蔡永红    

论文题目：教师职务绩效--结构及其影响因素研究

作者简介：蔡永红，女,1967年04月出生，1999年09月师从于北京师范大学林崇德教授，于2002年07月获博士学位。

摘  要

教师职务绩效的结构是教师绩效评价的理论基础。它是目前教师研究中的一个重要的理论问题，属于教师研究中的基础理论研究。长期以来，教师心理研究主要关注了对教师素质的探讨，而教师管理研究则主要关注了对教师评价的探讨，这两类研究长期分离在两个领域讨论。本研究从一种全新的视角来探讨教师发展与教师管理中的一个重要问题——教师职务绩效。这一新的问题研究角度将人事心理学、教育心理学及发展心理学三个领域的研究结合在一起，并一改过去分离地讨论教师素质与教师评价问题的传统，把两类研究结合在对教师评价类型的划分上，将其分为三种类型，即教师胜任力评价，教师绩效评价及教师有效性评价，认为这三种类型的评价应该在教师教育及管理的不同阶段使用，它们具有不同的功能，并提出教师素质就是教师胜任力，教师素质结构是教师胜任力评价的基础，而教师绩效结构则是教师绩效评价的基础。教师素质是教师选拔的基础，而教师绩效则是教师管理与教师教育培训的基础。在此基础上，通过两年多的实证研究，该研究系统地探讨了以下问题：

（1）教师职务绩效的结构；

（2）不同评价方式下教师绩效结构的一致性与差异性；

（3）不同评价者绩效评价结果的影响因素；

（4）不同评价方式所获得的绩效评价结果的比较分析

   本研究采用了理论与实践相结合的研究思路，深入实践，从实践中提出问题，在实践中研究并解决问题。该研究的问题提出来源于实践——即教师聘任制改革的需要，来源于中小学教师管理实践的需要（包括北大附中，北京市育才学校，北京市太平桥中学、浙江富阳中学，浙江富阳永兴中学等）。在深入实践两年多的时间里，系统地对大量教师进行了关键事件访谈，关键事件调查，并对人事心理学，教育心理学及发展心理学三个领域的相关文献进行了系统分析，在此基础上，结合作者所在博士点多年来所提出的教师素质结构理论，提出了教师绩效的结构，并在五所学校进行实证研究，对所提出的教师绩效结构进行了验证性分析。

  在对教师职务绩效结构的探讨中，在对大量文献进行回顾的基础上，通过开放式调查、个别访谈、关键事件访谈、关键事件调查等多种方式，对教师的工作进行了系统的分析，并基于教师素质结构的理论，我们提出了对教师职务绩效的内涵及结构的看法。我们认为：教师职务绩效是教师在教育教学过程中所表现出来的与教育教学目标相一致的行为。在教师的职务绩效中，包含职业道德、职务奉献、助人合作、教学效能、师生互动与教学价值六个维度，而这六个维度又可以进一步概括为两个二阶因素，即，关系绩效与任务绩效。前三个维度属于关系绩效，后三个维度属于任务绩效。认为，教师的工作具有高度的自主性，因而，关系绩效对其总体绩效的贡献很大。此外，关系绩效与任务绩效有相关，但二者各自所受到的影响因素有所不同。认知因素主要影响任务绩效，而人格因素则主要影响关系绩效。

  以这一结构为基础，我们编制了教师绩效的评价量表，并在北京和浙江杭州两地，共选取了五所学校，让不同的评价者用这个量表对同一个教师进行评价。其中，学生评价采用等距抽样方式，分别选取了5-20名学生对一个教师进行评价，同事评价采用了抽签方式，分别随机选取了2-10个同事来评价一个教师，领导评价采用了直接上级评价，每一个教师分别由2-7个领导进行评价。有完整学生评价数据的教师有312人，有完整领导评价、同事评价及自评的教师分别是168,168和181人。我们对学生评价的实证数据进行了验证性因素分析。接着，又对同事评价、领导评价及自评数据也进行了验证性因素分析，以探讨了不同评价方式下教师绩效在结构上的一致性与差异性。在此基础上，我们进一步对学生评价、领导评价及同事评价结果的影响因素进行了分析。最后，我们进一步分析了评价方式对评价结果的影响。

研究结果表明：

（1）教师的职务绩效包含6个维度，即职业道德、职务奉献、助人合作、教学效能、教学价值与师生互动。教师职务绩效的二阶因子包含任务绩效和关系绩效两部分。在教师的职务绩效中，任务绩效与关系绩效并不完全独立。二者有较高的相关。在教师职务绩效中，关系绩效对总体绩效的影响明显。

（2）不同评价方式下，绩效的结构维度不太一样，自评与其它评价方式的结构明显不同。在自评中，助人合作与师生互动合为一个维度，在二阶因子上属于任务绩效。学生评价、同事评价及领导评价具有相同的结构维度，但学生评价的结构效度最高，其次是同事评价，领导评价的结构效度不太好。以四种评价结果的相关分析为基础的效度分析表明，在四种评价方式中，学生评价的效度最高，自评的效度最差，同事评价与领导评价的效度一般。

（3）不同评价者用于做出绩效判断的信息来源不同，评价方式对评价结果的影响非常显著。在不同评价方式下，关系绩效与总体绩效的关系有所不同。在领导评价中，关系绩效对总体绩效的影响大于任务绩效，而在学生和同事评价中，任务绩效对总体绩效的影响大于关系绩效。领导评价更加重视关系绩效，而学生和同事评价更加重视任务绩效。

（4）代表知识、经验的职称、学历、教龄等因素，对任务绩效的影响更加明显，而人格因素对关系绩效的影响更加明显。人格因素对不同评价者的影响有所不同，责任心更多地影响领导所做出的关系绩效的判断，经验的开放性更多地影响学生所做出的任务绩效判断。而随和性对同事评价的任务绩效具有明显的负向影响，情绪稳定性明显地影响同事所做出的关系绩效的判断。

（5）职称、学历及教龄等因素，对绩效的影响不是线性的，绩效并不是随职称、学历及教龄的增加而提高，而是存在某个关键的职称、学历及教龄段，在这一阶段，被评者的任务绩效的表现最佳。

（6）领导评价受到刻板印象的影响较大。领导在评价绩效时，明显受到了对性别、学历等因素的刻板印象的影响。

根据本研究的结果，我们着重讨论了以下五个方面的问题：

（1）教师职务绩效包含关系绩效和任务绩效两个维度，不同评价方式下，任务绩效与关系绩效的所包含的具体内容维度有一定的差异。

（2）评估方式对评估结果有明显的影响，不同评价者做出绩效判断时，所依据的信息来源有所不同，因而，其做出的绩效判断也有一定差异。

（3）影响任务绩效与关系绩效评价结果的因素也有所不同，知识、经验主要影响任务绩效，人格因素主要影响关系绩效，情感因素对绩效评估的影响明显。

（4）对目前绩效评估研究的反思。我们认为，目前绩效评估研究领域存在一些问题，影响了研究的深入，这些问题是：以可直接观察的行为定义绩效，对于有些工作来说并不适合；当关系绩效影响个体的报酬及升迁时，它将不再是一种关系绩效行为；对组织系统及技术环境的作用仍然重视不够。

（5）本研究今后努力的方向，本研究还需要扩大取样范围，进行重复性研究，进一步探讨不同评价方式的绩效结构问题，并进一步建构教师个体绩效的因果模型。

Abstract

 The construction of Teacher job performance is the theory basis of teacher job performance appraisal. At present, it is a very important theory problem in the teacher research area. It is also a basic theory study of teacher research.  For ages, teacher psychological research has primarily paid attention to probing into teacher quality; however, teacher management research has primarily paid attention to inquiring into teacher appraisal. These two kinds of study have always been discussed in two distinct areas for a long time. This study inquired into an important issue of teacher development and teacher management from a fire-new perspective------ teacher job performance. This point of view combines three areas research -----personnel psychology, education psychology and development psychology----- with each other. It shifted from the tradition of separately discussing teacher quality and teacher appraisal to integrate these two kinds of research into dividing teacher appraisal type. It divided teacher appraisal type into three sorts: teacher competence appraisal, teacher performance appraisal and teacher effectiveness appraisal. We thought that each kind of teacher appraisal has different function and should be use in different stage of teacher education and teacher management. We also suggested that teacher quality is teacher competence, the construction of teacher quality is the basis of teacher competence appraisal, but the construction of teacher performance is the basis of teacher performance appraisal, teacher quality is the basis of teacher selection, but teacher performance is the basis of teacher management and teacher education or training. Through more than two years’ experimental study, this dissertation systematically explored such questions:

(1) the construction of teacher job performance;

(2) the construction consistency and diversity of different appraisers;

(3) influential factors to different appraisers’  rating results ;

(4) the compare of different raters’ appraisal results。

This study used the approach of combining theory with practice. It embed in practice, advanced problem from practice, researched and settled problem in practice. The problem was presented by practice-----the need of teacher appointment system reform. It advanced by the need of elementary and middle school’s reform. In more than two years of embedding in practice, we used key incidence interview, key incidence survey systematically investigated large numbers of teachers. We also systematically analysis literature of personnel psychology,  education psychology and development psychology. Then, on the basis of teacher quality construction theory, we suggested the construction of teacher performance.  We carried through empirical research in five schools. Then we did confirmative analysis to teacher performance construction.

In order to explore the teacher job performance construction, we based our research on the broad review of related research literatures. We adopted open-ended survey, individual interview，critical incident interview and critical incident survey to analysis the teacher job. On the basis of teacher quality construction theory, we put forward our view of the mining and construction of teacher job performance. We suggested that: teacher job performance is teacher’s behavior that is in accordance with teaching and education aim in the process of teaching and education. Teacher job performance involved six dimensions, such as :occupational morality, job dedication, assistance and cooperation, teaching effectiveness, teacher-student interaction and teaching value. The six dimensions could be generalized into two high-layer factors: contextual performance and task performance. The front three dimensions belong to contextual performance, and the last three belong to task performance. We suggested that teacher’s job has high autonomy. So contextual performance would have high contributions to total performance. Moreover, contextual performance would have correlation with task performance. But their influential factors would be different. Cognition factors would mostly influence task performance. But personality factors would mostly influence contextual performance.

According to above construction, we formed our teacher performance rating scale. In Beijing and Hangzhou , we let different evaluators rate the same teacher using the same performance rating scale. In student rating , we adopted fixed interval sample method. We respectively let five to twenty students to evaluate one teacher. In the colleague rating, we respectively selected two to ten peers by lots to evaluate one teacher. In the administrator evaluation, we adopted directive superior rating. Two to seven superiors rated each teacher. 312 teachers were rated by students. Therein, 168 teachers were rated by their superiors and peers. 181 teachers had self-rating data. We used Confirmative Factor Analysis (CFA) to analyze student’s rating data. In the following, in order to explore the construction consistency and diversity of different appraisers, we also used CFA to analyze peer rating data, superior rating data and self- rating data. Then we analyzed the influential factors to student rating, superior rating and peer rating. In the end, we analyzed the influence of rating method to rating results.

Results showed:

(1)Teacher job performance involves six dimensions. The six dimension is: occupational morality, job dedication, assistance and cooperation, teaching effectiveness, teacher-student interaction and teaching value. The six dimensions generalized into two high-layer factors: contextual performance and task performance. Contextual performance and task performance do not independent each other. They have high correlation. In teacher performance, contextual performance markedly influence total performance.

(2) Different rating methods have different constructions. Self rating has  evidently different construction with other methods. In self-rating, assistance and cooperation dimension incorporates with teacher-student interaction. It belongs to   task performance. Student rating, peer rating and superior rating have the same construction. But student rating has good construction validity. Peer rating’s construction validity takes the second place. Superior rating’s construction validity is bad. The validity analysis based on the correlation of four kinds rating results shows that: in all these four, student rating has the highest validity; self rating has the worst validity; peer rating and superior rating have common validity.

(3) Information used by different raters to draw performance judgment is different. Rating method has very marked influence to rating results. In different rating methods, contextual performance has different influence to total performance. In superior rating, contextual performance contributes more to total performance than task performance. In student and peer rating, task performance contributes more to total performance than contextual performance. Superior rating thinks much more of contextual performance. But student and peer rating think much more of task performance.

(4) Representatives of knowledge and experience, such as technical post, educational level, teaching age and so on, influence task performance more evidently. But personality influence contextual performance more markedly. Personality has different influence to different rater. Conscientiousness has more influence on superior’s rating of contextual performance. Openness to experience influences student’s rating on task performance more. But agreeableness has more influence on peer’s of task rating in the opposite direction. Neuroticism has evident influence on peer’s rating of contextual performance.

(5) Factors, such as technical post, educational level, teaching age and so on, have non-linear influence to performance. There is a key technical post, educational level, teaching age. In this stage, ratee has best performance.

(6) Superior rating is greatly influenced by stereotype. Superior is markedly influenced by sex and educational level stereotype.

Based on the results, we mainly discussed five problems: (1) Teacher performance involves contextual performance and task performance. In different rating methods, the contents of contextual performance and task performance have difference.  (2) Rating method has evident influence on the rating results. Different rater relies on different information to draw their performance judge. So their performance judges have discrepancy. (3) Factors that influence contextual performance and task performance are different. Knowledge and experience mainly influence task performance. Personality mainly influence contextual performance. Affective factor has evident influence on performance rating. (4) Reflection on present performance appraisal research. We suggest that: some issues in present performance appraisal research restrict the depth of study. These issues are: basing performance on direct observed behavior do not suit some job;  when contextual performance affects one’s pay and promotion, it will not be contextual behavior again; attention  paid  to organization system and technical condition is not enough. (5) This study’s further direction. This research needs to extend its sample numbers, carry repeated studies, and explore different rating method’s construction further. It needs to construct teacher personal performance casual model.