Emotional Branding and Consumer Behavior
Marketers try to influence consumers and their buying decisions about products and services through various means. Emotional branding is just one of many tools in the toolbox. An emotional connection helps them to create "loyalty without reason" for their product or service. Read this study to learn how emotional branding influences women's consumption behavior. Closely study Figure 1 in the literature review for a detailed model of the impact of emotional advertising on consumer behavior.
Results
Evaluation of measurement model
Table 2 results indicate that all loading items of the measurement model are greater than 0.70 and are found significant at a 0.05% level. It shows the items' level convergent validity as well. Composite reliability Cronbach's alpha values are more than 0.70, indicating acceptable reliability levels. Moreover, the AVE values of all items at the construct level are found to be more than 0.50. The heterotrait-monotrait (HTMT) values in all items are below 0.90, as mentioned in Table 3, which demonstrates the study's discriminant validity.
Table 2 Model Measurement
| Constructs | Items | LVs | CR | α | AVE | R2 |
|---|---|---|---|---|---|---|
| Emotions (E) | – | |||||
| Hope | E1 | 0.807 | 0.851 | 0.783 | 0.663 | |
| E2 | 0.811 | |||||
| Pride | E3 | 0.752 | 0.786 | 0.817 | 0.681 | |
| E4 | 0.773 | |||||
| Sadness | E5 | 0.837 | 0.793 | 0.849 | 0.652 | |
| E6 | 0.808 | |||||
| Fear | E7 | 0.819 | 0.825 | 0.825 | 0.661 | |
| E8 | 0.794 | |||||
| Appeal Drivers (A) | – | |||||
| Message | A1 | 0.835 | 0.788 | 0.721 | 0.585 | |
| A2 | 0.786 | |||||
| Picture quality | A3 | 0.822 | 0.857 | 0.746 | 0.637 | |
| A4 | 0.817 | |||||
| Color | A5 | 0.773 | 0.810 | 0.717 | 0.593 | |
| A6 | 0.741 | |||||
| Music | A7 | 0.832 | 0.863 | 0.733 | 0.539 | |
| A8 | 0.818 | |||||
| Celebrities (C) | – | |||||
| Showbiz | C1 | 0.853 | 0.811 | 0.738 | 0.675 | |
| C2 | 0.827 | |||||
| Sports | C3 | 0.877 | 0.865 | 0.716 | 0.611 | |
| C4 | 0.846 | |||||
| Attention (AT | ||||||
| AT1 | 0.776 | 0.901 | 0.855 | 0.701 | 0.605 | |
| AT2 | 0.862 | |||||
| AT3 | 0.863 | |||||
| Consumer Behavior (CB) | CB1 | 0.861 | 0.921 | 0.887 | 0.745 | 0.709 |
| CB2 | 0.873 | |||||
| CB3 | 0.856 | |||||
- Note: All loadings are significant at 0.05% level of significance (2-tailed)
- LV Loading Values, C.R Composite Reliability, α Cronbach’s alpha, AVE Average Variance Extracted, R2 Coefficient of Determination
Table 3 Heterotrait monotrait (HTMT) analysis results
| HTMT | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|
| 1 Emotions | |||||
| 2 Appeal Drivers | 0.46 | ||||
| 3 Celebrities | 0.42 | 0.61 | |||
| 4 Consumer Behavior | 0.44 | 0.59 | 0.58 | ||
| 5 Attention | 0.55 | 0.64 | 0.62 | 0.68 |
Structural model evaluation
This study shows path coefficients and coefficients of determination (R2) for structural model measurement. Moreover, a bootstrapping method (resampling) adopted with 5000 for the observations as the original sample (n = 240) for t-values and standard error. Magnitude and signs of path coefficients (as mentioned in Table 4 and Fig. 2) estimate the path relationship among the model variables.
Table 4 Path coefficients of model and significant testing of hypotheses (Direct Effect)
| Structural Path | Path Coefficients (t-value) | Confidence Interval | P-value (0.05%) | Results |
|---|---|---|---|---|
| E➙CB | 0.223 (5.836) | (0.152–0.305) | 0.001 | Accepted |
| A➙CB | 0.231 (4.462) | (0.132–0.342) | 0.000 | Accepted |
| C➙CB | 0.082 (2.733) | (0.023–0.143) | 0.004 | Accepted |
| AT➙CB | 0.348 (7.924) | (0.262–0.432) | 0.000 | |
| E➙AT | 0.331 (7.826) | (0.247–0.413) | 0.000 | |
| A➙AT | 0.301 (5.472) | (0.192–0.408) | 0.011 | |
| C➙AT | 0.094 (2.523) | (0.018–0.171) | 0.001 |
Fig. 2 Structural model results
Coefficients of determination (R2) values for endogenous variables are found adequate in this study (i.e., above the 10% level) (Falk and Miller, 1992). R2 values for attention (0.605) and consumer behavior (0.709) are moderate and considered appropriate for the study (Fig. 2). Beta and p values in Table 4 results show that the direct relationship of emotions (β = 0.223 (5.836), p = 0.001), appeal drivers (β = 0.231(4.462), p = 0.000) and celebrities (β = 0.082(2.733), p = 0.004) with consumer behavior are highly significant at 0.05% level. These results provide valid support for the proposed hypotheses H1, H2 and H3.
Mediation testing and analysis
Non-parametric bootstrapping was applied to measure the mediating effect significance. Results in Table 5 show that the indirect/mediation between emotions, appeal drivers, celebrity endorsement, and consumer behavior is significant (i.e., t > 1.96 at p = 0.05%). However, the indirect/mediation beta value effects are less than the direct beta value effects for each related construct. It shows that the mediation variable (attention levels) absorbed some percentage of the direct effect as well. These results support the proposed hypotheses H4a, H4b and H4c. Since the indirect effect in this study is significant between the constructs with absorption effect, it justifies the mediation requirement effect mentioned by Hair et al.. Hence, bootstrapping analysis results, shown in Table 4, meet the necessary requirements for a mediator measurement effect. VAF (Variance account for) results in Table 5 are found to be between 20% to 80% which identifies the mediation effect of attention levels on emotions, appeal drivers, and celebrities as partial mediation.
Table 5 Bootstrapping Mediation test (Indirect / Mediation Effect)
| Structural Path | Direct Effect (t- Value) | Indirect Effect (t- Value) | Total Effect` | VAF (%) | Interpretation | Results |
|---|---|---|---|---|---|---|
| E➙AT➙CB | 0.223 | 0.114 | 0.337 | 34.16 | Partial Mediation | Accepted |
| (5.836) | (5.686) | |||||
| A➙AT➙CB | 0.231 | 0.103 | 0.334 | 31.24 | Partial Mediation | Accepted |
| (4.462) | (4.236) | |||||
| C➙AT➙CB | 0.082 | 0.034 | 0.116 | 29.12 | Partial Mediation | Accepted |
| (2.733) | (2.476) |
- VAF variance accounted for, n.s not significant; |t| > =1.96 at p = 0.05 level; The VAF > 80% indicates full mediation, 20% ≤ VAF ≥ 80% shows partial mediation while VAF <20% is no mediation
The ranking analysis shows that overall the participants positively preferred music and color appeal drivers over the other appeal drivers being tested in the present study with the highest mean values (i.e., Meanmusic = 5.46, Meancolors = 4.55) and lowest standard deviation values comparatively (i.e., S.Dmusic = 1.53, S.Dcolors = 1.79) as compared to the other appeal drivers (Meanpicture = 4.38, S.Dpicture = 1.82; Meanmessage = 3.34, S.Dmessage = 1.88). This is shown in Table 6 and Fig. 3. Likewise, the study respondents positively preferred the emotions of happiness and pride over the other emotions being tested with the highest mean values (i.e., Meanhappiness = 5.86, Meanpride = 4.43) and the lowest standard deviation values (i.e., S.Dhappiness = 1.52, S.Dpride = 1.80) as compared to the other emotional stimuli (Meanfear = 4.31, S.Dfear = 1.87; Meansadness = 3.28, S.Dsadness = 1.90). This is shown in Table 7 and Fig. 4. In terms of celebrity choice, a celebrity endorser from the showbiz industry was preferred, with a mean value of 5.67 and low standard deviation value of 1.55, over a sport celebrity (Meansports = 4.58, S.Dsports = 1.69). This trend has been summarized in Table 8 and Fig. 5.
Table 6 Most preferred Appeal Drivers
| Music | Colors |
|---|---|
| 1 | 2 |
| Mean = 5.46 | Mean = 4.55 |
| S.D = 1.53 | S.D = 1.79 |
| Message | Picture quality |
| 4 | 3 |
| Mean = 3.34 | Mean = 4.38 |
| S.D = 1.88 | S.D = 1.82 |
Fig. 3 Most preferred appeal drivers
Table 7 Most preferred emotions
| Happiness | Pride |
|---|---|
| 1 | 2 |
| Mean = 5.86 | Mean = 4.43 |
| S.D = 1.52 | S.D = 1.80 |
| Sadness | Fear |
| 4 | 3 |
| Mean = 3.28 | Mean = 4.31 |
| S.D = 1.90 | S.D = 1.87 |
Fig. 4 Most preferred emotions
Table 8 Most preferred type of celebrity
| Showbiz | Sports |
|---|---|
| 1 | 2 |
| Mean = 5.67 | Mean = 4.58 |
| S.D = 1.55 | S.D = 1.69 |
Fig. 5 Most preferred type of celebrity
