardized
Coefficients Beta
t Sig. R2 H1a Perceived Trust,
Perceived Risk -.56*** .05 -.43 -10.74 .000 .43 H1b Perceived Trust, Purchasing Intention .64*** .04 .55 14.87 .000 .55 H1c Perceived Risk, Purchasing Intention -.27*** .04 -.30 -8.32 .000 .30
N=1018; *** = significant at the 0.001 level
Regression with purchasing intention and perceived risk in low-risk situation
Variables B SE Standardized
Coefficients Beta
t Sig. R2
H1a Perceived Trust,
Perceived Risk -.50*** .05 -.38 -9.12 .000 .38 H1b Perceived Trust, Purchasing Intention .59*** .05 .47 12.14 .000 .47 H1c Perceived Risk, Purchasing Intention -.23*** .04 -.25 -5.73 .000 .25 N=509 ; *** = significant at the 0.001 level
Regression with purchasing intention and perceived risk in high-risk situation
Variables B SE Standardized
Coefficients Beta
t Sig. R2 H1a Perceived Trust,
Mediation Effect across risk situations
Test Variables B SE Standardized
Coefficients Beta
t Sig. R2 Step 1 Perceived Trust >
Purchasing Intention
.64*** .04 .55 14.87 .000 .55
Step 2 Perceived Trust > Perceived Risk
-.56*** .05 -.43 -10.74 .000 .43
Step 3 Perceived Trust > Purchasing Intention
.61*** .05 .52 12.68 .000 .56
Perceived Risk > Purchasing Intention
-.07 .04 -.07 -1.78 .076
N=1018; *** = significant at the 0.001 level Mediation Effect in low-risk situation
Test Variables B SE Standardized
Coefficients Beta t Sig. R 2 Step 1 Perceived Trust >
Purchasing Intention .59*** .05 .47 12.14 .000 .47 Step 2 Perceived Trust >
Perceived Risk
-.50*** .05 -.38 -9.12 .000 .38
Step 3 Perceived Trust >
Purchasing Intention .56*** .05 .45 10.57 .000 .48 Perceived Risk >
Purchasing Intention
-.08 .04 -.08 -1.90 .058
N=509; *** = significant at the 0.001 level
Mediation Effect in high-risk situation
Test Variables B SE Standardized
Coefficients Beta
t Sig. R2
Step 1 Perceived Trust > Purchasing Intention
.54*** .04 .48 12.39 .000 .48
Step 2 Perceived Trust > Perceived Risk
-.54*** .05 -.41 -10.11 .000 .41
Step 3 Perceived Trust >
Purchasing Intention .46*** .05 .41 9.74 .000 .51
Perceived Risk > Purchasing Intention
-.15*** .04 -.18 -4.27 .000
Indirect effect of Trust through mediator Risk Sobel Test across risk situations
Sobel Test in low-risk situation
Correlation Between Dependent Variables across both risk situations
Perceived Trust Perceived Risk Purchasing Intention Perceived Trust 1 Perceived Risk -.38** 1 Purchasing Intention .48** -.30** 1
**. Correlation is significant at the 0.01 level (2-tailed). N=1018
Correlation Between Dependent Variables in Low and High Risk Service Situation
Perceived Trust Perceived Risk Purchasin g Intention Perceived Trust 1 -.41** .48** Perceived Risk -.38** 1 -.35** Purchasing Intention .47** -.25** 1
** Correlation is significant at the 0.01 level (2-tailed). N=509. Correlation above the diagonal is high risk service situation; below low risk situation.
HYPOTHESIS 2
