From a conditional adaptation vantage point, early life caregiving adversity likely enhances aspects of cognition needed to manage interpersonal threats. Yet, research examining early life care and offspring cognition predominantly relies upon experiments including affectively neutral stimuli, with findings generally interpreted as “early-life caregiving adversity is, de facto, ‘bad’ for cognitive performance.” Here, in a Southeast Asian sample, we examined observed maternal sensitivity in infancy and cognitive performance 3 years later as preschoolers took part in three tasks, each involving both a socioemotional (SE) and non-socioemotional (NSE) version: relational memory (n = 236), cognitive flexibility (n = 203), and inhibitory control (n = 255). Results indicate the relation between early life caregiving adversity and memory performance significantly differs (Wald test = 7.67, (1), P = 0.006) depending on the SE versus NSE context, with maternal sensitivity in infancy highly predictive of worse memory for SE stimuli, and amongst girls, also predictive of better memory when NSE stimuli are used. Results concerning inhibitory control, as well as cognitive flexibility in girls, also tentatively suggest the importance of considering the SE nature of stimuli when assessing relations between the caregiving environment and cognitive performance. As not all approaches to missing data yielded similar results, implications for statistical approaches are elaborated. We conclude by considering how an adaptation-to-context framework approach may aid in designing pedagogical strategies and well-being interventions that harness pre-existing cognitive strengths.

Small‐scale eye‐tracking research lends support to behavioral studies of relational memory by 6 months of life. Here, in the largest eye‐tracking test of relational memory to date (n = 276), we replicate these findings and examine the impact of excluding data based on looking behavior characteristics at test. Past work examining infants' preferential looking toward arbitrary‐paired objects and scenes has excluded infants from analysis based upon “insufficient looking” at test. Yet, research suggests that variation in looking behavior may be associated with looking patterns during encoding, as well as trait‐like differences in visual and cognitive processing. Similar to past research, we observed evidence for relational memory among 6‐month‐olds. In keeping with past research, when infants were excluded based on “insufficient looking,” we observed evidence for relational memory only when infants were tested immediately. However, when exclusion criteria were relaxed, infants specifically demonstrated preferential looking during a presumably more difficult delay‐plus‐interference condition. Moreover, analyses revealed that looking behavior during encoding was associated with looking behavior at test. Together, results suggest that infants do possess rudimentary relational memory capabilities, but that experimenters' ability to detect these capabilities is influenced by both experimental conditions and individual differences in looking behavior.