To summarize, all children interacted with the experimenter and actively participated in the imitation task. There was evidence of improvement in performance from baseline to recall as would be expected with attention to, and memory for, the actions that were modeled by the experimenter. All participants evidenced a decrease in performance as the difficulty of the task increased, as would be expected. When the maltreated children were compared to the nonmaltreated children in a 2-group design, there was no statistically significant difference in performance. However, when the maltreated group was divided into two subtypes of either neglected or abused, and performance was compared in a 3-group design, it was revealed that the neglected children experienced deficits in performance relative to abused children. For production of target actions, the neglected children's performance trended toward significance when compared to the nonmaltreated children's performance. However, there was no significant difference between the performance of the abused children and the nonmaltreated children for either production of target actions or productions of ordered pairs. The children in this longitudinal study were assessed previously at 12 months of age in a mother-child play situation (Valentino et al., 2006). Interactions during structured play between mother and child were evaluated for maternal directives and child responses. Interestingly, the difference in social interactions that was most reliable was the finding that the abused children imitated their mothers more often than did the nonmaltreated children. There was no difference between the imitative behaviors of the neglected children and the abused or nonmaltreated children. The researchers note that by imitating their mothers, the abused children might be attempting to prevent further abusive incidents. Limit setting behaviors of the mothers in response to child initiations were positively related to the children's imitative behaviors. Thus, it would appear that maternal negative feedback to child-initiated behaviors is related to an increase in imitative behaviors that are most likely met with positive reinforcement. The continued pursuit of this positivity may impede the development of self-initiated behaviors; delayed development of self-initiated behavior has been linked to disorders of social competence (Landry, Smith, Miller-Loncar, & Swank, 1998). However, imitation has long been known to be a mechanism of learning (Piaget, 1962) and has become an accepted tool for assessment of declarative memory (Bauer, 2004). Whereas the adaptation to abuse posited by Valentino et al. (2006) may be detrimental to social development, our data for this same sample indicate that the reliance on imitative behavior exhibited by the abused children may afford them an advantage at 21 months of age in imitation paradigms. The neglected children are thus at a disadvantage relative to the abused children in the study reported here in that they were not reinforced by mothers for imitative behavior. It is important to note that all children in this sample were from low-income homes. Scores on these events for both target actions and ordered pairs are higher in samples of higher SES children (e.g., Bauer et al., 2000). Thus, the low SES of the families affected performance across the groups. It is possible that the factor responsible for the difference between the abused group and the neglected group is resilience in the face of poverty. Resilience is the ability to recover following a traumatic event or adversity (Masten, 2001), and has been related to child characteristics, such as general intelligence (Masten et al., 1988). It has been proposed that neural plasticity may be responsible for this recovery (Cicchetti & Curtis, 2006). Alternatively, as has been detailed earlier in this chapter, the advantage afforded abused children could arise from the strengthening of neural pathways. It would be adaptive to develop exceptional event memory so as to avoid the events that lead to abuse. Mechanisms of plasticity are responsible for the laying down of memories (Aimone, Wiles, & Gage, 2006). Thus, the higher performance seen in the abused group could be related to a preservation of brain plasticity that facilitates resilience in the face of poverty, stress, and/or trauma. Plasticity in the brains of the neglected children may be lost due to the lack of stimulation, leaving them more vulnerable to the stress of poverty and neglect. In conclusion, maltreated children have often been studied as a single group. However, it is becoming clear from research conducted by our group and others that the subtypes of maltreatment may have different developmental sequelae. It is important that we understand the differential pathways involved in the development of abused versus neglected children. As discussed in other chapters of this volume, the imitation paradigm has emerged as a valuable tool in the identification of at-risk infants and toddlers. With the data reported here, it is evident that data from the elicited imitation procedure utilized herein differentiates between the subtypes of maltreatment. Research must be conducted to further elucidate the correlates of resilience in toddlers who have been abused. A longitudinal investigation would enable investigation of the questions of continuity of the observed increase in imitative behavior and whether increased imitation has a detrimental social effect while exerting a bolstering cognitive effect.