Innovative Methods for Affectivity Profiling : Quantitative Semantics
Garcia, Danilo LU
and Sikström, Sverker
LU
(2023)
p.67-88
- Abstract
Background: Affectivity has been suggested as a complex adaptive meta-system composed of positive affect and negative affect, two independent but interrelated markers of well-being, that can be represented as four distinct affective profiles: self-fulfilling (high positive affect/low negative affect), high affective (high positive affect/high negative affect), low affective (low positive affect/low negative affect), and self-destructive (low positive affect/high negative affect). This model has been extensively studied during the last two decades and operationalized through well-developed self-report inventories (e.g., the Positive Affect Negative Affect Schedule, PANAS) that use fixed items and rating scales (e.g., 5-point Likert... (More)
Background: Affectivity has been suggested as a complex adaptive meta-system composed of positive affect and negative affect, two independent but interrelated markers of well-being, that can be represented as four distinct affective profiles: self-fulfilling (high positive affect/low negative affect), high affective (high positive affect/high negative affect), low affective (low positive affect/low negative affect), and self-destructive (low positive affect/high negative affect). This model has been extensively studied during the last two decades and operationalized through well-developed self-report inventories (e.g., the Positive Affect Negative Affect Schedule, PANAS) that use fixed items and rating scales (e.g., 5-point Likert scales). However, this type of self-reports preimpose which feelings and emotions that are part of people’s affective experiences. In other words, they do not allow people to freely describe their own emotional well-being at all levels of health: physical, psychological, and social. Aim: In this Chapter, we use computational methods (i.e., quantitative semantics) to study how the meaning of words that people freely generate to describe their physical, psychological, and social well-being can be quantified to measure positive affect and negative affect. We then use these semantic estimates of affect for affectivity profiling. The “semantic” affective profiles were validated by (1) mapping them to people’s self-reported affectivity (i.e., PANAS-scores) and (2) by investigating which words significantly discriminate between individuals with distinct “semantic” affective profiles. Method: Participants (N = 523) were asked to describe their physical, psychological, and social well-being using five words for each well-being domain. In addition, participants self-reported their affectivity through the PANAS. The affectivity profiling was done using the median splits method with participants’ PANAS-scores and then with the language-based semantic estimates of positive affect and negative affect. Results: The diagnostic analyses showed that people used significantly different and semantically distinctive words to describe their well-being: physical well-being using material words and agentic verbs; psychological well-being using emotional adjectives; and social well-being using communal words and nouns. Regarding our main analyses, the semantic estimates of positive affect predicted the positive affect PANAS-scores (physical: r = 0.31, psychological: r = 0.33, and social: r = 0.19) and the semantic estimates of negative affect predicted the negative affect PANAS-scores (physical: r = 0.27, psychological: r = 0.41, and social: r = 0.17). Moreover, individuals with a self-fulfilling profile used agentic (e.g., power, strong, intelligent, determined), communal (e.g., loving, kind, caring, support, helping, sharing, honest), and transcendental (e.g., spiritual, creative, and meaningful) words when describing their well-being, while those with a self-destructive profile used words that mirrored an outlook of separateness and poorly developed character and health (e.g., depressed, sad, anxious, guilty, isolated, lonely, pain, overweight, unhealthy, sick). Finally, individuals with high affective and low affective profiles used neutral words (e.g., aging, normal, average) and function words (e.g., on, the, and at), but also words mirroring extrovert (high affective: work, fun, and co-worker) and introvert tendencies (low affective: connected, close, and respect). Conclusions: We conclude that quantitative semantics is a promising method for affectivity profiling that should be further investigated. More specifically, the semantic estimates of affectivity derived from people’s own descriptions of their physical, psychological, and social well-being capture better the true nature of affect-after all, the affectivity dimensions involve more mood-related and social features and are not purely a measure of unconscious emotions or only certain emotions, but rather a conscious apprehension of the full range of our affective experience, which is an independent and interactive part of all well-being domains.
(Less)
- author
- Garcia, Danilo
LU
and Sikström, Sverker
LU
- organization
- publishing date
- 2023-01-01
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- keywords
- Affective profiles model, Affectivity, Latent Semantic Analysis, Negative affect, Positive affect, Quantitative semantics
- host publication
- The Affective Profiles Model : 20 Years of Research and Beyond - 20 Years of Research and Beyond
- pages
- 22 pages
- publisher
- Springer International Publishing
- external identifiers
-
- scopus:85206020752
- ISBN
- 9783031242199
- 9783031242205
- DOI
- 10.1007/978-3-031-24220-5_4
- language
- English
- LU publication?
- yes
- id
- 1f1f03a2-4af7-4300-85aa-ff2b5b57d72d
- date added to LUP
- 2024-12-16 13:29:04
- date last changed
- 2025-07-01 05:25:19
@inbook{1f1f03a2-4af7-4300-85aa-ff2b5b57d72d,
abstract = {{<p>Background: Affectivity has been suggested as a complex adaptive meta-system composed of positive affect and negative affect, two independent but interrelated markers of well-being, that can be represented as four distinct affective profiles: self-fulfilling (high positive affect/low negative affect), high affective (high positive affect/high negative affect), low affective (low positive affect/low negative affect), and self-destructive (low positive affect/high negative affect). This model has been extensively studied during the last two decades and operationalized through well-developed self-report inventories (e.g., the Positive Affect Negative Affect Schedule, PANAS) that use fixed items and rating scales (e.g., 5-point Likert scales). However, this type of self-reports preimpose which feelings and emotions that are part of people’s affective experiences. In other words, they do not allow people to freely describe their own emotional well-being at all levels of health: physical, psychological, and social. Aim: In this Chapter, we use computational methods (i.e., quantitative semantics) to study how the meaning of words that people freely generate to describe their physical, psychological, and social well-being can be quantified to measure positive affect and negative affect. We then use these semantic estimates of affect for affectivity profiling. The “semantic” affective profiles were validated by (1) mapping them to people’s self-reported affectivity (i.e., PANAS-scores) and (2) by investigating which words significantly discriminate between individuals with distinct “semantic” affective profiles. Method: Participants (N = 523) were asked to describe their physical, psychological, and social well-being using five words for each well-being domain. In addition, participants self-reported their affectivity through the PANAS. The affectivity profiling was done using the median splits method with participants’ PANAS-scores and then with the language-based semantic estimates of positive affect and negative affect. Results: The diagnostic analyses showed that people used significantly different and semantically distinctive words to describe their well-being: physical well-being using material words and agentic verbs; psychological well-being using emotional adjectives; and social well-being using communal words and nouns. Regarding our main analyses, the semantic estimates of positive affect predicted the positive affect PANAS-scores (physical: r = 0.31, psychological: r = 0.33, and social: r = 0.19) and the semantic estimates of negative affect predicted the negative affect PANAS-scores (physical: r = 0.27, psychological: r = 0.41, and social: r = 0.17). Moreover, individuals with a self-fulfilling profile used agentic (e.g., power, strong, intelligent, determined), communal (e.g., loving, kind, caring, support, helping, sharing, honest), and transcendental (e.g., spiritual, creative, and meaningful) words when describing their well-being, while those with a self-destructive profile used words that mirrored an outlook of separateness and poorly developed character and health (e.g., depressed, sad, anxious, guilty, isolated, lonely, pain, overweight, unhealthy, sick). Finally, individuals with high affective and low affective profiles used neutral words (e.g., aging, normal, average) and function words (e.g., on, the, and at), but also words mirroring extrovert (high affective: work, fun, and co-worker) and introvert tendencies (low affective: connected, close, and respect). Conclusions: We conclude that quantitative semantics is a promising method for affectivity profiling that should be further investigated. More specifically, the semantic estimates of affectivity derived from people’s own descriptions of their physical, psychological, and social well-being capture better the true nature of affect-after all, the affectivity dimensions involve more mood-related and social features and are not purely a measure of unconscious emotions or only certain emotions, but rather a conscious apprehension of the full range of our affective experience, which is an independent and interactive part of all well-being domains.</p>}},
author = {{Garcia, Danilo and Sikström, Sverker}},
booktitle = {{The Affective Profiles Model : 20 Years of Research and Beyond}},
isbn = {{9783031242199}},
keywords = {{Affective profiles model; Affectivity; Latent Semantic Analysis; Negative affect; Positive affect; Quantitative semantics}},
language = {{eng}},
month = {{01}},
pages = {{67--88}},
publisher = {{Springer International Publishing}},
title = {{Innovative Methods for Affectivity Profiling : Quantitative Semantics}},
url = {{http://dx.doi.org/10.1007/978-3-031-24220-5_4}},
doi = {{10.1007/978-3-031-24220-5_4}},
year = {{2023}},
}