The Empirical Research On Scientific Evidence-based Reasoning Ability

Evidence-based reasoning ability,one of the high order thinking skills,always plays an important role in science education reform,which is the key and essential ability in solving scientific problems and conducting scientific inquiry,etc.With the reform and development of international science education and cultivation demand of students’ key competencies,this dissertation conducted systematic empirical research focusing on “Scientific Evidence-based Reasoning Ability”(SERA)Corresponding work and results are as follows:By combining with theoretical analysis of different fields,we clarified the connotation of SERA,that is the ability to recognize,transform and formulate evidence from the existing experience and question context,then conduct the process of reasoning leading to making conclusions and solving problems supported by such evidence.Based on the essential connotation and relevant literature review,we begun with the analysis on the curriculum documents that mainly involved standards from different countries and regions,and the discussion of the embodied “evidence-based reasoning” in the classroom teaching.They provided the theoretical and practice basis respectively.By merging the complexity of evidence and reasoning into an integrated structure,and combining with experts review,we constructed and modified a framework to categorize and define SERA(FSERA).The next work was the development of instrument on SERA according to the constructed framework.Followed by the initial instrument,pilot study was conducted to examine its quality.The items were evaluated and rectified based on the data results.Finally,an operable and qualified instrument for SERA assessment was obtained.A SERA investigation was conducted with 2261 junior middle school students by employing the modified instrument.The data analysis was operated through SPSS 22.0 and WINSTEPS 3.72.0 software.The results indicated that those participants held significant different performance in different grades,gender and academic levels.According to the ability levels in FSERA and interview case analysis,students’ performance in these variables(grade,gender and academic level)were compared in detail.The influence factor model of SERA was developed and evaluated following by the analysis approach of Structural Equation Model(SEM).In the Phase of Model Development,beginning with the theoretical development,five possible factors were proposed and investigated.They were reading academic achievement,spatial ability,mathematics academic achievement,science academic achievement,and interest in science learning.The identifiable measurement model,structural model and integrated model were specified and reported respectively.On the basis of model development,questionnaires were compiled and chose,the proposed factors were investigated in the Phase of Model Assessment.By applying the statistical techniques of SEM and AMOS 22.0 software,the collected data was analyzed,including parameter estimation and fit test,for determining the relationship path between these influence factors and students’ SERA performance,and the influence factor model of SERA.In the follow-up work of assessment phase,we modified the initial model to achieve a better one that ensuring its theoretical and statistical quality.Subsequently,the model was applied in covariant structural analysis of multi-group,involving crossvalidation and moderation analysis.Thus,the influence factor model was discussed and examined further,as well as the in-depth comparison on model structure among different samples.SERA is of significance in science learning and students’ competence development.This dissertation centred on this by conducting assessment study and influence factors study.Grounded on theoretical research,we obtained ample empirical data results and achieved the anticipated research goals.Accordingly,this research was beneficial for science education research and ability assessment,which supported reasoning skills hoisting,high order thinking development,disciplinary competence cultivation,and the acceleration of science learning.It can also provide evidence and direction for improvement of teachers’ instruction practice.While in the aspect of instrument rectification,sample selection,influence factors model modification etc.,we shall carry on more research in the future.