TL;DR
This paper introduces the first large-scale affective norms for Chinese words specifically designed to capture age-related differences. The AANC database, containing ratings for 2,061 four-character Chinese words across valence, arousal, dominance, and familiarity, reveals distinct emotional perception patterns between older and younger adults, particularly highlighting the "positivity effect" in older individuals.
What Problem Does This Paper Address?
Understanding how the affective meaning of words influences cognitive processes like recognition, attention, and memory is crucial across various fields, from psychology to natural language processing. However, existing affective norm databases often lack age-specific data, meaning they might not accurately reflect how different age groups perceive emotional stimuli. This is particularly problematic for Chinese, the world's most spoken language, where no comprehensive affective norms for older adults were available. The authors aimed to bridge this gap by creating a resource that accounts for age-related differences in emotional word perception, acknowledging that cultural and life-experience factors can shape these perceptions.
Methods
The researchers collected ratings for 2,061 four-character Chinese words from two distinct age groups: younger adults (16-40 years) and older adults (55+ years). Participants rated each word on four dimensions: valence (pleasantness), arousal (intensity), dominance (control), and familiarity. The data collection involved a computer-based questionnaire, with participants rating words on nine-point scales. To ensure data quality, rigorous outlier detection and trimming procedures were implemented. The study also analyzed correlations between these dimensions and explored how these relationships differed between age groups.
Datasets
The primary contribution of this paper is the creation of the Age-Related Affective Norms for Chinese Words (AANC) database. This new dataset comprises ratings for 2,061 four-character Chinese words, with data collected from both older and younger adult participants. The AANC database includes mean ratings and standard deviations for valence, arousal, dominance, and familiarity for each word, broken down by age group. It also incorporates word frequency, complexity, and character complexity data from the Chinese Lexicon. This database is intended to serve as a valuable resource for researchers, differing from existing Chinese affective lexicons by specifically including older adult ratings and focusing on four-character words, which often carry richer meanings.
Key Insights and Ideas
- The "Positivity Effect" is Alive and Well in Chinese: Older adults show a clear preference for processing positive information, rating positive words as more arousing and less controllable, while rating negative words as less arousing and more controllable compared to younger adults. This reinforces the well-established "positivity effect" in aging.
- Age Shapes Emotional Perception Beyond Valence: While both age groups largely agree on the pleasantness (valence) of words, significant age-related differences emerge in arousal, dominance, and familiarity ratings, suggesting that age influences how intensely words are perceived, how much control they exert, and how familiar they are.
- Cultural Nuances Matter: The study highlights that affective norms are language- and culture-specific. Direct translation of affective resources from other languages may not accurately capture the emotional nuances for Chinese speakers, underscoring the need for culturally relevant datasets like AANC.
- Four-Character Words Offer Richer Emotional Data: The focus on four-character words, which often function as idioms or convey complex meanings, provides a more nuanced understanding of emotional language compared to shorter word sets.
Key Results
- The AANC database contains ratings for 2,061 four-character Chinese words.
- Younger adults rated words significantly higher for dominance and familiarity than older adults.
- Older adults rated words slightly higher for arousal than younger adults.
- No significant age differences were found in mean valence ratings across the entire dataset.
- Correlations between older and younger adults' ratings were very high for valence (r = 0.95) but lower for arousal (r = 0.73), dominance (r = 0.62), and familiarity (r = 0.53).
- Linear mixed-effects models revealed significant age-by-polarity interactions, indicating that older adults rated positive words as more arousing and less controllable, and negative words as less arousing and more controllable, compared to younger adults.
Limitations and Future Work
The authors acknowledge several limitations:
- Limited Word Length: The study focused only on four-character words due to time and funding constraints, excluding two- and three-character words.
- Gender Differences Not Detailed: Small sample sizes for each gender prevented a detailed analysis of gender differences.
- Discrete Emotions Not Explored: The study focused on valence, arousal, and dominance, not exploring discrete emotions (e.g., happiness, anger) which can also influence information processing.
Future work could expand the database to include two- and three-character words, conduct detailed gender analyses, and incorporate ratings for discrete emotional categories. Further research is also needed to explore the underlying mechanisms driving these age-related differences in emotional perception, potentially linking them to life experience, cultural factors, and biological changes.
Summary updated