It’s complicated: explaining the relationship between trust, distrust, and ambivalence in online transaction relationships using polynomial regression analysis and response surface analysis

Abstract

Trust and distrust are considered crucial elements affecting online relationships – particularly those involving electronic transactions. Although some studies propose that they are distinct, others claim that they are merely opposite ends of one continuum. Further adding to the debate is the possibility of ambivalence, a topic that has not been examined in electronic transaction relationships. Unfortunately, current models of trust and distrust have limitations that impede explanations of how – or even if – ambivalence is generated by feelings of trust and distrust and how these two constructs can best coexist. We thus propose a hybrid model which considers the limitations and strengths of previous models. Namely, we posit that trust and distrust can coexist as separate components with related continua. We use polynomial regression analysis (PRA) and response surface analysis (RSA) to test these complex relationships. Using an empirical study of online consumer behaviour with 521 experienced online consumers, strong empirical validation is found for the model. We examine the effects of ambivalence on the truster’s intentions towards a website and find a small positive effect which increases such intentions. PRA and RSA confirm that trust and distrust are most likely separate components – not opposite ends of a continuum – with related continua. The continua within the subconstructs of trust and distrust likely have more complex and interesting relationships than have been considered previously. These findings lead to interesting future research opportunities on trust, distrust and ambivalence using advanced techniques such as PRA and RSA.

Keywords:

• trust
• distrust
• ambivalence
• e-commerce
• electronic commerce
• attitudes

Electronic supplementary material

The online version of this article (doi:10.1057/s41303-016-0027-9) contains supplementary material, which is available to authorized users.

Electronic supplementary material

The online version of this article (doi:10.1057/s41303-016-0027-9) contains supplementary material, which is available to authorized users.

Notes

1 Trusting beliefs is composed of three subconstructs, namely benevolence, competence and integrity (McKnight et al., 2002). Benevolence is exhibited by an organisation that cares about the individual and attempts to act in his/her best interests. Competence is exhibited by an organisation that has the capability to perform the desired behaviour. Finally, a firm with high integrity is honest in its interactions with the individual and will fulfil its promises to him/her.

2 This body of literature defines feelings as the emotional response and attachments which an individual ascribes to other individuals or objects (Kachadourian et al, 2005); beliefs are the logically held information regarding the characteristics of other individuals or objects (Kachadourian et al, 2005). Feelings thus involve affect, whereas beliefs involve cognition; thus, these concepts can also be referred to as affective beliefs and cognitive beliefs (Trafimow & Sheeran, 1998). Finally, behaviours are actions that are performed by an individual (Kachadourian et al, 2005) that are intended to reflect the held feelings and beliefs of the individual.

3 For example, an online consumer can have trust in the TurboTax® website (the trustee) and believe that it has competent advice to assist consumers in the completion and filing of an accurate tax return. However, a consumer may simultaneously distrust advice from the website regarding money management software, largely because the company sells a product in that category. Thus, the proper response to whether an online consumer trusts the organisation should not be ‘yes’ or ‘no’ but ‘to do what?’ (Hardin, 1993). In complex relationships, which only magnify when introducing organisations as the object of trust, it is most important to refer to specifics to understand whether a consumer trusts an organisation via its website. For example, consumers are likely to trust that online orders to Apple’s iTunes online store will be conducted without risking their future credit card transactions on other websites. However, they might also believe that their shopping history on the iTunes store might result in future target marketing. The various facets that make up a relationship allow trust and distrust to coexist, and thus support the bidimensional model of trust and distrust.

4 For example, if an online consumer believes that Amazon.com will ship a purchased item in a timely manner, the consumer cannot also believe the item will not be shipped in a timely manner. Information that is used to form the positive or negative expectations that will lead to trust or distrust cannot be inherently contradictory: Either the information will lead to a positive expectation that the trustee will perform some exact behaviour (e.g., ship an item), or it will lead to a negative expectation (e.g., not ship the item).

5 For example, consumers of Delta Airlines will value information regarding Delta Airlines from the official website differently than information posted on websites such as DeltaReallySucks.com or other travel review websites.

6 For example, a consumer who uses YouSendIt.com for the transmission of files to various colleagues around the globe might believe that the organisation is able to receive and host these files. By using the service, the consumer accepts this belief and disregards the potential negative belief that the organisation is not able to receive and host the same files. Ultimately, the consumer either believes that the organisation is competent or incompetent in relation to this action. The various cues that are present on the website can be used to form both trust and distrust towards YouSendIt.com.

7 Ability is defined by the subdimensions competence and incompetence. Ability forms the assessment of the seller’s proficiency (or lack thereof) to complete a given task (i.e., competence and incompetence).

8 Orientation is defined by the subdimensions benevolence and malevolence. Orientation is the idea that the seller intends to do harm or good to the buyer.

9 Dependability is defined by the subdimensions integrity and deceit. Dependability is the notion that a buyer expects a seller to adhere to a set of guiding principles of being honest, or expects the seller to deceive him or her.

10 Of the subjects, 59% were male and 41% female. The average age was 28.1 years, with a standard deviation of 5.6. The respondents reported an average of 7.1 completed collegiate semesters, with a standard deviation of 1.9.

11 The use of such participants for this type of study follows the precedents set forth in past e-commerce research (Dinev & Hart, 2006; Pavlou & Fygenson, 2006; Lowry et al, 2008; Parboteeah et al, 2009; Lowry et al, 2012). Participants in this young but educated demographic had extensive experience with e-commerce, the Internet, and various computing technologies – particularly as users and consumers – which qualifies them as excellent targets for this study.

12 Because of the nature of formative measures, reliability checks cannot be reasonably made for formative measures (Diamantopoulos & Winklhofer, 2001). To establish reliability, which refers to the degree to which a scale yields consistent and stable measures over time (Straub, 1989), PLS computes a composite reliability score as part of its integrated model analysis. This score is a more accurate measurement of reliability than Cronbach’s α because it does not assume that loadings or error terms of the items to be equal (Chin et al, 2003). However, as a conservative check, Cronbach’s α can also be used as a basis of comparison (Fornell & Larcker, 1981; Nunnally & Bernstein, 1994).

13 The height of the dependent variable (z) is represented by the graphical display, and augmented with colour. Warmer colours (those near the red spectrum) represent higher scores for the dependent variable, while the cooler colours (those near the blue spectrum) represent lower scores.