元刻板印象是个体关于外群体成员对其所属群体(内群体)所持刻板印象的信念。它与外群体刻板印象在内容上较为一致，但比外群体刻板印象更消极。元刻板印象对群际关系具有极大的负面影响，认知上会使人们更加关注威胁信息，情绪上会增加群际焦虑和偏见，行为上会降低群际互动的意愿与互动质量。然而，以往研究却较少关注元刻板印象是否会对群际间的亲社会行为（如群际帮助行为）产生消极影响。在本研究中，我们以刻板印象内容模型、群际帮助的地位关系模型、系统公正理论和共同内群体认同模型为主要的理论依据，试图以不同群体大学生为例来建构元刻板印象的内容与结构，在此基础上探索不同维度的元刻板印象对自主定向与依赖定向的群际帮助的不同影响及其潜在的中介变量，并尝试利用元刻板印象的特征和共同内群体认同来降低元刻板印象对群际帮助的消极影响。 研究1以汉、藏族大学生为例来考察少数人群体与多数人群体之间的元刻板印象的内容与结构。结果发现，从内容上看，虽然少数人群体的元刻板印象的内容整体较为积极，但他们高估了消极元刻板印象的消极性，表现出一种“过于悲观”的倾向；虽然多数人群体元刻板印象的内容整体偏消极，但他们却高估了元刻板印象的积极性，表现出一种“盲目乐观”的倾向。从结构上看，元刻板印象包含三个维度：道德、热情与能力/支配性，其中，汉族大学生元刻板印象的主要特点是低道德、低热情与高支配性，而藏族大学生元刻板印象的主要特点是高道德、高热情与低能力。 研究2分别以藏族大学生（研究2a）、彝族大学生（研究2b ）和夜大学生（研究2c）为被试，采用测量（研究2a）或操纵（研究2b、2c）元刻板印象的方法，并结合多种群际帮助的测量方式（想象的一般帮助测量，如研究2a、2b；具体假设情境下的帮助测量，如研究2c），考察了元刻板印象对群际帮助的影响。结果发现，积极元刻板印象越高，被试越多地认为系统是公正的、外群体的帮助是自主定向的帮助（研究2a）；消极元刻板印象激活后，元刻板印象的消极度越高，低地位群体的被试会越多地认为系统是不公正的（研究2a、2b），并且会越多地认为外群体的帮助是依赖定向的帮助（研究2a、2c），越少地认为外群体的帮助是自主定向的帮助（研究2b ）；系统公正信念在元刻板印象与自主定向的帮助之间起中介作用（研究2a、2b ）。 研究3利用元刻板印象的特征及凸显共同内群体认同来降低元刻板印象对群际帮助的消极影响。研究3a发现，同时呈现积极与消极元刻板信息比单独呈现消极或单独呈现积极信息时，被试会更少地认为汉族人对藏族人的帮助是依赖定向的帮助。研究3b发现，共同内群体认同启动比单纯藏族人认同会让藏族人更多地认为外群体的帮助是自主定向的帮助。 本研究考察了少数人群体及多数人群体之间元刻板印象的内容与结构，及其对群际帮助的影响，并探索了两条降低元刻板印象对群际帮助所产生的消极影响的措施。这些研究结果对于群际关系的理论研究及在实践中如何促进群际关系的和谐、避免群际间的误解与冲突都有一定的意义。

Beliefs held by in-group members regarding how the in-group is stereotyped by out-groups are known as meta-stereotypes. They are consistent with the stereotypes that the out-groups have of the in-group in content. However, people tend to overestimate the negativity of meta-stereotypes. It has great negative impact on intergroup relationships, for example, when negative meta-stereotypes are salient, people cognitively pay more attention to threaten information, emotionally have higher level of intergroup anxiety and prejudice, are less willing to interact with an out-group member, and exhibit less favourable behaviour and more unfavourable behaviour. However, it is unclear whether meta-stereotypes would have great negative impact on intergroup prosocial behavior, such as intergroup helping. Therefore, in the present research, based on the stereotype content model, the intergroup helping as status relations model, the system justification theory and the Common Ingroup Identity Model, we took Han and Tibetan college students as examples to explore the content and construct of their meta-stereotypes in Study 1, conducted three studies in total to examine the effect of meta-stereotypes on intergroup helping , as well as the potential mediating variable in Study 2, and attempted to take advantage of the characteristics of meta-stereotypes and the common ingroup identification to reduce the negative impact of meta-stereotypes on intergroup helping in Study 3. As mentioned, in Study 1, we took Han and Tibetan college students as examples to explore the content and construct of their meta-stereotypes. The results showed as follows: for the minority group, although the content of their meta-stereotypes is predominately positive, they are too pessimistic to overestimate the negativity of negative meta-stereotypes; for the majority group, although the content of their meta-stereotypes is predominately negative, they are too optimistic to overestimate the positivity of meta-stereotypes. There are three dimensions for the meta-stereotypes: the morality dimension, the warmth dimension, and the competence or dominance dimension. The meta-stereotypes of Han college students are characterized by low morality, low warmth, and high dominance, whereas the meta-stereotypes of Tibetan college students are characterized by high morality, high warmth, and low competence. In Study 2, 78 Tibetan college students (Study 2a), 65 Yi college students (Study 2b) and 101evening college students (Study 2c) participated in this study. We measured (Study 2a) or manipulated participants’ meta-stereotype (salient or non-salient, Study 2b and Study 2c), measured their evaluation of outgroup helping with de-contextualizing helping (imaging an outgroup helping scenario, Study 2a and 2b) or a specific hypothetical scenario of outgroup helping (Study 2c) to examine the effect of meta-stereotypes on intergroup helping. We found that, the higher participants perceived the positive meta-stereotypes, the higher their system justification belief and the more they believed that outgroup provided them with autonomy-oriented help (Study 2a); the higher participants perceived the negative meta-stereotypes, the lower their system justification belief (Study 2a and Study 2b), the more they believed that outgroup provided them with dependency-oriented help (Study 2a and Study 2c) and the less they believed that outgroup provided them with autonomy-oriented help (Study 2b). Moreover, the relationship between meta-stereotypes and autonomy-oriented was mediated by system justification belief. Furthermore, we conducted study 3 to explore how to reduce the negative impact of meta-stereotypes on intergroup helping. In Study 3a, we found that, compared with when presenting only the positive or negative stereotypes information, when presenting the positive and negative stereotypes information at the same time, participants were less likely to perceive that outgroup provided them with dependency-oriented help. In Study 3b, we found that, compared with when ingroup identity was salient, when common ingroup identity was salient, participants were more likely to perceive that outgroup provided them with autonomy-oriented help. In summary, the present research attempted to explore the content and construct of meta-stereotypes and examine its negative impact on intergroup helping, and how to reduce this negative impact. These results have significant implications not only for the theoretical research of the intergroup relations, but also in practice, for how to promote intergroup harmony and avoid misunderstandings and conflicts between different groups.