From Information Experience to Consumer Engagement
Read this comprehensive study on the UX experience within the fashion sector. Examine the research to identify how values on positive emotions influence a user's engagement toward a fashion brand. At what point in this research was the user experience enhanced (the conversion process)?
Results
Sample characteristics
The demographic characteristics of respondents are summarized in Table 1. The mean age of the respondents was 28 years old, ranging from 20 to 49, and 69.0% were female. The respondents were fairly well-educated, with 56.6% indicating that they had completed a university degree or above. As for SNS usage, respondents listed Facebook (79%), Instagram (77.2%), Kakao Story (39%), Twitter (27.9%), and Pinterest (6.2%) as their favorite SNSs. The average time per day respondents spent on SNSs was approximately 35 min.
Table 1 Demographic characteristics
|
Characteristic |
n |
% |
|---|---|---|
|
Gender |
||
|
Male |
90 |
31.0 |
|
Female |
200 |
69.0 |
|
Age |
||
|
20–25 |
138 |
47.6 |
|
26–30 |
61 |
21.0 |
|
31–35 |
53 |
18.3 |
|
36–40 |
21 |
7.2 |
|
41–49 |
17 |
5.9 |
|
Education |
||
|
High school graduate |
17 |
5.9 |
|
Bachelor's and college student |
109 |
37.6 |
|
Bachelor's degree |
147 |
50.7 |
|
Master's degree or above |
17 |
5.9 |
|
Occupation |
||
|
Self-employed |
4 |
1.4 |
|
Sales and service work |
12 |
4.1 |
|
Manufacturing work |
3 |
1.0 |
|
Office work |
109 |
37.6 |
|
Business administration and management |
5 |
1.7 |
|
Professional |
18 |
6.2 |
|
Student |
113 |
39.0 |
|
House work |
9 |
3.1 |
|
Others |
17 |
5.9 |
|
Household monthly income (Korean Won) |
||
|
Less than 1,000,000 |
60 |
20.7 |
|
1,000,000–less than 2,000,000 |
62 |
21.4 |
|
2,000,000–less than 3,000,000 |
56 |
19.3 |
|
3,000,000–less than 4,000,000 |
37 |
12.8 |
|
4,000,000–less than 5,000,000 |
22 |
7.6 |
|
5,000,000–less than 6,000,000 |
16 |
5.5 |
|
More than 6,000,000 |
37 |
12.7 |
Data analyses
Hypotheses one to three were tested using a two-stage structure equation modeling approach. First, confirmatory factor analysis (CFA) was performed in order to ensure the quality of the proposed measurement model. Second, structural equation modeling was conducted to test the proposed hypotheses. Both analyses were performed with Amos 25.0 with maximum likelihood estimation of the covariance matrix. To test Hypotheses 4a and 4b, data analyses were conducted using the PROCESS macro from SPSS.
Measurement model
CFA was performed to establish the fit of the measurement model for structural analysis. The CFA results exhibited an acceptable fit (χ2 = 495.19, df = 317, χ2/df = 1.56, p < 0.001, CFI = 0.97, TLI = 0.94, IFI = 0.97, RMSEA = 0.04, SRMR = 0.04). All the coefficients were significant (C.R. > 12.16). Table 2 provides the items used in the model, standardized factor loadings, Cronbach's alpha coefficients, and the construct reliabilities. As shown in Table 3, the AVEs of all the constructs were greater than the threshold value of 0.5, so the convergent validities of all constructs were established. In addition, the AVE of each construct was greater than the shared variances (squared correlation coefficients) between all possible pairs of constructs, providing evidence for discriminant validity. Consequently, the analyses confirmed the construct validity of all the latent constructs.
Table 2 Measurement model assessment results
|
Variables |
Factor loading |
Cronbach's α |
Construct reliability |
|---|---|---|---|
|
Usefulness (Interacting with information, such as postings, in this brand page was…) |
|||
|
Ineffective–effective |
0.85 |
0.90 |
0.90 |
|
Unhelpful–helpful |
0.88 |
||
|
Impractical–practical |
0.75 |
||
|
Useless–useful |
0.83 |
||
|
Enjoyment (While I was interacting with information, such as postings and comments, in this brand page…) |
|||
|
I had fun |
0.79 |
0.88 |
0.88 |
|
It provided me with a lot of enjoyment |
0.89 |
||
|
I enjoyed the process of the information interaction |
0.86 |
||
|
Positive emotion (My feelings during the interaction with information in this brand page was close to…) |
|||
|
Bad–good |
0.75 |
0.86 |
0.86 |
|
Sad–happy |
0.81 |
||
|
Positive–negative (reversed) |
0.71 |
||
|
Unpleasant–pleasant |
0.84 |
||
|
Satisfaction (While I was interacting with information, such as postings and comments, in this brand page…) |
|||
|
The overall experience with this brand page was delighted |
0.75 |
0.79 |
0.79 |
|
Overall, I am very satisfied with this brand page |
0.86 |
||
|
Cognitive engagement (While I was interacting with information, such as postings and comments, in this brand page…) |
|||
|
I was deeply engrossed |
0.78 |
0.89 |
0.89 |
|
I concentrated fully on what I was doing |
0.89 |
||
|
I was totally absorbed in what I was doing |
0.88 |
||
|
Elaboration (While I was interacting with information, such as postings and comments, in this brand page…) |
|||
|
I thought about what actions I myself might take based on some posts on this brand page |
0.72 |
0.87 |
0.87 |
|
I found myself making connections between the contents (posts) and what I've read or heard about elsewhere |
0.75 |
||
|
I thought about how what I had read/saw on this brand page related to other things I know |
0.67 |
||
|
I tried to think of the practical applications of what I read/saw from the posts in the brand page |
0.78 |
||
|
I tried to relate the ideas in the stories (of the posts) to my own life |
0.84 |
||
|
Engagement intention |
|||
|
I intend to stay on as a friend/follower (customer member) in this SNS site |
0.75 |
0.91 |
0.90 |
|
I will recommend this SNS site to others |
0.83 |
||
|
If friends or relatives who has the same interest ask for advice, I would recommend this SNS site |
0.84 |
||
|
I want to help other customers/members/friends in this SNS with their questions |
0.79 |
||
|
I want to help the brand to improve its service |
0.76 |
||
|
Curiosity |
|||
|
Interacting with the information interaction excited my curiosity |
0.75 |
0.83 |
0.82 |
|
Interacting with the information made me curious |
0.82 |
||
- Stem question in parentheses
Table 3 Convergent and discriminant validity
|
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
|
|---|---|---|---|---|---|---|---|---|
|
1. Usefulness |
0.69 |
|||||||
|
2. Enjoyment |
0.42 |
0.72 |
||||||
|
3. Positive emotion |
0.49 |
0.40 |
0.61 |
|||||
|
4. Satisfaction |
0.55 |
0.52 |
0.14 |
0.65 |
||||
|
5. Cognitive Engagement |
0.27 |
0.37 |
0.31 |
0.29 |
0.72 |
|||
|
6. Elaboration |
0.28 |
0.35 |
0.14 |
0.27 |
0.31 |
0.57 |
||
|
7. Engagement intention |
0.38 |
0.38 |
0.24 |
0.59 |
0.41 |
0.32 |
0.63 |
|
|
8. Curiosity |
0.44 |
0.67 |
0.36 |
0.45 |
0.28 |
0.44 |
0.24 |
0.70 |
- Diagonal entries are AVE for each construct. Off-diagonal entries are the squared correlation coefficients between constructs
Hypotheses testing
Structural equation modeling was performed to test hypotheses one to three. The results exhibited an adequate model fit (χ2 = 585.80, df = 285, χ2/df = 2.06, CFI = 0.94, TLI = 0.93, IFI = 0.94, RMSEA = 0.06, SRMR = 0.07). The model accounted for 76.8%, 66.2%, 32.6%, 30.6%, and 64.8% of the variances in positive emotions, satisfaction, cognitive engagement, elaboration, and consumer engagement intentions, respectively (see Fig. 2). Hypotheses 1a and 1b predicted the impact of the instrumental and non-instrumental values of information interactions - usefulness and enjoyment - on positive emotions. Positive emotions were significantly influenced by usefulness (β = 0.54, t = 8.12, p < 0.001) and enjoyment (β = 0.42, t = 6.65, p < 0.001). Therefore, Hypothesis 1a and 1b were supported. Furthermore, in order to see the relative strength of the two values, the bootstrap method (using 1000 re-samples) was employed. The substantial overlap of over 50% of 95% bias-corrected confidence intervals (CIs) suggests that the beta weights of the two were not statistically different. The results inferred that both usefulness and enjoyment perceived from information environments were equally important in arousing positive emotions in one's information experiences in the context of fashion brand pages, supporting the current perspective on user experiences in the disciplines of information systems and HCIs. Hypotheses 2a–2c explicated the associations between positive emotions and the three dimensions of experiential states, including satisfaction, cognitive engagement, and elaboration. Results showed that positive emotions had a significantly positive impact on satisfaction (β = 0.81, t = 11.96, p < 0.001), cognitive engagement (β = 0.57, t = 8.58, p < 0.001), and elaboration (β = 0.55, t = 8.18, p < 0.001), in support of Hypotheses 2a, 2b, and 2c. Hypotheses 3a–3c postulated the associations between experiential states and consumer engagement intentions toward brand SNS pages. The results revealed that engagement intentions were significantly influenced by satisfaction (β = 0.53, t = 7.68, p < 0.001), cognitive engagement (β = 0.30, t = 5.20, p < 0.001), and elaboration (β = 0.16, t-value = 2.95, p < 0.01), yielding support for Hypotheses 3a, 3b, and 3c. Through bootstrapping (1000 re-samples), we examined the extent of overlaps of CIs for the beta estimates for the effects of the three experiential states (satisfaction, cognitive engagement, and elaboration) on consumer engagement intentions. The results revealed that the overlaps for all pairs were less than 50%, indicating that the strength of the effects of experiential states on engagement intentions differed significantly from each other. Specifically, satisfaction was found to have a greater impact on engagement intentions (β = 0.53), followed by cognitive engagement (β = 0.30) and elaboration (β = 0.16).
Fig. 2
Resulted model. Numbers are standardized regression weights. **p < 0.01, ***p < 0.001
The PROCESS macro from SPSS (model 1, 5000 bootstrap samples) was used to test Hypotheses 4a and 4b regarding the possibility that curiosity moderates the relationships between perceived values (usefulness, enjoyment) and positive emotions. We first tested whether or not curiosity moderated the effect of perceived usefulness on positive emotions. The results revealed that the overall model was significant: F(3286) = 69.47, p < 0.001, R2 = 0.422. The significant interaction between usefulness and curiosity (b = 0.11, SE = 0.03, t = 3.32, p < 0.01) suggested that as curiosity increases, the relationship between perceived usefulness and positive emotions becomes stronger. The second procedure to determine the moderating effect of curiosity on the relationship between enjoyment and positive emotions also indicated that the overall model (F(3286) = 53.09, p < 0.001, R2 = 0.422) and the interaction between enjoyment and curiosity (b = 0.12, SE = 0.03, t = 4.25, p < 0.01) were significant. Therefore, the relationship between perceived enjoyment and positive emotions becomes stronger with increasing degrees of curiosity. The results of the moderation analyses are highlighted in Table 4. As expected, the influence of positive emotions related to one's instrumental and non-instrumental values involved in information interactions becomes stronger for highly curious consumers, supporting Hypotheses 4a and 4b.
Table 4 Moderation effect of curiosity
|
Predictor |
Curiosity |
Effect |
SE |
t |
CI95% |
|---|---|---|---|---|---|
|
Perceived usefulness |
3.99 (Mean − 1SD) |
0.43 |
0.07 |
6.57 |
0.30, 0.56 |
|
5.04 (Mean) |
0.54 |
0.06 |
8.92 |
0.42, 0.66 |
|
|
6.08 (Mean + 1SD) |
0.65 |
0.07 |
8.98 |
0.51, 0.79 |
|
|
Perceived enjoyment |
3.99 (Mean − 1SD) |
0.23 |
0.07 |
3.15 |
0.08, 0.37 |
|
5.04 (Mean) |
0.36 |
0.07 |
5.44 |
0.23, 0.48 |
|
|
6.08 (Mean + 1SD) |
0.48 |
0.07 |
6.74 |
0.34, 0.62 |