In 2007, the Chinese Ministry of Education implemented a policy of tuition-free teacher education to improve the quality of basic education. Many undergraduates in teacher-training college do not pay tuition and lodging fees and receive cost of living bonuses supported by state revenue. However, often, these undergraduates cannot afford to make career choices, and return to their hometown and work in local secondary schools as middle school teachers for more than 10 years after their graduation. They are known as tuition-free teacher candidates. Currently, there are approximately 90,000 such candidates (Weng and Qi, 2014). However, many tuition-free teacher candidates do not like teaching (Zhang and Zheng, 2007; Wu et al., 2010). Some of them plan to reimburse the tuition in the future so that they can enjoy the freedom of occupation selection; others have no clear plan about their future work (Lian, 2007; Wang, 2010). Obviously, teacher candidates will perform better in their future careers if they want to be a teacher. Thus, it is important to know whether teacher candidates want to be middle school teachers. However, some students hide their real thoughts about preferred professions and report that they want to be teachers because of economic reasons and parents' orders (Zhang and Zheng, 2007). We attempted to identify the stereotypes about middle school teachers held by teacher candidates and then used them to predict whether they wanted to be middle school teachers. Furthermore, it may help to select appropriate teacher candidates who ardently want to be a teacher.

People usually use stereotypes to simplify cognitions about individuals or certain ways of doing things. A stereotype is a cognitive structure that includes a set of beliefs about personal attributes of people in a social group. These personal attributes include personality traits, feelings, physical appearances, preferences, and occupations (Ashmore and Boca, 1979; Wong, 2004; Kanahara, 2006). The contents of stereotypes have been widely investigated (Anderson and Sedikides, 1991; Edwin and Hub, 1999). The stereotype content model (SCM) proposed that stereotypes are based on two dimensions: warmth and competence. One may view a group as warm but not competent (e.g., the old are nice but weak) or as competent but not warm (e.g., the rich are cold but efficient) (Fiske et al., 2002). This model has been supported by some studies. For example, the SCM of immigrant subgroups predicts that perception centers on competence and warmth and relates to targets' perceived status and competition within society (Lee and Fiske, 2006). In addition, warmth and competence can reliably differentiate social group stereotypes in individualist and collectivist nations (Cuddy et al., 2009). However, some researchers still doubt whether the contents or structures of stereotypes are limited to warmth and competence (Zuo et al., 2006). Leach et al. (2007) suggested that the warmth dimension encompasses two distinct aspects called sociability and morality; they concluded that morality is more important than sociability and competence traits when giving positive in-group evaluations. It also should be noted that some researchers have tended to use the label “morality” in place of “warmth” (Vonk, 1996; Wojciszke, 2005). In addition, other components can also be found for stereotypes of Chinese people. For example, one study extracted 10 dimensions for impressions of Chinese people: morality, wealth, attitude, reason-politeness, friendship ties, skin color, temperament, sense of community, appearance, and male chauvinism (Sun, 2004). That is to say, the SCM (warmth and competence) may include necessary components of various stereotypes, rather than representing all stereotypes.

Consistent with the SCM model, warmth and competence are two necessary dimensions for teacher stereotypes. Teachers are stereotyped as being incompetent and warm (Carlsson and Björklund, 2010). Specifically, they are presented as being poorly adjusted (Erskine and Andrew, 1951) and incapable of anything but teaching (Lieberman, 1956). However, teachers are also evaluated as being warm, because they are amicable and have a deep love for students (Yuan et al., 2005). Additionally, some other features considered to befit teachers include being responsible, having profound knowledge, disseminating traditional culture or existing knowledge, and fostering social, economic, and cultural development (Lin, 1994; Yuan et al., 2005). However, teachers usually experience some degree of poverty and they often solve this problem by leaving the profession (Gerbner, 1963). The occupation of teaching is seen as having a feminine orientation; thus the male teacher is considered as low in status, poor in terms of public affairs, and as lacking assertiveness, firmness, strength, activity, and confidence by college students (O'Dowd and Beardslee, 1961). Consequently, we inferred that the SCM (warmth and competence) contained necessary components of teacher stereotypes, and the stereotypes may also include other specific attributes of teachers.

Occupational stereotypes can not only provide quick and easy assumptions that will affect the behavior of social group members (Yzerbyt and Demoulin, 2010), but also have been shown to influence individuals' career development and aspirations (Sue and Kirk, 1972; Leong, 1985). Consequently, it is valuable to identify stereotypes about teachers held by teacher candidates. We can infer that teacher candidates' stereotypes about teachers will have an important influence on their desire for work and professional behavior.

In the present study, we focused on teacher candidates' stereotypes toward Chinese middle school teachers. To better understand these stereotypes, we compared stereotypes about university and middle school teachers. In addition, previous studies suggested that positive and negative descriptions could simultaneously exist in stereotypes about a social group (Greenwald et al., 2002). In order to enhance the credibility of the present study, we explored whether the same structure would be found in positive and negative descriptions. Furthermore, a study reported that the stereotype of accountants held by accountants was significantly different from people in other occupations (Imada et al., 1980); thus, it is hypothesized that the stereotype about teachers held by tuition-free teacher candidates is different from general undergraduates. Finally, we computed correlation coefficients between undergraduates' willingness to be middle school teachers and their stereotypes about middle school teachers and predicted that students with positive middle school teacher stereotypes would be more likely to be middle school teachers.

The present study explored stereotypes about middle school and university teachers, comparing them between regular undergraduates and tuition-free teacher candidates. With both positive and negative descriptions, exploratory factor analysis revealed a consistent three-factor model (occupational cognition, personality, and emotion) for stereotypes of two teacher types (middle school and university). Our data also revealed some differences between teacher candidates and general undergraduates in the specific stereotypes of two teacher types. Students' stereotypes of middle school teachers have a significant correlation with their willingness to be middle school teachers.

The SCM focuses on two dimensions: competence and warmth (Fiske et al., 2002). The essence of competence is confidence, independence, competitiveness, and intelligence. The essence of warmth is tolerance, good-naturedness, and sincerity. This two dimensional hypothesis has already been verified by static assessment, dynamic change (Cuddy et al., 2004; Lin et al., 2005), and cross-cultural evidence (Fiske et al., 2002). In fact, competence and warmth can be summarized by occupational personality in the present study. Specifically, the words “enthusiastic,” “friendly,” and “facetious” and so on were positively related to warmth; the words “arrogant,” “autocratic,” and “severe” and so on were negatively related to warmth. The words “enterprising,” “erudite,” and “confident” and so on were positively related to competence; the words “mediocre intelligence,” “inferior,” and “conservative” and so on were negatively related to competence. Thus, the present study provided evidence that warmth and competence exist in stereotypes held by Chinese people.

Stereotypes were not limited to a set of beliefs about the personal attributes of a group, but also included feelings, physical appearances, preferences, and occupations (Ashmore and Boca, 1979; Wong, 2004). In addition to occupational personality according to classical SCM, the present study also revealed two factors named occupational cognition and emotion in teachers' stereotype content. Occupational cognition referred to one's cognition about a specific occupation, such as welfare, remuneration, social prestige, development prospects, etc. As for occupational cognition, the evaluations of the two teacher types all involved salary, welfare, prospects, fame, and working atmosphere. However, the evaluations of university teachers considered more than the foregoing, such as freedom (corresponding words are “free,” “restricted,” and “open”), achievement (corresponding phrases are high and low achievement), and working artistry (corresponding words are “elegant” and “vulgar”). The results are consistent with a previous study that found occupational prospects, working strength, working conditions, and incomes have important effects on occupational cognition (Chen, 2011). In this paper, occupational emotion referred to cognitions about teacher's emotional states when they were working. For example, undergraduates think that middle school teachers are always anxious; this is undergraduates' cognition about teachers' emotions. The occupational emotions of two teacher types included labor intensity (corresponding words or phrases are “effortless,” “arduous,” “leisurely,” “busy,” “relaxed,” “laborious,” “stressful teaching,” and “stressful researching”) and teachers' personal emotions (corresponding words or phrases are “worry-free,” “worried,” “less anxious,” and “apprehensive”). Adelmann (1989) suggested that higher emotion workers belonged to the teaching profession and middle school teachers suffered from students' bad behavior, time pressure, excessive workload, and so on. The present study suggests that occupational cognition and emotion are parts of teacher stereotypes held by undergraduates.

For both the positive and negative descriptions of occupational cognition, there were no significant differences between teacher candidates and general undergraduates when evaluating middle school and university teachers. Generally, people's beliefs about others partly depend on their experience. Stereotypes also result from the complicated interaction between their holder and targets (Hakel et al., 1970; Imada and London, 1979). To some degree, the two student types have contact with the same type of middle school and university teachers. Thus, they may have received similar information about teachers' salaries, welfare, prospects, and so on. In addition, there were mutual contacts and sharing of viewpoints between teacher candidates and general undergraduates; thus, it is natural that they have similar occupational cognitions toward the two teacher types.

In terms of the occupational personality of middle school teachers, teacher candidates gave significantly lower scores than undergraduates in the context of negative descriptions (Figure 2B). Similarly, for occupational emotion about middle school teachers, compared with general undergraduates, teacher candidates had significantly higher scores in the context of positive descriptions (Figure 1C) and significantly lower scores in the context of negative descriptions (Figure 2C). The results are consistent with social identity theory (Tajfel, 1981, 1982) that social identity consists of social categorization, social comparison, and positive distinctiveness (Tajfel, 1982). In emerging intergroup comparisons, individuals would increase self-esteem by preserving or achieving positive social identity. Simultaneously, individuals tended to give in-group members more positive appraisal. In addition, even with no explicit conflict and competition in groups, there is also a tendency toward in group-favoring behavior (Tajfel, 1974, 1981; Tajfel and Turner, 1979). Therefore, compared with general undergraduates, teacher candidates held a superior stereotype toward middle school teachers in terms of occupational personality and emotion.

Furthermore, in terms of occupational emotion about university teachers, teacher candidates had significantly lower scores in the context of positive descriptions, but significantly higher scores in the context of negative descriptions when compared with general undergraduates (Figures 1C, 2C). According to in-group favoritism, there should not have been significant differences between the scores given by teacher candidates and general undergraduates, because university teachers were an out-group for both kinds of undergraduates. The present study suggested a discrepant result that can be interpreted well by social comparison and self-evaluation maintenance (SEM) models. The SEM model is composed of reflection and comparison processes. According to social comparison theory, if another's performance is highly relevant to oneself, the comparison process will be relatively important and one will suffer by comparison with a close other's better performance. The outstanding performance of a close other could make one's own performance pale by comparison and even lead to decreased self-evaluation (Tesser et al., 1988). In fact, the SEM model suggested that persons behave in a manner that will maintain or increase self-evaluation in various ways including disparaging others (Brown, 1986; Epstein and Feist, 1988). It is common to compare middle school teachers with university teachers (Liu et al., 2007; Samek et al., 2010). As shown in Figures 1A, 2A, both teacher candidates and general undergraduates agree that university teachers have advantages in possessing more resources like salary, welfare, prospects, fame, and so on. There is no doubt that teacher candidates will become middle school teachers in the future. According to the SEM comparison process, when evaluating university teachers, teacher candidates' self-evaluations may be affected. To maintain or increase positive self-esteem, they will decrease university teachers' evaluations. Therefore, teacher candidates had more negative occupational emotions for university teachers compared with general undergraduates.

Finally, we compared the stereotypes about middle school teachers with university teachers rather with than primary teachers, which was due to the following considerations. First, undergraduates are more familiar with these two types of teachers. Specifically, now they are pursuing higher education inspired by university teachers. And they graduated from middle school not more than 4 years, but graduated from primary school at least 6 years. Therefore, the stereotypes about middle school and university teachers are clearer than the stereotypes about primary teachers (Wang, 2004; Liu et al., 2007). Second, by comparing stereotypes toward university and middle school teachers, we can understand the relative status of middle school teachers as perceived by undergraduates, because university teachers are widely considered to possess higher status and incomes in China (Wang et al., 2014). Moreover, the present study did not measure stereotypes about preschool or primary teachers because tuition-free teacher candidates will work as middle school teachers rather than as primary school or nursery teachers. It is valuable to identify the stereotypes about preschool and primary teachers held by undergraduates majoring in nursery and elementary education in the future.

In sum, the present study develops a three-factor (SCM) about middle school and university teachers, and found differences between the stereotypes held by tuition-free teacher candidates and general undergraduates. Simultaneously, we proposed that this three-factor stereotype model can be used as a potential tool to assess teacher candidates' actual willingness to work as middle school teachers, because of a significant correlation between stereotypes held by students and their willingness to work as middle school teachers. To some degree, it is good for tuition-free teacher candidates to form appropriate stereotypes about middle school teachers and form a clear career plan. Despite its contributions, the present study has certain limitations. An insufficient sample size may limit the generalizability of our findings. Future research could increase the sample size. In addition, the present model may not be applicable in other countries with different cultures. Thus, it needs further investigation in settings outside China.

YGC, Substantial contributions to the design of the work, and revising the paper critically for important intellectual content. YXZ, Completing the whole study and be responsible for writing the paper. YFZ, CHP, Giving a great deal of help during the acquisition, analysis and interpretation of data. What's more, and also give the pertinent suggestions for the revise of paper. YGC, YXZ, YFZ, and CHP, Final approval of the version to be published, and got agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

This study was supported by a grant from the Key Research Institute of Humanities and Social Science in Chongqing (16SKB008).