Real-world social interactions are complex and multimodal, involving visual, semantic, and prosodic cues. However, much of the existing research on social information processing relies on simplified stimuli and controlled experimental designs. This study aimed to address this gap by investigating the effects of more naturalistic social feedback (static photos and dynamic videos) on motivation, learning, and socio-emotional processing, with a specific focus on potential age-related differences. The researchers hypothesized that dynamic feedback would be a stronger motivator than static feedback, and that age would influence both motivational and emotional responses. They employed a social incentive delay (SID) task, a variation of the monetary incentive delay (MID) task, where faster response times (RTs) indicate increased motivation within a reinforcement learning context. Prior research has shown that anticipation of reward and punishment, regardless of whether the feedback is textual, pictorial, or video-based, influences motivation and behavior, typically leading to faster RTs when anticipating positive or negative feedback compared to neutral feedback. However, the relative effectiveness of dynamic video feedback compared to static photos remains unclear. Evaluative learning, the process of transferring emotional valence from a valent stimulus to a previously neutral stimulus, is also relevant. The study further considered age-related changes in reward processing, focusing on potential learning impairments in older adults. Studies have shown age-related changes in emotional memory and valence judgments of emotional faces, particularly with static stimuli, but the impact of dynamic stimuli is less clear. The socioemotional selectivity theory (SST) provides a framework for understanding age-related changes in emotional processing, suggesting that older adults might show a positivity effect, prioritizing emotional goals due to limited remaining lifetime. This study aimed to examine the interaction between dynamic social feedback and age-related changes in socio-emotional processing within a motivational setting. Finally, the study's implications for educational contexts, specifically the use of dynamic social feedback in digital learning tools and the potential age-related differences in responses to such feedback, are considered.

Existing literature highlights the motivational power of social rewards and punishments, influencing behavior across various feedback modalities (text, photos, videos). Studies using the MID and SID tasks show faster RTs when anticipating high rewards or positive/negative feedback compared to neutral feedback. While video feedback, particularly those involving biological motion and social presence, offers more engaging properties, the direct comparison of video versus photo feedback in SID tasks is limited. Evaluative learning, the transfer of emotional valence from valent to neutral stimuli, has been demonstrated in aversive conditioning paradigms, but its role in social reward/punishment contexts and age-related effects remains under-investigated. Age-related changes in reward processing and emotional responses are well-documented. Neuroimaging studies suggest that aging is associated with compromised updating of predictive value, potentially linked to dopaminergic mesolimbic system degeneration. While motivational effects (faster RTs) are preserved in older adults in less demanding tasks, more complex learning tasks reveal age-related impairments. Older adults often demonstrate a positivity effect, showing reduced negative affect, better memory for positive stimuli, and decreased amygdala activation to negative stimuli. This positivity effect has been mainly observed using static stimuli, but the impact of dynamic stimuli requires further investigation. The interplay between age, dynamic social feedback, and motivational settings remains an open question.

The study combined new and previously collected data to compare the effects of static (photos) and dynamic (videos) social feedback in a SID task across two age groups (young adults: 18–35 years; older adults: 50–84 years). Each study involved a simple RT task followed by the SID task, where participants responded to a target preceded by a cue signaling the feedback type (neutral, positive, negative). The SID-Photo task used photos of volunteers displaying emotional mimics and gestures, while the SID-Video task used videos with audio of volunteers conveying the same feedback types with both mimics, gestures and speech. The SID task used a staircase method to adjust difficulty, aiming for an average hit rate of 83%. Study 2 (SID-Video) included additional measures: an explicit cue rating task assessing subjective emotional valence of cues after each trial and a Single Category Implicit Association Test (SC-IAT) to measure implicit cue valence. Both studies concluded with a rating task assessing the emotional valence and credibility of the feedback videos. For the video stimuli, the researchers recorded a new set of videos with two actors (one male and one female). These videos were pre-tested (n=76 participants) to evaluate emotional valence and credibility, selecting videos based on specific criteria to ensure suitability for the study. The simple RT task (50 trials) served to establish a baseline RT for each participant, used to adjust the difficulty of the SID task. Statistical analyses included permutation F-tests and t-tests, both frequentist and Bayesian methods were employed. Permutation tests controlled for multiple comparisons using Bonferroni correction. Bayesian ANOVAs were also utilized for model comparison.

In both the SID-Photo and SID-Video tasks, response times (RTs) were faster for positive and negative cues compared to neutral cues in both young and older adults. This result indicates that anticipation of both positive and negative social feedback accelerated response times regardless of age, and regardless of whether the feedback was given via photos or videos. There was no significant difference in the effect on RTs between static and dynamic feedback. There was a main effect of age on RTs, with older adults having slower RTs than young adults. Analysis using a 2 × 2 × 3 mixed Bayesian ANOVA provided extreme evidence for the main effects model including the factors of age and condition (neutral, positive, negative), strongly supporting that there was no interaction between age and condition. In the SID-Video task, the explicit cue rating task revealed a significant main effect of condition, with cues associated with positive feedback rated as more positive than neutral and negative cues. A significant interaction effect indicated that this difference was smaller in older adults. The SC-IAT showed a significant main effect of condition, with positive cues showing a more positive association than negative cues. However, there was no main effect of age on implicit cue valence. Emotional valence ratings of the feedback videos revealed a significant main effect of age, with older adults rating negative feedback videos as less negative than young adults, consistent with the positivity effect in older adults. There was a significant interaction effect between age and condition, with the age difference being most pronounced for negative feedback. Credibility ratings of the feedback videos revealed a significant interaction between age and condition. Older adults rated neutral and positive feedback videos as more credible than young adults. Partial correlations showed a significant correlation between explicit cue valence and feedback video valence in the positive condition, but not in the neutral or negative conditions. No significant correlation was found between implicit cue valence and feedback video valence.

The results confirm that anticipation of positive and negative social feedback motivates faster responses in both young and older adults, consistent with previous research. The lack of a significant difference between the effects of static and dynamic feedback suggests a potential context-dependent value normalization, where the relative value of each feedback type is normalized within the context of the experimental task. The positivity effect in older adults, observed in the emotional valence ratings of negative feedback videos, aligns with the existing literature on age-related changes in emotional processing. This effect might reflect cognitive change strategies, particularly reappraisal, for emotion regulation, allowing older adults to still be motivated to avoid negative feedback while managing emotional responses to negative social content. Evidence of evaluative learning was found, with a significant transfer of positive valence from positive feedback videos to predictive cues, but not negative valence. This could be influenced by the different proportions of trials in each condition. The difference between explicit and implicit measures of cue valence might reflect the influence of higher-order cognitive processes (explicit) versus automatic responses (implicit). The absence of significant correlations between implicit cue valence and feedback video valence suggests that these automatic processes are less directly influenced by the valence of the feedback stimuli. Potential age biases based on the age of actors were discussed. The study considered potential biases from different experimenters across studies but acknowledges these as possible limitations.

This study demonstrates the motivational power of positive and negative social feedback across the lifespan, with no added benefit of dynamic over static stimuli. Evaluative learning was present for positive feedback, and age-related positivity bias was found. These findings highlight the interplay of motivation, learning, and socio-emotional processing across different ages. Future studies should investigate these factors within subjects using more precisely matched stimuli and a more balanced proportion of positive and negative feedback. The implications of the findings for the development of educational tools are promising.

The study's between-subjects design, with separate groups for the SID-Photo and SID-Video tasks, might have introduced context effects, potentially masking some of the effects of feedback type. Also, the different durations of photo and video feedback stimuli and the unbalanced trial numbers in the three conditions (neutral, positive, negative) represent limitations of the study. The study acknowledges the influence of external factors such as the COVID-19 pandemic, which could have affected participant motivation and behavior during data collection. A within-subjects design may provide clearer results, although potentially increasing task duration and associated fatigue. Future research should explore these limitations to refine the understanding of dynamic social feedback effects.