Bipolar vs. unipolar scaling in dynamic network analyses of Ecological Momentary Assessment data

Abstract

This study examined the statistical and clinical benefits of using bipolar versus unipolar scaling in dynamic network analysis of Ecological Momentary Assessment (EMA) data. Methods: Forty-seven students completed EMA reports three times daily for five weeks via either unipolar (n = 24) or bipolar (n = 23) scales. The data were analyzed to construct idiographic network models. Results: The bipolar scaling group presented significantly lower zero inflation (2.37% vs. 10.31%, U = 2407756, r = 0.75, p < .05) and greater response variability. Network analysis revealed more participants with significant network edges in the bipolar group (69.57% vs. 41.67%, χ²(1) = 12.06, p = .0007). Additionally, the bipolar group had lower odds of zero responses than the unipolar group did (p = .038). Conclusion: Bipolar scaling enhances EMA data quality by reducing zero inflation and increasing variability, resulting in richer dynamic network models. Further research is needed to confirm these findings in clinical populations.