本研究以安德森的信息整合理论为理论基础、以儿童的生活事件为实验任务，随机选取5-6岁、7-8岁、11-12岁儿童和大学生四个年龄段的被试162名，采取实验法和观察法相结合的方式，深入探讨了儿童期望值判断的发展特点及其对风险决策的影响。本研究分研究一、研究二和研究三三个部分：研究一考察儿童在单奖励模型中的期望值判断发展特点。结果表明：（1）在单奖励模型中，5-6岁儿童就能直觉进行期望值判断，概率是其考虑的最重要因素；但当奖励很大时，奖励（价值）显著影响其期望值判断。（2）在单奖励模型中，7-8岁儿童能使用乘法策略进行期望值判断，但不稳定。（3）在单奖励模型中，11-12岁儿童能比较稳定地使用乘法策略进行期望值判断，期望值判断水平甚至比大学生更客观。研究二考察儿童在双奖励模型中的期望值判断发展特点。结果表明：（1）在双奖励模型中，所有被试的期望值都有向中间集中的趋势，其中5-6岁儿童向中间集中的趋势非常明显，7-8岁儿童向中间集中的趋势减弱，11-12岁儿童的期望值与标准期望值比较接近。（2）在双奖励模型中，部分5-6岁、7-8岁儿童能直觉进行期望值判断， 11-12岁儿童期望值判断水平有很大提高、与标准期望值最接近，大学生明显高估期望值。（3）在双奖励模型中，对于每个可能的结果，小部分5-6岁儿童能使用乘法策略，但很不稳定；多数7-8岁儿童不能使用乘法策略；11-12岁儿童和大学生能稳定地使用乘法策略进行期望值判断。（4）在双奖励模型中，在整合两个可能的结果时，5-6岁儿童基本上不能使用加法策略，7-8岁儿童不能稳定地使用加法策略，11-12岁儿童和大学生主要使用加法策略。研究三考察儿童的风险决策发展特点以及期望值判断对风险决策的影响。结果表明；（1）在获得情境下，儿童选择冒险决策和保守决策的数量相当；在失去情境下，儿童倾向于冒险决策以避免损失。（2）在获得情境下且单奖励模型中，儿童的期望值越高，越可能选择保守决策。在失去情境且单奖励模型中，当期望值很高或很低时，儿童的冒险决策稍多于保守决策，差异不明显；当期望值处于中间位置时，儿童的冒险决策明显多于保守决策。（3）在获得情境下且双奖励模型中， 11-12岁组儿童的冒险决策明显多于保守决策，其余儿童的冒险策略与保守策略数量相当；在失去情境且双奖励模型中，各个年龄阶段被试的期望值判断与风险决策的关系比较复杂。

Three experiments used Information Integration Theory and children’s daily events to study how children judge expected value and how children make risk decision. Five-six years old, seven-eight years old, eleven-twelve years old and undergraduates were investigated.The purpose of experiment 1 was to explore how children judge expected value of simple gambles. The results indicated: (1) five-six years old children can judge expected value intuitively; (2) seven-eight years old children can judge expected value with multiplicative strategy, however, the use of the strategy wasn’t stable;(3) eleven-twelve years old children can judge expected value with multiplicative strategy stably, even the expected value was more impersonal.The purpose of experiment 2 was to explore how children judge expected value of complex gambles. The results indicated: (1) Expected value of all the participants was coming the middle in contrast to normal expected value, especially the 5-6 years old children; (2) Part of 5-6 years old and 7-8 years old children can judge expected value intuitively, 11-12 years old children’s expected value was close to normal, undergraduates’ expected value was above normal obviously; (3) Within a single outcome, small part of 5-6 years old children used multiplicative strategy, most 7-8 years old children couldn’t use multiplicative strategy, 11-12 years old children and undergraduates can use it stably; (4) Across multiple outcome, 5-6 years old children couldn’t use additive strategy, 7-8 years old children used additive strategy but not stable, 11-12 years old children and undergraduates used additive strategy mainly.The purpose of experiment 3 was to explore how children make risk decision. The results indicated: (1) In the domain of gains, risk decision was equivalent to riskless decision, in the domain of losses, the participants made more risk decision to avoid loss; (2) In the domain of gains and simple gambles, the higher the expected value the more the riskless choices; in the domain of losses and simple gambles, when the expected value was high or low risk choices were equivalent to riskless choices, when the expected value was in the medium risk choices were more than riskless choices; (3) In the domain of gains and complex gambles, 11-12 years old children’s risk choices were more than riskless , other participants’risk choices and riskless choices were equal; in the domain of losses and complex gambles, the relation of expected value and risk decision was complicated.