| With the development of society and the demand for high-quality citizen,assessment has shifted from examine what students have learned in the past to the ability to solve problems in the future.Focus on the development of students’ life-long learning ability.For example,PISA,an international large-scale educational assessment program,aims to determine whether students master the knowledge and skills needed to participate in society are emphasis on the use of contextualized items.The contextualized items can better examine the students’ scientific literacy,and diagnose students’ problems in extracting and applying knowledge,improve the teaching’。and train the talents that meet the future social development for the country.However,the proportion of contextualized items in the whole set of questions is more reasonable.How to build a situational classification framework for localization contextualized items? These problems still need to be broken.In view of the above problems,the general idea of this study is: First,based on literature research,the paper defines the contextualized items and constructs a localized situation classification framework.Then apply for data to carry out research,because the author’s Institute of Science Education has been undertaking research and development for junior high school science assessment projects for 13 years,accumulating a large number of test questions and test data.The author used Excel to count the 789 test questions of the junior high school science assessment project in the past 13 years,and obtained the ratio of the proportion of contextualized items to the five categories of contextualized items.Next,author used the SPSS to analyze the test data of four regions in 2016 to find out the factors that influence the score of contextualized items.In addition,SPSS was also used to analyze the test data of 2015 and 2016 in C region to find out the scoring factors that affect the five categories of contextualized items.Finally,according to the research results,the author puts forward some suggestions on the improvement of scientific evaluation and proposition.The results were as follows:1.The ratio of contextualized items for the middle and early 13 th year science assessment project was between 50% and 80%,which was higher than TIMSS(42.4%)and NAEP(41.4%)but lower than PISA(92.2%).2.The overall student’s score on the contextualized items is lower than the noncontextualized items,which indicates that the situation makes the test more difficult.3.The scoring factors affecting the contextualized items are gender differences,urban-rural structural differences,schools of different nature,whether they are only children,and whether they are not left behind.4.Nearly 13 years ago,the proportion of the five major contextualized items in the middle school science assessment project: the first 6 years are the most accounted for the number of life-contextualized items,the middle 5 years are the scientific practice-contextualized items.The proportion of technical and engineering-contextualized items after 2014 has increased significantly.And in 2016,the proportion reached the maximum.At the same time,the respective proportions of the five contextualized items vary from year to year.5.The scores of the five contextualized items: the highest scores for all students are the technical and engineering-contextualized items,followed by the life-contextualized items,next is the natural and scientific practice-contextualized items.The lowest score is the environmental-contextualized items.6.It was found that there was no significant difference between the five contextualized items.This paper puts forward the following suggestions based on the above research.(1)Control the complexity of the context.(2)The creation of contextualized items must be true and fair.(3)The contextualized items must be specific.(4)Encourage teachers to implement contextual teaching. |
