Translation and psychometric evaluation of Smartphone Addiction Scale—Short Version (SAS-SV) among Chinese college students

Abstract

Background

Smartphone addiction is very prevalent among college students, especially Chinese college students, and it can cause many psychological problems for college students. However, there is no valid research instrument to evaluate Chinese college students’ smartphone addiction.

Objective

This study aimed to translate the Smartphone Addiction Scale—Short Version (SAS-SV) into Chinese and evaluate the psychometric characteristics of the Smartphone Addiction Scale- Chinese Short version (SAS-CSV) among Chinese college students.

Methods

The SAS-SV was translated into Chinese using the forward-backward method. The SAS-CSV was completed by 557 Chinese college students (sample 1: n = 279; sample 2: n = 278). 62 college students were randomly selected from the 557 Chinese college students to be meas- ured twice, with an interval of two weeks. The reliability of the SAS-CSV was evaluated by internal consistency reliability and test-retest reliability, and the validity of the SAS-CSV was evaluated by content validity, structural validity, convergent validity, and discriminant validity.

Results

The SAS-CSV presented good content validity, high internal consistency (sample 1: α = 0.829; sample 2: α = 0.881), and good test-retest reliability (ICC: 0.975; 95% CI: 0.966–0.985). After one exploratory factor analysis, three components (tolerance, withdrawal, and negative effect) with eigenvalues greater than 1 were obtained, and the cumulative variance contribution was 50.995%. The results of confirmatory factor analysis indicated that all the fit indexes reached the standard of good model fit (χ ²/df = 1.883, RMSEA = 0.056, NFI = 0.954, RFI = 0.935, IFI = 0.978, TLI = 0.969, CFI = 0.978). The SAS-CSV presented good convergent validity for the factor loading of all the items ranged from 0.626 to 0.892 (higher than 0.50), the three latent variables’ AVE ranged from 0.524 to 0.637 (higher than 0.50), and the three latent variables’ CR ranged from 0.813 to 0.838 (higher than 0.70). Moreover, the square roots of the AVE of component 1 (tolerance), component 2 (withdrawal) and component 3 (negative effect) were 0.724, 0.778, and 0.798, respectively, higher than they were with other correlation coefficients, indicating that the SAS-CSV had good discrimination validity.

Conclusion

The SAS-CSV is a valid instrument for measuring smartphone addiction among Chinese college students.