Still under stress? Post pandemic change in the relationship of affective dimensions of words

Abstract

During the last two years, we witnessed the coronavirus pandemic and its impact over the mental health (e.g. Damnjanović et al., 2020; Rudroff et al., 2020). Previous research showed that emotional valence (EV) and arousal (A) ratings change under suspense (Delatorre et al., 2019), while Planchuelo et al. (2020) recorded lower A estimates during the COVID-19 lockdown. Having this in mind, we assumed that isolation due to quarantine would change how we emotionally experience words and that the EV and A estimates from post COVID-19 period should be similar to those collected in 2018. To answer this hypothesis, we compared the EV and A estimates collected during 2018 (the first wave), with the new ratings collected at the beginning of the pandemic in 2020 (the second wave) and in the summer of 2022 (the third wave). In the first and second wave, participants were psychology students (N1 = 40, N2 = 42; Dage = 19, ~90% women); in the third wave, participants were accessed via social networks (N3 = 100; Mage= 41.7±8, 86% women). The number of words presented to participants varied across three data collection waves (N1 = 2100, N2 =8 02, N3 = 882). For EV, extremes of the bipolar scale represented negative (1) and positive (7) words. Arousal was rated on a unipolar scale, where low extreme represented words low in arousal and high extreme highly arousing words. The middle point (4) represented neutral words in both scales. The EV and A estimates from the second and third waves showed high correlations with those collected during 2018 (the first and second wave correlations: rEV(800) = .93; .90, p < .001; rA(800) = .76; .70, p < .001). The EV is a more stable indicator of the emotional experience of words, compared to the A, which was in accordance with previous normative studies. A new finding concerns a relationship between EV and A estimates usually described via a quadratic or U-function (Bradley & Lang, 1999), indicating that negative and positive words excite us more, while we are indifferent to neutral words. Such a relationship was recorded in the first wave (r(800) = - .29, p < .001) but not during the second and third waves. The relationship became nearly linear (r = - .55, p < .001) in the second wave and perfectly linear in the third (r = -.77, p<.001). Our results revealed that our participants became more sensitive to the negative content and less sensitive to the positive content. This pattern could be linked to reduced resilience and may represent a mental health risk.