Table 1 The statistical characteristic of subjects

2.2 Materials

(1) With reference to the Teachers’ Teaching Regulated Ability Scale constructed by Beijing Normal University, We constructed the Middle School Teachers’ Teaching Strategy Scale (The MSTTS). The MSTTS, a 66-item, 5-point, Likert-type instrument, was divided into three dimensions, nine components. The three dimensions are pre-class strategy (planning strategy and preparing strategy) , in-class strategy (managing strategy, motivation strategy, teaching method and guiding strategy ) and after-class strategy (evaluation strategy, reflectivity strategy and compensation strategy). Through the examination, the consistency reliability was 0.920 and split-half reliability was 0.848. The Middle School Teachers’ Teaching reflection Scale made by Shen J L. (coefficient alpha was 0.864, split-half reliability was 0.756 ) was acted as the criterion scale. The pearson correlation coefficient between the two scales was 0.810 ( N=126,p<0.01) and the correlation between the dimensionalities in the two scales have reached the significant level. The correlation coefficient matrix showed the correlation coefficient of the nine factors ranged from 0.264 to 0.600 (p<0.01). Hence it showed the scale had strong construct validity. The indexes of the confirmatory factor analysis ranged from 0.87 to 0.98,showing better goodness-of-fit among the three sub-scales.

(2) With reference to Walle’s (1997) Three Factors Scale, the Teachers’ Achievement Goal Orientation Questionnaire (The TAGOQ ) was developed. The TAGOQ ,a 18-item, 7-point, was made to measure the task goal and performance goal of the teachers. (1 means absolutely agree, 7 means absolutely disagree) The consistency reliability of the TAGOO, the subscale of the task goal and the subscale of the performance goal were respectively 0.792, 0.870 and 0.781. The test-retest reliability was 0.823. The outcome of the content validity showed that the correlation between the items of task goal subscale and the total score of it ranged from 0.64 to 0.80 (p﹤0.01). The correlation coefficient between the items of performance subscale and the total score of the TAGOQ ranged from 0.58 to 0.83 (p﹤0.01). The confirmatory factor analysis showed better goodness-of-fit, with indexes ranging from 0.86 to 0.96.

(3) Eysenck Personality Questionnaire Revised. (Modified by Chen Z G. ,1981) . The Eysenck Personality Questionnaire, Revised (EPQ-R) is a self-report instrument that uses 85 items to determine the strengths of three dimensions of personality: psychoticism, neuroticism and extroversion. The fourth is the validity subscale. High score in the extraversion-introversion dimension indicates extraversion personality. Such as sociable, easygoing, optimistic, fond of change, adventurous, impetuous. Low score shows introversion personality. Like quiet, introspective, reticent, unconcerned about others except close friends, fond of orderly life style and emotionally stable etc. In the neuroticism dimension, high score indicates anxiety, depression, heavy-heartedness, response so strongly to stimuli that often results in less rational behavior. In the psychoticism dimension, high score shows loneliness, unconcerned about others, lack of emotion, unreasonable, fond of provocation, like doing something peculiar regardless of danger, difficult in adapting to the social environment. The assessment is 2-point, 1 is yes,2 is no. The consistency reliability is 0.554.

(4) With reference to the scale designed by Meger, Allen and Long L R. (2001), the Middle School Teachers’ Affective Commitment Scale (The MCTACS) was constructed. The MCTACS uses 22 items to form three dimensions: affective commitment, means the degree of love towards teaching, how deep they throw themselves into teaching career, and their cognition of coincidental degree among teaching career, personal ideal and value; continuance commitment, means a person is unwilling to quit the teaching career due to the fear of losing benefits and difficulties of finding another job; normative commitment, it means on the one hand a person doesn’t want to quit the teaching career because he accepts social ethics ,on the other hand, he has a sense of obligation and responsibility towards teaching career. A 6-point assessment system is used in the scale. Through statistics analysis, the consistency coefficient is 0.853, split-half reliability is 0.769. The analysis of content validity shows the correlation between each item and the scale ranges from 0.48 to 0.65 (p﹤0.01) ,the correlation between the three subscales and the scale are respectively 0.766,0.555,0.778 (p﹤0.01). The indexes of confirmatory factor analysis ranged from 0.65 to 0.99,showing better goodness-of-fit.

(5) With reference to the scale designed by Maslach(1997), the Middle School Teachers’ Job Burnout Scale was developed. The MSTJS consists of 21 items formed three dimensions: emotional exhaustion, means one feels exhausted and forfeit one’s emotional resource because one cannot deal with the problems and demands around him during his teaching; depersonalization, when the teacher get together with the students, he response in an indifferent attitude, that is the teacher gradually comes up with a negative, unconcerned attitude towards colleagues and students and a sense of detestation to the world and its ways; reduced personal accomplishment, the teacher feels a sense of low achievement which comes from three aspects: a sense of lack of ability in cooperation with colleagues, a sense of lack of work ability and a sense of lack of working significance. Through the statistics analysis, the consistence coefficient of the scale was 0.883, the split-half reliability was 0.801. Factor analysis indicates good construct validity and content validity. The analysis of content validity shows the correlation between each item and the whole scale ranged from 0.301 to 0.682 (p﹤0.01), the correlation between the three subscales and the whole scale are respectively 0.766,0.555,0.778 (p﹤0.01). The indexes of the confirmatory factor analysis ranged from 0.60 to 0.99, showing a better goodness-of-fit.

2.3 Procedure

Group measure was adopted. And instruction was used to instruct the subject to fill in the questionnaire.

3.Results

Table 2 The mean on teaching strategies of different types of teachers