Experiment 2

To investigate whether the above phenomenon exists in different product buying situations, different experimental conditions were designed. In experiment 1, participants chose between two kinds of apartments, one is smaller but convenient and the other is bigger but farther away from their company. Considering there are more and more online entertaining services turning up these days, especially after the commercial environment was changed by COVID-19 (Yeung, 2020), in experiment 2, participants chose between two online music streaming services. Experiment 2 was conducted to test whether the effect of imagery perspective and construal level on decision difficulty and moderation of the need for cognition can apply in the online choosing scenario.

3.3.1. Participants and Design

The experiment was designed as a single factor between-subject one (imagery perspective: actor versus observer). The participants were randomly assigned to one of two conditions. Then, the decision difficulty, construal level and need for cognition were tested. A total of 185 participants took part in this study: 98 participants were assigned to the actor perspective and 87 to the observer perspective (95 female participants and 90 males), with most participants “21-25 years old”, interval (N21-25=45).

22

3.3.2. Procedure and Materials

First, this experiment used manipulation the same as in the pilot test and experiment 1 to verify that participants were fully into the assigned imagery perspective state. Participants read the explanation about the two perspectives and then watched two demonstration videos shot from corresponding perspectives to understand the two kinds of imagery perspectives. The participants closed their eyes to imagine riding a bike for thirty min and then described the scene they pictured.

After the practice of imagining riding a bike, the participants imagined themselves choosing from two online music streaming services with desirability and feasibility trade-off. In the choosing scenario, the number of songs and the efficiency to upload new releases signified desirability, ease to use and smoothness of user interface, representing feasibility. The two brands are at the same price and other unmentioned condition are also the same.

Participants read the following description:

You are an office worker and you are used to listen to music through your cellphone.

You are looking for an online music streaming service to subscribe. After spending time looking for an appropriate brand, you excluded other alternatives and target the two brands below:

Brand A:

 more songs, you can enjoy every song you want

 low efficiency to upload new releases

 the smoothness of user interface scored 1 (up to 5) and the APP lags or crashes sometimes

23

Brand B:

 fewer songs

 high efficiency to upload new releases

 the smoothness of user interface scored 5 (up to 5) and the APP is easy to use and is also aesthetic

The price and sound quality of the two brands are the same. Please adopt the actor perspective (or observer perspective) and imagine your daily life after subscribing to Brand A and Brand B, separately, i.e., picture your daily routine or habits and how you would feel using the service to listen to music such as enjoying music on Sunday mornings o, killing time when you are going to work. Which brand do you prefer?

After deciding which brand to subscribe, participants rated the decision difficulty, did the behavioral identification form and evaluated the need for cognition. To ensure the accuracy when comparing the two experiments, this experiment used the same scale as experiment 1. Participants’ decision difficulty was measured by the 9 items, including perceived difficulty, readiness, confidence, satisfaction with their choice, and tendency to postpone (Fitzsimons, 2000; Thompson et al., 2009); Cronbach’s alpha=.92.

Construal level was rated by the Behavioral Identification Form (BIF) designed by Vallacher and Wegner (1989). Finally, the need for cognition was measured by an 18-item scale (Cacioppo & Petty, 1982); Cronbach’s alpha=.92.

3.3.3. Results Main effect

According to the results of the independent samples T-test, imagery perspective affects people’s decision difficulty (referred to as “IP” and “DD” in the table below).

The decision difficulty of people adopting the actor perspective significantly differed

24

from those adopting the observer perspective. People adopting the actor perspective tended to perceive more difficulty in regard to their decision. In contrast, those imagining from the observer perspective tended to experience less difficulty regarding their decision (Mactor=3.34, SD=.76, Mobserver=3.13, SD=.58; t=2.06, p=.04, d=-.299).

Table 5. Results of Independent Samples T-test between IP and DD (Study 2)

N M SD df t p d

Decision Difficulty

Actor

Perspective 98 3.34 .76

179 2.06 .04 -.299 Observer

Perspective 87 3.13 .58

The statistics results support the following hypothesis:

H1: Imagining from the actor perspective causes more decision difficulty, while imagining from the observer perspective leads to less decision difficulty.

Mediating effect

To verify whether there’s a mediating effect of the construal level toward the imagery perspective and decision difficulty, this study employed the Sobel test for its analysis. The statistics results reveal that the mediating effect is significant (z=-8.791, p<.00). The standardized effect size of mediating effect was -0.17, which indicated that adopting the observer perspective affects the construal level and then decreases the standard deviation of decision difficulty by 0.17.

In other words, the construal level can explain the relationship between the imagery perspective and decision difficulty since the imagery perspective alters one’s

25

construal level. People adopting the actor perspective tend to construe the options at a lower level and have a relatively concrete mindset; thus, they see the options more distinctly and have more decision difficulty. Those who adopt the observer perspective tend to represent at a higher level and have a relatively abstract mindset, perceiving more similarity, and thus having less decision difficulty.

Table 6. Results of Sobel Test (Study 2)

d Sobel’s z p

Construal

Level -.017 -8.791 .00

The statistics results support the mediating effect of construal level toward imagery perspective and decision difficulty:

H2: The construal level mediates the relationship between imagery perspectives and decision difficulty. In other words, adopting the actor perspective lowers one’s construal level and leads to more decision difficulty.

Moderating effect

The result of second order linear regression is demonstrated in table 3 below.

According to the result, imagery perspective and need for cognition (referred to as “IP”

and “NFC”) can explain 32% of variance of construal level (Fstep1=44.45, p<.001).

Considering the interaction of imagery perspective and need for cognition, the variance of construal level increases 2% (Fstep2=6.315, p<.05).

26

Table 7. Regression of Second Order Interaction (Study 2) Construal Level

IP=Imagery Perspective; NFC=Need for Cognition

In terms of main effect, the ability of imagery perspective to explain construal level is significant (p<.001). Yet, the effect of NFC is not significant (β=-.02, p>.05).

Moreover, the interaction between imagery perspective and need for cognition significantly explains the construal level (β=-.21, p<.01).

Figure 3. Interaction of IP and NFC (Study 2)

27

As the above interaction graph shows, observer perspective leads to higher construal levels. Moreover, for people with low NFC, the effect of imagery perspective influencing their construal level is greater than for people with higher NFC. In summary, NFC is a moderator and supports Hypothesis 3.

H3: The need for cognition moderates the relationship between imagery perspective and construal level. People with higher NFC tend to think at lower construal levels while people with lower NFC have higher construal levels.

Although the moderating effect exists in both experiments 1 and 2, NFC influences the relationship between imagery and construal level in different ways. The phenomenon may result from the difference in online-offline choosing scenarios, a difference that has been studied for many years. The two channels provide different values, affecting individuals’ decision-making process and their behavior. Research on attitude formation has confirmed that NFC influences the way consumers form their attitude toward the advertisement and the brand (Sicilia et al., 2006); in their study, the effect caused by one’s NFC differs between online websites and traditional advertising media.

Furthermore, Xu et al. (2021) and Lee et al. (2021) suggest that when consumers evaluate a product in an online channel, they construe at higher levels than those buying offline. In summary, this study posits that the online-offline difference affects how NFC moderates the relationship between imagery perspective and construal level.

28