Parents are key stakeholders in the production of human capital, acting as primary investors in allocating resources for their child's education. These decisions are fundamental in shaping children's future outcomes. How do parents invest in their children's education based on their academic abilities? Do they reinforce the advantages of high-achieving children by making complementary investments? Alternatively, do they take a compensatory approach, directing more resources toward children with fewer advantages? Or do they make educational investments independently of their children's academic performance? Understanding these investment patterns holds both theoretical and policy significance. Theoretically, this question remains unresolved because a child's academic endowment is inherently endogenous--it can be impacted by parental investment--making the relationships estimated through quasi-experimental or correlational methods potentially spurious. From a policy standpoint, understanding these dynamics is crucial for designing effective social policies that address disparities arising from unequal parental resources, ensuring that all children, regardless of their home environment, have equal opportunities to thrive. This study offers robust experimental evidence on how parents allocate educational resources in response to their children's academic abilities within low-income multi-child families, using randomized information shocks as an exogenous factor. The author conducted a large-scale randomized controlled trial (RCT) involving 3,750 households with two children in grades 3 through 10 in rural China. Half of the households (N = 1,878) are assigned to the treatment group, where parents receive their children's test scores, while the control group (N = 1,872) does not. The author then analyzed the effects by comparing changes in parental educational investment among those with similar baseline beliefs about their children's abilities, but who receive either high- or low-ability signals. The central method builds on the concept proposed by Dizon-Ross (2021): if parents who receive high-ability signals increase their investments more than those who receive low-ability signals, it is inferred that parents are following a complementary investment approach. Conversely, if parents who receive low-ability signals increase their investments more, it is inferred a compensatory approach. Additionally, the author examines heterogeneous treatment effects by household socioeconomic status (SES), the gender composition within the family (girl-girl, girl-boy, boy-girl, and boy-boy), and the child's age.