Will AI-generated relationships bring a new wave of dependency? In my latest blog, I depict a plausible near-future scenario of AI-assisted resurrection; I discuss who might be at risk and what should be done to prevent such a looming crisis. rad-blog.com/2024/04/15/a...

Sorry for the late reply (I'm less active on bsky). But yes, glad you found the link to this paper (osf.io/preprints/psyarxiv/dk3e2). And great that the results intrigued you! We proposed a few possible explanations on the rather counterintuitive results, but they're for future research to test.

Of course, thanks also to my wonderful co-authors who kindly allowed me to work with their datasets and contributed to the strong discussion in the pre-print: @eeskevanroekel.bsky.social , Sarah , Gillian, and Jolien.

A big thank you to Maaike , @loespouwels.bsky.social , Jacqueline , and @domimaciejewski.bsky.social . My supervisors guided me from the conceptualization to writing of this paper. And they were super patient with my long detour of working on the ER variability paper😉

But overall, this well-powered study shows ED and ER variability temporally negatively influence each other in 👥 daily lives. Results prompt reconsideration of existing assumptions on how emotion differentiation facilitates emotion regulation.

Research on ER variability is still in its infancy. So, we don't know yet how variability is different developmentally, and how it relates to other emotion (regulation) characteristics, and context.

2) ED and ER variability compete for similar resources outside the emotion system, so that when one process is more active, the other is less so. E.g., if both need effort and lead to fatigue, so it's difficult to have both high.

Possible explanations? 1) ED and ER variability may both serve to dampen emotion intensity. So, if one process already does the work, the other doesn't need to in the subsequent moment, hence a lower value.

It is also true reciprocally: When the more 👥 deviated from their typical emotion regulation strategies (i.e., the higher their emotion regulation variability), the less they differentiated their emotions at the next moment.

Surprise finding! When 👥 can better differentiate their emotions, they have LOWER emotion regulation variability the next moment, i.e., more stable in their use of emotion regulation strategies.

👥 also reported ~5 positive and ~5 negative emotions in each ESM beep. From these we calculated +&- momentary ED: The more intensities of emotions are deviating in the same direction (i.e., positively or negatively) with regard to a person’s mean, the lower ED it is, vice versa.

Adolescents (👥) reported their use of ~6 ER strategies since the last beep. We calculated momentary ER variability as Bray-Curtis dissimilarity (and its subcomponents). It shows how deviated 👥 are from their usual pattern of ER. Details in 1st tweet link.

Preprint: osf.io/preprints/psyarxiv/dk3e2. This is pre-registered. N👥=750+ & 25000+ESM obs from 5 datasets. Thanks Tilburg #Emotions2023 conference where I met amazing researchers also interested in emotion dynamics and that are now also co-authors of this paper! 🧵 below...

tldr: NO. We found an unexpected NEGATIVE reciprocal influence between ED and ER variability, regardless of the types of variability (intensity or switching) or emotions (positive or negative). The benefits and interplay between ED and ER variability are not that straightforward.

My first PhD paper on emotion regulation (ER) variability is published in Emotion! doi.org/10.1037/emo0001344
With this, we ask: Within adolescents, does emotion differentiation (ED) increase ER variability? Knowing what one feels informs ways to adjust ER, right?

Does it mean a similar weighted treatment is needed for the dissimilarity index when we adopt the weighting treatment for SD?

Hi Ruben, it's public now: osf.io/vzh2n/ & github.com/taktsun/ERV_...
Actually, upon further thoughts, I think Bray-Curtis dissimilarity, while by formula takes into consideration the sum of values, suffers the same instability of values when it is close to the lower bound.

Hi Ruben, it's public now: osf.io/vzh2n/ & github.com/taktsun/ERV_...
Actually, upon further thoughts, I think Bray-Curtis dissimilarity, while by formula takes into consideration the sum of values, suffers the same instability of values when it is close to the lower bound.