Trust the Facts: The Impact of Reviews’ Written Style and Subject-Focus on Peer-to-Peer Accommodation Consumption

Peer-to-Peer (P2P) Accommodation

The sharing economy, which is established on the concept of sharing resources on demand rather than possessing them, has been on the rise and thriving in response to the global recession and resource depletion (Botsman & Rogers, 2010). As an emerging business model, the sharing economy commercial service has become increasingly popular and allows people to share resources in many innovative ways (Cohen & Kietzmann, 2014). Among the various sharing economy business types, the P2P accommodation platform (e.g., Airbnb and Vrbo) is one of the most outstanding and successful examples that well represents the sharing economy concept (Mauri et al., 2018). The P2P accommodation platform refers to such an online channel through which people can share their unoccupied room(s) or entire house with others, particularly travelers, for a relatively short term (Shin et al., 2020). The P2P accommodation option has dramatically impacted and even changed the traditional hospitality business model (e.g., hotels) and people’s travel behaviors (Guttentag, 2015).

Along with the on-going development of the P2P accommodation business, researchers have also paid attention to this new form of accommodation practice, particularly interested in what drives people to choose P2P accommodation (De Canio et al., 2020). Previous P2P accommodation literature has indicated a range of driving factors, such as economic benefits (e.g., lower price compared with hotels), homely facilities (e.g., fully equipped kitchen and other home appliances), authentic and novel experiences, social benefits, sustainability value, and positive emotional benefits (Guttentag, 2015, 2018; Shin et al., 2020; Tussyadiah & Pesonen, 2016). In the process, researchers have acknowledged that trust is a very important construct driving and influencing prospective customers’ decision-making regarding choosing P2P accommodation (Ye et al., 2021). While P2P accommodation may offer many benefits and values to customers, the lack of trust would deter customers’ participation in P2P accommodation consumption (Hawlitschek et al., 2016), particularly as it is a relatively new business model compared with the traditional hotel accommodation option. Therefore, a thorough and in-depth understanding of customers’ trust issues and their decision-making processes in the P2P accommodation context is pivotal for both practitioners and researchers.

Conceptualizing Trust in the P2P Accommodation Context

Trust has been extensively researched in various e-commerce settings, such as customer-to-customer as well as business-to-consumer transactions (Kim & Peterson, 2017; Luo et al., 2021). Being a prerequisite for reciprocal transactions, customers’ trust in the product/service or the seller/service provider reduces their perceptions of consumption risk during the decision-making process (Shulga et al., 2021) and subsequently drives purchase decisions (Matute et al., 2016). Previous literature confirms that trust is a crucial psychological mechanism explaining the impact of online reviews on customers’ purchase decisions (Matute et al., 2016). Trust is particularly important to investigate in the P2P accommodation context, where two parties of “strangers” (i.e., host and guest) participate in a transaction to share the partial or entire usage of a “home” through the online platform (Park & Tussyadiah, 2020; Ye et al., 2021). Thus, prospective customers tend to be hesitant to participate in a sharing economy due to a lack of trust (Hawlitschek et al., 2016). Specifically, in the context of P2P accommodation, customers have vulnerability and expectations toward the property quality and the host’s behavior (Park & Tussyadiah, 2020). As such, Hawlitschek et al. (2016) propose that there are three targets of trust in the sharing economy context: product (i.e., accommodation property), peer (i.e., host or other guests), and platform. Previous research suggests that trust in the property and the host play a critical role in the booking process (Park & Tussyadiah, 2020). Given the high level of uncertainty for service quality in the context of P2P accommodation, potential guests seek to obtain high-quality information that aids in reducing their perceptions of uncertainty about the property and the host (Ye et al., 2021).

As trust is a multi-dimensional latent construct (Hawlitschek et al., 2016), it is important to identify the specific dimensions of trust (i.e., competence, integrity, and benevolence) for precisely grasping trust in the P2P accommodation context and its effect on booking intention. Following McKnight and Chervany’s (2001) study, we categorize trust into three conceptual dimensions in this research: competence-, integrity-, and benevolence-based trust. Competence-based trust is derived from the performance of the individuals/objects, referring to one’s judgment of the targeted individual/object’s ability (or inability) to meet or exceed a certain set of performance expectations (McKnight & Chervany, 2001). Integrity-based trust, meanwhile, focuses on judgments of the moral uprightness of the individual or the agent: the belief that the other party will honor their commitments and act in consistency with their promises (McKnight & Chervany, 2001). Finally, benevolence-based trust revolves around the judgments of how much a target group can be expected to exhibit care, reflecting one’s belief that others care about one and will put one’s interest ahead of their own (McKnight & Chervany, 2001). With each making a distinct contribution to the integral perception of trust, these three dimensions of trust may have varying effects on customers’ evaluations of different subjects.

Customer trust can be classified into two categories based on subject association: trust toward the product/service and trust toward the seller/service provider (Mayer et al., 1995). Previous research suggests that product/service-associated customer trust mainly revolves around the theoretical dimensions of competence and integrity, respectively addressing how well the product performs in functional aspects (Lee & Lee, 2005) and whether it consistently delivers superior value as promised (Montecchi et al., 2019). Prior studies find that the competence and integrity dimensions of product-associated trust can influence customers’ quality perceptions and purchase intentions toward food products (Janssen & Hamm, 2012; Wulandari & Suryaningsih, 2019). Addressing the competence-based dimension of product-associated trust, Hawlitschek et al. (2016) reveal that customers’ trust in the performance of a P2P rental service positively affects their consumption intention. Integrity-based product-associated trust, as manifested in customers’ perception that the product can deliver its functional promise, can reduce customers’ perceived risks and enhance their purchase intentions toward financial services (Montecchi et al., 2019). Seller/service provider-associated customer trust, on the other hand, can be either competence-, integrity- and/or benevolence-based (Xu et al., 2016). Cao et al. (2014) found that, when customers purchase products associated with high levels of consumption risk (e.g., dairy products with higher levels of food safety concerns and risks), all three dimensions of trust in the seller (i.e., competence-, integrity-, and benevolence-based trust) can positively affect customers’ repurchase intentions. Furthermore, Xu et al. (2016) disclose that, compared with integrity- and benevolence-based trust, competence-based trust in a financial service provider substantially impacts the buyer’s purchase behavior.

In light of the literature reviewed above, we conceptualize trust in the P2P accommodation context as a multi-subject focus and multi-dimensional construct delineating competence-, integrity- and/or benevolence-based trust toward either the P2P property or the P2P host (as seen in Table 1). Following previous research (Hawlitschek et al., 2016; Janssen & Hamm, 2012; Montecchi et al., 2019; Wulandari & Suryaningsih, 2019), customer trust toward the P2P property consists of competence-based and integrity-based trust. Specifically, in line with Hawlitschek et al. (2016), competence-based trust toward the P2P property refers to how a guest believes that a property has the essential and adequate facilities/amenities to ensure excellent service performance. Consistent with Montecchi et al. (2019), integrity-based trust toward the P2P property refers to how a guest believes a property has a consistent and trustworthy quality record that lives up to its service promises.

OPEN IN VIEWER

Table 1.

Conceptualization of Trust in the P2P Accommodation Context.

Subject Focus of Trust	Theoretical Sub-Dimensions of Trust
	Competence	Integrity	Benevolence
P2P Host	The degree to which a guest believes that a host is knowledgeable and competent at hosting.	The degree to which a guest believes that a host is dependable and reliable.	The degree to which a guest believes a host is concerned with the guest’s welfare and puts the guest’s interest ahead of the host’s self-interest.
P2P Property	The degree to which a guest believes that a property has the essential and competent facilities and amenities.	The degree to which a guest believes that a property has a consistent quality.	N/A

On the other hand, trust toward the P2P host comprises competence-, integrity- and benevolence-based sub-dimensions. Along with Hawlitschek et al. (2016) and Ma et al. (2017), competence-based trust toward the P2P host refers to how well a guest believes that a host is competent and knowledgeable when performing service duties as a host. In line with Hawlitschek et al. (2016), integrity-based trust toward the P2P host refers to how well a guest believes that a host will be honest and reliable in fulfilling their service promises. Benevolence-based trust toward the P2P host is about to what extent a guest believes that a host cares about the guest’s welfare (Hawlitschek et al., 2016; Ma et al., 2017).

The existing literature indicates that online reviews can influence customers’ trust perceptions toward P2P properties and hosts, driving booking intention. Extending such a line of theoretical inquiry, the current research examines how and why the written style (i.e., fact-based vs. emotion-based) and subject focus of online reviews (i.e., property-focused vs. host-focused) influence customers’ booking intention via competence-, integrity-, and benevolence-based trust perceptions toward P2P properties and hosts. In the next section, literature related to the written style of online reviews is discussed, focusing on differentiating fact-based versus emotion-based reviews.

Written Style of Online Reviews: Fact-Based Versus Emotion-Based Reviews

Depending on the written style, customer reviews can be either primarily fact-based or emotion-based. In the context of P2P accommodation, fact-based reviews focus on describing specific attributes and features of the P2P property or the P2P host (Shin et al., 2019). Meanwhile, emotion-based reviews may focus on using emotional expressions to illustrate one’s subjective assessment of the P2P property or the P2P host (Dong et al., 2019; Shin et al., 2019). Previous studies have found that fact-based objective information is better received by customers and has a more substantial effect on customer judgment than emotion-based subjective information (Filieri et al., 2018). For instance, research finds that when evaluating hotels, customers perceive overly emotional reviews as less trustworthy (Baker & Kim, 2019) and less helpful (Bronner & De Hoog, 2011) than fact-based reviews.

Under the dual-process theory (Wason & Evans, 1974), we argue that fact-based versus emotion-based reviews can lead to different customer responses toward online reviews of P2P accommodation, given the distinct mental processes elicited. Dual-process theory (Wason & Evans, 1974) proposes two distinct types of information processing: Type I processing, characterized as automatic, spontaneous, and emotional; and Type II processing, which is systematic, deliberative, and cognitive (Wason & Evans, 1974). Compared with Type I processing, Type II processing requires much more mental effort to systematically analyze the information (Wason & Evans, 1974). Moreover, the two processing types are mutually exclusive in nature, with the occurrence of one processing type typically superseding the other (Cheyne et al., 2008). For instance, the notion of “don’t get carried away by feelings” suggests that the emotional style of Type I processing may impede the deliberate reasoning associated with Type II processing.

Utilizing the dual-process theory, we infer that fact-based and emotion-based reviews will demonstrate different influences on customer trust toward P2P properties and hosts. Cognitive elaborations are found to be the core processes shaping customer trust along with the conceptual bases of competence, integrity, and/or benevolence (McKnight & Chervany, 2001). When going through fact-based reviews, people tend to follow Type II processing to concretely elaborate the factual information disclosed (Filieri et al., 2018), which facilitates the formation of trust judgments. Contrarily, the processing of emotion-based reviews is likely to follow Type I processing, which inhibits logical analysis and cognitive elaborations, distracting individuals from forming cognitively-based trust judgments (Dong et al., 2019). In line with such theorization, previous research suggests that fact-based reviews are perceived as more trustworthy and informative compared with emotion-based reviews, which further leads to higher purchase intentions (Carbonell et al., 2019). On the other hand, emotion-based reviews focus on expressing personal emotions and often leave a vague impression of the product/service or the seller/service provider (Dong et al., 2019). With a heavy load of emotional expressions blended in, emotion-based reviews make it difficult for customers to derive trust judgments based on objective information (Dong et al., 2019). For example, Grabner-Kräuter and Waiguny (2015) found that emotional (vs. factual) reviews could lower customers’ attitudes toward their physicians.

Thus, in the context of P2P accommodation, prospective customers may value fact-based reviews more than emotion-based reviews during information search. Therefore, we posit that fact-based (vs. emotion-based) reviews will be more effective in helping customers form trust judgments toward the P2P property and the host, which further leads to higher levels of booking intention. Specifically, we argue that fact-based (vs. emotion-based) reviews will enhance booking intention due to the indirect effect via competence-, integrity-, and benevolence-based trust. Fact-based (vs. emotion-based) reviews feature concrete information about the property/host. Therefore, fact-based (vs. emotion-based) reviews may increase customers’ trust in the property for providing adequate facilities and amenities to fulfill its function or enhance customers’ trust in the host for being knowledgeable and skillful enough to perform their part of the transaction (Park & Tussyadiah, 2020). The enhanced competence-based trust would minimize perceived transaction risk and increase customers’ intent to book P2P accommodation (Sthapit & Björk, 2019; Vázquez-Casielles et al., 2013). As such, we propose that competence-based trust may mediate the relationship between fact-based reviews and booking intention. Integrity-based trust also plays a significant role in online transactions (Amatulli et al., 2021; Ponte et al., 2015). Previous literature indicates that integrity-based trust is a key factor in building successful relationships (Sthapit & Björk, 2019): If customers believe the property or the host show integrity based on the factual information they have read (e.g., consistent quality of the property or a reliable host), integrity-related trust would emerge—and customers tend to book such P2P accommodation (Amatulli et al., 2021; Tussyadiah & Park, 2018) due to enhanced integrity-based trust. Lastly, benevolence-based trust is associated with the host’s genuine care and kindness to others and has a strong impact on the customer–host connection (Park & Tussyadiah, 2020). Prior research suggests that customers may develop benevolence-based trust toward the host when they discover the considerate amenities prepared by the host (Park & Tussyadiah, 2020), and such a benevolence-based trust in the host will further increase customers’ willingness to use the P2P accommodation (Cheng et al., 2019). The host-focused, fact-based reviews usually provide information regarding the amenities provided or actions taken by the host, which could trigger such a benevolence-based trust in the host and then further enhance customers’ booking intention. Taken together, we propose the hypotheses as follows:

Online Reviews’ Subject Focus: Host-Focused Versus Property-Focused Fact-Based Reviews and the Moderating Role of Customers’ Gender

Not all fact-based reviews, however, will demonstrate an equal impact on customers’ P2P accommodation booking intention. Before making a booking decision, customers gather information about the host and the property to establish trust (Nisar et al., 2019). Correspondingly, fact-based reviews can be either host-focused or property-focused. Host-focused fact-based reviews illustrate what hosts provide for their guests in detail, with narrative descriptions enabling potential customers to make inferences about the host (Tussyadiah & Park, 2018). Such inferences can be regarding the host’s competence in service performance, personal quality of integrity, and benevolence characteristics such as a warm and caring nature.

Meanwhile, property-focused fact-based reviews may describe how the property is furnished, assisting potential customers in anticipating the property’s service attributes and quality performance (e.g., service excellence and quality record). Although both host-focused and property-focused fact-based reviews are informative in helping customers build trust in the accommodation (Nisar et al., 2019), male and female customers may respond differently toward the two types of fact-based reviews. Gender differences in information processing have been well recognized in previous literature (e.g., Kiecker et al., 2000; Sun et al., 2021). The current research particularly focuses on contrasting males’ versus females’ responses toward host-focused and property-focused fact-based reviews. Previous research suggests that males tend to be independent and driven by individualistic goals, while females place a higher value on interpersonal ties (Kiecker et al., 2000; Sun et al., 2021). Such gender differences also influence how males and females process information (Meyers-Levy, 1986). For instance, unlike males, who are more likely to be object-focused, task-specific, and self-focused, females tend to be other-focused and rely more on their social interactions with others to form cognitive judgments (Mulac et al., 2001). Gender differences are also found in social influence processes (Awad & Ragowsky, 2008). Human cues in the communication medium can influence individuals’ trust and behaviors (Park et al., 2021), and males and females can be affected by the same human cue in different ways (Meyers-Levy, 1986). Specifically, females tend to be more responsive to humanistic elements and socio-emotional traits in information communications (Awad & Ragowsky, 2008). As such, females (as opposed to males) are more likely to be influenced by online reviews that contain human-related information (Awad & Ragowsky, 2008). In addition, males and females follow different approaches when establishing social trust (Sun et al., 2007). Compared with males, females rely on social interactional cues when forming trust judgments (Sun et al., 2007). For example, Awad and Ragowsky’s study (2008) shows that females trust online retailers more when they perceive retailers’ social presence in business transactions and communication processes.

Given that females rely more on social information to derive trust judgments and are more predisposed to trust others, we theorize that female customers will be more influenced by host-focused (vs. property-focused) fact-based reviews when making trust judgments and booking decisions within the P2P accommodation context. In other words, among reviews following a fact-based written style, those that are host-focused (vs. property-focused) will lead to higher booking intention among female customers. Trust perceptions mediate such effects. Meanwhile, among male customers who are less reliant on social information to form trust judgments and consumption evaluations, online reviews’ subject-focus may not influence their trust perceptions and booking intention for P2P accommodation. Therefore, we propose the following hypotheses:

Study Overview

Guided by the proposed conceptual model as demonstrated in Figure 1, a series of three scenario-based studies were conducted to test the proposed hypotheses. Study 1 examined the impact of different written styles of property-focused online reviews (i.e., fact-based vs. emotion-based) on customers’ P2P accommodation booking intention (to test H₁) and the underlying mechanism of various property-associated trust dimensions (competence-based and integrity-based, to test H_2a and H_2b). Study 2 examined host-focused online reviews (fact-based vs. emotion-based, to test H₃) and the underlying mechanism of host-associated trust dimensions (competence-based, integrity-based, and benevolence-based, to test H_4a, H_4b, and H_4c). Lastly, Study 3 investigated the gender-moderated impact of review subject focus (i.e., property-focused vs. host-focused, to test H₅) on customers’ P2P accommodation booking intention and the related mediation effects of trust (to test H₆).

OPEN IN VIEWER

Fig. 1.

Conceptual Model.

Online reviews were extracted from a major P2P accommodation platform (Airbnb) popular in the United States, and further refined to develop scenarios of different types of online reviews (fact-based vs. emotion-based; property-focused vs. host-focused; see Appendix). Study participants were recruited from the online research panels of MTurk.com. In each study, participants were first told to imagine searching for P2P accommodation information for an upcoming trip and then directed to read an online review depending on their assigned scenario. Afterward, participants answered a short survey questionnaire based on the scenarios they had read.

Study 1: Property-Focused Online Reviews (Fact-Based vs. Emotion-Based)

Data Analysis and Results

Before testing the proposed hypotheses, exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were first conducted to validate the multidimensional trust construct.

EFA Results

EFA was conducted using principal component analysis to identify the two dimensions of trust in the property (i.e., competence-based and integrity-based). The two factors together explained 74.61% of the variances. Competence-based trust and integrity-based trust accounted for 30.03% and 44.58%, respectively, of the variances. Detailed results are provided in Table 2.

OPEN IN VIEWER

Table 2.

EFA and CFA Results of Trust in the Property.

Factors/Items	Component		Standardized Factor Loading	α	CR	AVE
	F1	F2
Competence-Based Trust
COM1. This house will meet my requirements.	.33	.85	.82	.81	.79	.65
COM2. This house will be fully equipped.	.26	.89	.83
Integrity-based trust
INT1. The quality of this house is trustworthy.	.84	.23	.80	.85	.81	.52
INT2. The quality of this house is reliable.	.81	.28	.80
INT3. It will be safe when staying at this house.	.75	.31	.76
INT4. It will be comfortable to stay at this house.	.76	.26	.73

ANOVA Results

To test H₁, a one-way ANOVA on booking intention was conducted. The results revealed significant between-group (fact-based vs. emotion-based) differences: fact-based reviews (vs. emotion-based) led to significantly higher booking intention (Mean_Fact-based = 5.40, Mean_{Emotion-based} = 5.12; F_(1,152) = 4.84, p-value = .03). Therefore, H₁ was supported.

CFA Results

To confirm the validity of the two dimensions of trust generated by EFA, CFA was conducted. As shown in Table 2, the standardized factor loadings of six observation variables are between .73 and .83, confirming each variable’s strong ability to interpret latent constructs (Hair Jr. et al., 2010). The χ² statistics and other goodness-of-fit indices of measurement model respectively were χ²/df = 2.21, GFI = .91, CFI = .94, IFI = .94, RMR = .05, and RMSEA = .09, which altogether confirmed the measurement model’s overall goodness of fit as the two dimensions of trust. Also, all Cronbach’s alpha (α > .80), composite reliability (CR > .70), average variance extracted (AVE > .50) indicate good internal consistency, reliability, and discriminant validity (Fornell & Larcker, 1981).

PROCESS Results

To verify the mediation effect of H_2a and H_2b, the analytic software of PROCESS (Hayes, 2013; Model 4) with the recommended bootstrap technique was used to test multiple mediators in a parallel manner. The results showed that the two parallel mediators (competence-based trust and integrity-based trust) both fully mediated the impacts of different written styles (independent variable = X) on customers’ booking intention (dependent variable = Y): insignificant direct effect of X on Y (β = -.01, 95% bootstrap CI [-.20, .18]), significant relative indirect effect of X on Y through the mediator of competence-based trust (β = -.12, 95% bootstrap CI [-.25, -.02]), and also significant relative indirect effect of X on Y through the other mediator of integrity-based trust (β = -.14, 95% bootstrap CI [-.27, -.05]). H_2a and H_2b were thus supported.

Study 2: Host-Focused Online Reviews (Fact-Based vs. Emotion-Based)

Data Analysis and Results

As with Study 1, EFA and CFA were conducted to validate the multidimensional trust construct before testing the proposed hypotheses by ANOVA and PROCESS procedures.

EFA results

EFA was conducted with principal component analysis to identify the three dimensions of trust in the host (i.e., competence-based, integrity-based, and benevolence-based). As expected, the principal component analysis generated three factors with an eigenvalue greater than 1. The three factors together explained 74.26% of variances: competence-based trust (28.96%), integrity-based trust (29.31%), and benevolence-based trust (15.99%), respectively. Detailed results are provided in Table 3.

OPEN IN VIEWER

Table 3.

EFA and CFA Results of Trust in the Host.

Factors/Items	Component			Standardized Factor Loading	α	CR	AVE
	F1	F2	F3
Competence-Based Trust
COM1. This host is competent.	.34	.76	.16	.81	.86	.86	.61
COM2. This host is capable.	.33	.76	.11	.78
COM3. This host will help me when I need.	.25	.74	.23	.74
COM4. This host will make me satisfied.	.22	.78	.32	.81
Integrity-Based Trust
INT1. This host is honest.	.83	.23	.26	.83	.90	.85	.59
INT2. This host will keep their words.	.84	.21	.17	.79
INT3. This host appears trustworthy.	.75	.42	.18	.86
INT4. This host will provide good accommodation.	.75	.40	.13	.82
Benevolence-Based Trust
BEN1. This host will mean no harm to me.	.22	.13	.85	.68	.67	.62	.45
BEN2. This host feels like a caring person.	.18	.35	.73	.74

CFA results

CFA test results showed that the standardized factor loadings of 10 observation variables are between .68 and .86, confirming each variable’s strong ability to interpret latent constructs (Hair Jr. et al., 2010). The χ² statistics and other goodness-of-fit indices of measurement model were χ²/df = 2.17, GFI = .90, CFI = .95, IFI = .95, RMR = .05, RMSEA = .08. Hence, the measurement model’s overall goodness of fit can be considered proper and has validated the three dimensions of trust in the host. As shown in Table 3, all Cronbach’s alpha, CR, and AVE indicate acceptable internal consistency, discriminant validity, and reliability (Fornell & Larcker, 1981).

ANOVA results

To test H₃, a one-way ANOVA on booking intention was conducted. The results indicated significant between-group differences: fact-based review (vs. emotion-based) significantly leads to higher booking intention (Mean_Fact-based = 5.54, Mean_{Emotion-based} = 5.22; F_(1,198) = 6.31, p value = .01). Therefore, H₃ is supported.

PROCESS results

To verify the mediation effect of H_4a, H_4b, and H_4c, PROCESS (Hayes, 2013; Model 4) was employed. The results revealed that competence-based trust and integrity-based trust fully mediated the impacts of different written styles (independent variable = X) on customers’ booking intention (dependent variable = Y), however, benevolence-based trust had no mediating effect: insignificant direct effect of X on Y (β = -.09, 95% bootstrap CI [-.29, .12]), significant relative indirect effect of X on Y through competence-based trust (β = -.14, 95% bootstrap CI [-.30, -.02]), and significant relative indirect effect of X on Y through integrity-based trust (β = -.09, 95% bootstrap CI [-.22, -.01]), while non-significant relative indirect effect of X on Y through benevolence-based trust (β = -.01, 95% bootstrap CI [-.04, .03]). Therefore, H_4a and H_4b were supported, and H_4c was not supported.

Study 3: The Moderation Effect of Gender on Host-Focused Versus Property-Focused Fact-Based Reviews

Data Analysis and Results

ANOVA results

To test H₅, a 2-way ANOVA between the subject focus (property-focused vs. host-focused) and customer gender (male vs. female) was performed on booking intention. As shown in Table 4, a significant interaction effect emerged (F_(1,187) = 17.56, p value = .00). The simple main effect test was then performed to explore any significant interaction effects among customers of the same gender. According to the results, female customers showed higher levels of booking intention when the online review was host-focused versus property-focused (Mean_{Property-focused} = 5.08, Mean_Host-focused = 5.91; F_(1,187) = 22.20, p value = .00), on the other hand, male customers showed no significant difference when reading the two types of online reviews (Mean_{Property-focused} = 5.41, Mean_Host-focused = 5.22; F_(1,187) = 1.32, p value = .25), supporting H₅.

OPEN IN VIEWER

Table 4.

ANOVA Results for Booking Intention.

Source	Sum ofSquare	df	MeanSquare	F	p
Corrected model	18.40	3	6.13	8.55	.00
Intercept	5512.73	1	5512.73	7685.68	.00
Subject focus	4.84	1	4.84	6.75	.01
Gender	1.52	1	1.52	2.12	.15
Interaction effect	12.59	1	12.59	17.56	.00
Error	134.13	187	.72
Total	5696.20	191

PROCESS results

To test H₆, the moderated mediation tests were conducted using the PROCESS procedure (Hayes, 2013; Model 8). In accordance with our theoretical predictions, among female customers, different subject-focused online reviews influenced their booking intention via the two-dimensional trust (competence-based trust: β = .14, 95% bootstrap CI [.05, .24]; integrity-based trust: β = .18, 95% bootstrap CI [.09, .29]). However, the indirect effect via trust was not significant among the male customers (competence-based trust: β = .04, 95% bootstrap CI [-.01, .10]; integrity-based trust: β = .01, 95% bootstrap CI [-.07, .08]). Thus, H₆ was supported.

Theoretical Implications

The present research contributes to the stream of online review research in hospitality and tourism. Although trust is a pivotal factor in shaping customers’ booking decisions for P2P accommodations, no research has yet pursued an in-depth investigation of the three sub-dimensions of trust (i.e., competence-, integrity-, and benevolence-based) in the P2P accommodation context. By filling this gap in the literature, our research contributes theoretically to the P2P accommodation literature by proposing a multi-subject-focus and multi-dimensional conceptualization of trust. Drawing on classic social psychology literature on trust judgment (Mayer et al., 1995; McKnight & Chervany, 2001), our conceptualization delineates the competence-, integrity- and/or benevolence-based dimensions of trust toward either the P2P property or the host. Accordingly, the empirical findings from the current research extend the relevant literature on P2P accommodations, providing a more robust and comprehensive measure of customer trust. Moreover, our findings demonstrate the multidimensional trust construct as the underlying mechanism driving customers’ decision-making processes behind P2P accommodation bookings. In line with previous research on customer trust (e.g., Hawlitschek et al., 2016; Ma et al., 2017; Montecchi et al., 2019), our research finds that customers exhibit a higher level of booking intention with fact-based (vs. emotion-based) reviews. Such results are mediated by competence- and integrity-based trust perception; however, benevolence-based trust does not mediate the relationship between the host-focused fact-based review and customers’ booking intention. A host’s benevolence is typically manifested in their altruistic behaviors and genuine care for guests’ welfare (Wu et al., 2017). An online review written by the third party (i.e., other customers), however, may not be able to effectively convey a host’s level of altruism and care. That is, the formation of benevolence-based trust requires direct personal interactions (either face-to-face, or verbal, or textual communication) in facilitating one to interpret the interactional others’ altruistic intention (Cheng et al., 2019; Mody et al., 2019). Thus, benevolence-based trust is more likely to impact only in direct customer-host interactions, but hardly formed in the context of third-party endorsement.

Extending existing hospitality and tourism literature on online reviews, our research contrasts the effectiveness of fact-based versus emotion-based reviews and finds that fact-based reviews can lead to higher P2P accommodation booking intention. Such findings are consistent with the dual-process theory (Wason & Evans, 1974) and the previous literature that endorses the importance of factual information in influencing consumption decisions (Baker & Kim, 2019). Given that the P2P accommodation sector has high quality heterogeneity and consumption risks (Shin et al., 2020), reducing perceived consumption risk and enhancing customer trust are keen steps to ensure bookings. In this regard, our findings demonstrate that fact-based (vs. emotion-based) reviews can help customers know better about the property or the host and enhance their trust perceptions, ultimately leading to higher booking intention.

Furthermore, our findings suggest that not all fact-based reviews are equally effective in eliciting customers’ booking intention for P2P accommodations. Among fact-based reviews, host-focused (vs. property-focused) reviews may be more effective in driving female customers’ booking intention. This effect, however, is mitigated among male customers. These findings align with the selectivity hypothesis (Meyers-Levy, 1986) that explains gender differences in information processing. Our findings support the theoretical notion that females (vs. males) tend to be more influenced by social information about interactional others when deriving trust judgments (Awad & Ragowsky, 2008; Mulac et al., 2001). The current discoveries provide future researchers with a potential avenue to explore further gender differences in information processing and consequent behavioral responses.

Managerial Implications

The findings from the current research can help P2P accommodation practitioners, including both P2P platform managers and property hosts, better understand customers and their decision-making process. Such insights can help P2P accommodation platforms further enhance the effectiveness of their online marketing efforts. For example, as fact-based reviews are demonstrated as being more influential than emotion-based reviews in magnifying customers’ booking intention for P2P accommodation, we recommend P2P accommodation platforms leverage features and functions within the platform to induce customers to write more fact-based reviews. P2P accommodation platforms can encourage greater posting of fact-based reviews by suggesting a series of attribute-specific (e.g., facilities, amenities, hospitality excellence, and service quality) questions in the review posting tab as review writing references and guidelines. Such questions will facilitate customers to recall service details about the property or the host and guide them to write fact-based reviews. The P2P accommodation service attracts customers with distinctive characteristics (e.g., homely amenities, fully equipped kitchen, authentic living style, local experiences with the host; Guttentag et al., 2018). Reviews specifying guest experiences with such unique features would help enhance potential customers’ trust toward the property and the host, which subsequently leads to enhanced booking intention (Subroto & Christianis, 2021; Visser et al., 2017).

Second, we recommend P2P accommodation platforms leverage the selective display of customer reviews to attract potential guests and enhance the effectiveness of online marketing efforts. To increase customers’ booking rates, platforms may exhibit those fact-based reviews that delineate specific information about the property and the host (i.e., how the property is equipped and how the host served the guests) at the top of the review page.

Finally, our findings highlight the importance of establishing customer trust by consistently improving the P2P property’s quality record while providing genuine care to satisfy customers. Our findings suggest that competence- and integrity-based trust perceptions are keen to drive customers’ booking intention for P2P accommodations. As excellent service performance has been demonstrated as the prerequisite of customer trust and satisfaction (Shin et al., 2020), P2P accommodation platforms should equip hosts with the relevant quality audit devices to help them ensure service excellence and gain positively written reviews.