Validation of the Electronic Cigarette Attitudes Survey (ECAS) for Latinx and Non-Latinx Black Youth

Background and purpose: Electronic cigarette (e-cigarette) use among adolescents has rapidly increased in recent years, especially among Latinx and non-Latinx Black adolescents. Yet, limited research exists on adolescent attitudes that may contribute to the popularity of these products. This is in part due to the lack of measures focused on assessing adolescent attitudes towards e-cigarette use. The aim of this study was to develop a measure of adolescent e-cigarette use attitudes. Items for this measure were derived by both prior qualitative and theoretical work on factors that may contribute to the recent rise in popularity of these products. More specifically, the measure reflects theorized health risks, product characteristics (e.g., inconspicuous use), ease of purchase, and adolescent attitudes associated with e-cigarette use.  

Methods: Data were collected using a cross-sectional survey design. The Electronic Cigarette Attitudes Survey (ECAS) was administered to a sample of 247 adolescents (M age = 16.0, SD = 1.2) from a high school in Florida. Most of the sample (63.6%) identified as female, a majority (69.2%) were non-Latinx Black, and 27.1% were Latinx. To determine the factor structure, a parallel analysis scree plot and exploratory factor analysis (EFA) identified and extracted a likely factor structure using Mplus 7.11. To verify the factor structure, confirmatory factor analysis (CFA) of the ECAS was conducted using Mplus 7.11.

Results: A parallel analysis scree plot was conducted to compare 1000 observed eigenvalues to the 95th percentile of permuted eigenvalues for determining factor retention, the parallel analysis and eigenvalue supports a one-factor solution. The model fit for the one-factor was acceptable across indices, with the exception of the RMSEA (RMSEA=0.099, CFI=0.936, TLI=0.922, and SRMR=0.046). It is important to note that prior work demonstrates that models with smaller sample sizes (~100) had RMSEA values that often falsely indicated poor model fit (Kenny, Kaniskan, & McCoach, 2015). All factor loadings were above 0.60 and significant at the 5% level. The overall measure demonstrated good internal consistency (Cronbach's α=0.93). A CFA based on a one-factor solution of the ECAS was conducted on the second half of the sample chosen at random (n=121). Although all standardized loadings were significant (p < .001) and above the 0.50 criterion (range=0.55–0.83), the model fit on several indices were below threshold (RMSEA=0.124, CFI=0.875, TLI=0.847, and SRMR=0.063). Modification indices improved the model fit so that all indices were acceptable (RMSEA=0.092, CFI=0.934, TLI=0.917, and SRMR=0.053). Latinx adolescents were more likely than non-Latinx black adolescents to report lifetime use of e-cigarettes in the current study. No significant differences were found on ECAS scores based on race/ethnicity or e-cigarette use status (i.e. lifetime e-cigarette use vs. no use).

Conclusions and Implications: By investigating possible motives for the popularity of e-cigarettes among adolescents, researchers can help inform public health initiatives such as regulation policies and prevention efforts. Accordingly, the ECAS represents a viable measure for assessing e-cigarette attitudes among adolescents examining factors relevant to the upsurge of e-cigarette use among adolescents. Lifetime use of e-cigarettes, other tobacco products, and marijuana among adolescents are also discussed.