Determine the temporal ordering of distress-related language and polarization

Determine whether distress-related language expressed in computational social-media measures precedes or follows political polarization.

Background

Computational studies use sentiment, emotional expression, and cognitive-distortion language as indirect proxies for psychological distress in polarized online environments. However, these measures are generally cross-sectional and correlational, and they lack clinical validation at the individual level.

The authors explicitly identify the temporal ordering between distress-related language and polarization as unresolved. Establishing this ordering would clarify whether such language functions as a psychological antecedent of polarization, a consequence of polarized interaction, or merely a correlated signal.

References

Second, it limits causal interpretation because most studies using computational proxies are cross-sectional and correlational, making it unclear whether distress-related language precedes or follows polarization.

— Political polarization and mental wellbeing: asymmetric evidence for bidirectionality  (2610.11601 - Mahir et al., 8 Oct 2026) in Section “Operationalization of mental wellbeing in polarization research”