**3. Results**

## **3.1 The relationship between fandom and PERMA**

The respondents' (n = 2046) participation in the fandom was as follows (multiple answer):


Those who chose the fourth option (n = 994, mean age 45.3 years, 54.0% male) were classified as non-fandom participants, and the rest were classified as participants (n = 1052, mean age 40.4 years, 55.0% male).

First, the level of well-being of the two groups was compared using the PERMA model, which consists of positive emotion (α = 0.85), engagement (α = 0.79), relationship (α = 0.81), meaning (α = 0.89), and accomplishment (α = 0.82) (see details in **Table 2**). It includes an item with a Cronbach's alpha slightly below 0.8, but in view of the design of the scale itself, the original version of the scale was retained. As a result, fandom participants showed significantly higher levels of well-being in all criteria—positive emotion (*t* = 5.95, *df* = 2010.18, *p* < 0.001), engagement (*t* = 10.04, *df* = 2003.63, *p* < 0.001), relationship (*t* = 5.44, *df* = 2029.87, *p* < 0.001), meaning (*t* = 5.40, *df* = 2027.15, *p* < 0.001), and accomplishment (t = 4.91, *df* = 2019.62, *p* < 0.001)—than the nonparticipants (**Table 3**).

Furthermore, regression analysis was conducted to examine the impact of participation in fandom after eliminating other factors that could influence well-being. The dependent variables are the elements of PERMA, and the independent variables are the fandom dummy, gender dummy, age, marital status dummy, educational background, employment status, and annual income. Before the analysis, to avoid multicollinearity, it was confirmed that the residuals of each dependent variable followed a normal distribution and the VIF of each independent variable is less than 3 (**Table 4**). The results showed that participation in knowledge sharing was a significant predictor of all PERMA elements: positive emotion (*β* = 0.12, t = 4.80, *p* < .001), engagement (*β* = 0.19, t = 7.56, *p* < .001), relationship (*β* = 0.10, t = 3.93, *p* < .001), meaning (*β* = 0.11, t = 4.28, *p* < .001), and accomplishment (*β* = 0.10, t = 4.16, *p* < .001).


### **Table 3.**

*Comparison of the level of well-being in two groups.*

Therefore, it was shown that participation to the fandom had a significantly positive impact on the level of well-being.

### **3.2 Classification of fandom participants**

To gain a better understanding of fandom, cluster analysis was conducted by the frequency of fandom activities by adopting the k-means method. After applying the number of clusters from two to five, the participants were classified into four clusters (**Table 7**). Each cluster was characterized as follows: Cluster 1 participants (n = 321, mean age = 41.0 years, 55.1% male) are more willing to engage in activities with peers than doing them alone (**Table 7**). They seem to go out to see a live concert or enjoy viewing their favor's contents in home if someone does together or invite them and would not do it alone. Thus, they are named "accompanying." In contrast, Cluster 3 participants (n = 176, mean age = 42.2 years, 54.0% male) prefer to do fandom-related activities alone (**Table 7**). They can accompany someone to the same extent as Cluster 1 participants, but they are more likely to go alone; accordingly, they are named "independent." Cluster 2 participants (n = 313, mean age = 35.5 years, 56.2% male) are frequent with respect to overall fandom activity (**Table 7**). They work with their peers, sometimes alone, and even contact their favorites directly. It shows that they invest a lot of time and effort into their fandom, and thus, they are named "enthusiastic." In opposition, Cluster 4 (n = 242, mean age = 44.5 years, 54.1% male) showed overall infrequency in fandomrelated activities (**Table 7**). They engage in fandom without going out and buying goods, in other words, spend less time and money. Therefore, they are named "casual".

Each cluster has a different degree of involvement in fandom and with others through it. This study examined the levels of PERMA for each cluster to see if such a difference in fandom involvement affects well-being. As a result of ANOVA, there were significant differences among the clusters for each of the PERMA factors: positive emotion (*F* (3, 1048) = 16.98, *p* < 0.001), engagement (*F* (3, 1048) = 17.35, *p* < 0.001), relationship (*F* (3, 1048) = 23.72, *p* < 0.001), meaning (*F* (3, 1048) = 17.46, *p* < 0.001), and accomplishment (*F* (3, 1048) = 15.41, *p* < 0.001).

Subsequently, differences in mean values of each PERMA element were assessed with Bonferroni multiple comparison (**Table 5**). The results indicated that the "enthusiastic" cluster (Cluster 2) showed significant higher levels in all PERMA elements than the other clusters (**Table 5**).


*The Relationship between Fandom and Well-Being DOI: http://dx.doi.org/10.5772/intechopen.106870*

**Table 4.**

 *The results of regression analysis.*


*Happiness and Wellness - Biopsychosocial and Anthropological Perspectives*

**Table 5.** *The result of multiple comparisons.*


**Table 6.** *Distribution of fandom objects by clusters.*

*The Relationship between Fandom and Well-Being DOI: http://dx.doi.org/10.5772/intechopen.106870*

**Table 6** shows the distribution of fandom objects by clusters. A chi-square test revealed that the distribution of fandom objective significantly differs in clusters (χ2 = 67.79, *df* = 18, *p* < .001). The "enthusiastic" Cluster 2 participants tended to have more multiples, which means that they have multiple favorites across genres and accordingly spent a lot of time and effort on fandom. The "independent" participants of Cluster 3 are more likely to target fandoms other than people and characters, which is consistent with them acting alone. The "casual" Cluster 4 participants had a relatively high percentage of respondents who supported only the person without crossing genres. The "accompanying" participants of Cluster 1 had a similar distribution as the total.
