Results
The initial sample size was calculated by using a rule of thumb calculation of 15 minimum subjects per indicator [98], resulting in 840 subjects for 56 indicators. By taking into account invalid responses, a total of 1,367 responses were collected. The average reported time to take the survey was 28 minutes and 30 seconds. We excluded incomplete responses (n = 287), careless responses that took less than one standard deviation of average time spent (n = 78), and responses where participants answered questions containing simple attention checks incorrectly (n = 139), resulting in 863 valid responses (63% of 1,367).
The average age of participants was 33 years old (SD = 7.00). The majority of the respondents identified as male (57%) and selected the married category (65%). Most of the participants were white (74%), had obtained a bachelor’s degree (55%), and were employed full-time (85%). The mode income fell between $25,000 and $49,999. Detailed sample characteristics are shown in Table 1.
Table. Sample characteristics. (See the original article for the full table.)
Measurement model
To identify measurements composing constructs, we used confirmatory factor analysis (CFA) to assess the measurement model. CFA was performed among six latent variables, including FoMO, social shopping, PSI, social wellbeing, psychological wellbeing, and financial wellbeing. The CFA results are shown in Table 2 and the correlation matrix is shown in Table 3. Due to the low factor loadings below.60, we eliminated two items of the FoMO scale, two items from the social shopping scale, six items from the PSI scale, three items from the social wellbeing scale, one items from the psychological wellbeing scale, and four items from the financial wellbeing scale. After removing these items, the measurement model showed a good fit to the data (χ2 = 1449.56, χ2/df = 2.25, CFI = .96, TLI = .95, RMSEA = .03; [99]). All final factor loadings were over.61, all average variance extracted (AVE) values were over.51, and composite reliability (CR) values were over.82, confirming convergent validity [99]. The square root of the AVE for each construct was larger than the corresponding correlation coefficient between the factors, confirming discriminant validity [99].
Table. Study measures and confirmatory factor analysis results. (See the original article for the full table.)
Table. Correlation matrix. (See the original article for the full table.)
Hypothesis testing
We tested all hypotheses using a structural equation modeling (SEM). The SEM result demonstrated an overall good fit to the data: χ2 = 1482.67, 2/df = 2.29, CFI = .95, TLI = .95, and RMSEA = .04 [99]. The results are shown in Fig 2 and Table 4 .
Table. Structural equation modeling results. (See the original article for the full table.)
Figure. Structural equation modeling results.Note: * p < .05, **p < .01, ***p < .001. Solid lines indicate significant results, p < .05. Dashed lines indicate insignificant results, p ≥ 0.05. (See the original article for the figure.)
The SEM results showed a positive relationship between the FoMO and PSI (β = .63, t = 13.79, p < .001). This indicates that consumers who experience higher levels of FoMO appeared to have stronger PSIs with social media influencers, supporting H1. Further, we found that FoMO is inversely related to social wellbeing (β = -.80, t = − 14.00, p < .001), psychological wellbeing (β = -.29, t = − 6.00, p < .001), and financial wellbeing (β = -.67, t = − 12.23, p < .001), supporting H2a, H2b, and H2c. The result also showed that the relationship between PSI and social shopping was positive (β = .69, t = 14.93, p < .001), which means that consumers who experience PSIs through social media are more likely to engage in social shopping, accepting H3.
PSI was positively associated with all three types of wellbeing: social wellbeing (H4a: β = .18, t = 2.96, p < .01), psychological wellbeing (H4b: β = .57, t = 8.24, p < .001), and financial wellbeing (H4c: β = .15, t = 2.40, p < .05). The result demonstrated that those with stronger PSIs are more likely to have higher levels of social, psychological, and financial wellbeing. As hypothesized, H4a and H4b showed a positive relationship. However, the relationship between PSI and financial wellbeing was negative, which was the opposite direction proposed in H4c. Thus, we accept H4a, H4b, but reject H4c. We did not find a significant relationship between social shopping and three types of wellbeing, rejecting H5a, H5b, and H5c.
This study further examined the mediating roles of PSI and social shopping using the bootstrapping method. Significant indirect effects of FoMO on social shopping (B = .31, p < .01; B is unstandardized coefficient), social wellbeing (B = .14, p < .05), psychological wellbeing (B = .32, p < .01), and financial wellbeing (B = .10, p < .05) through PSI were found. However, the mediating effects involving a single mediator of social shopping and two mediators of PSI followed by social shopping were not found to be significant.
CC BY 4.0. "Fear of missing out, social media influencers, and the social, psychological and financial wellbeing of young consumers" by Bartosiak A, Lee JE, Loibl C, PloS one (2025). https://doi.org/10.1371/journal.pone.0319034. Licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Peer-reviewed and open access; full text from PubMed Central (PMC11999132). Headings and layout adapted for reading; figures and tables are in the original.