Application of Correspondence Analysis and Related Methods to - PowerPoint PPT Presentation

Application of Correspondence Analysis and Related Methods to Evaluation of Knowledge and Skills of Students prof. Jozef Dziechciarz dr Alicja Grzeskowiak dr Agnieszka Stanimir

Education Policy Background Main goal of education reform (Poland, 1999): “the improvement of quality in formal educational system and its adjustment to requirement of the knowledge-base economy, assuring the access to lifelong learning system for the entire society” Source: Perspektywy gospodarki opartej o wiedzę w Polsce ; World Bank Report, KBN, 2004 Principles of the reform: new structure of the education system, ranging from • nursery school to doctoral studies an independent assessment, external examination • system supervision separated from management • new system of financing • new professional status of the teaching staff linked • with the promotion paths CARME Conference, Rennes, 8-11 February 2011

Education Reform and PISA (Programme for International Students Assessment) Surveys 1999 – PISA trial version of survey, introduction of • Education Reform in Poland 2000 – PISA first survey, participating Polish • students approaching the end of compulsory education in the old system (very good baseline for comparisons) 2009 – fourth PISA survey; after ten years after the • introduction of the reform Poland’s participation was summarized as: „on average, the reform was associated with significant improvements. Poland improved its score in mathematics by 0.25 of a standard deviation, in reading, by 0.28 of a standard deviation, and in science, by 0.16 of a standard deviation” Source: The Impact of the 1999 Education Reform in Poland , OECD Education Working Paper no.49, OECD, 2010 CARME Conference, Rennes, 8-11 February 2011

Scope of Research PISA results evaluation was published in • 2009 but the newest information taken into consideration was from 2006 the need for evaluation based on the up- • to-date results national exam results from 2010 • analyzed, similar competences as PISA test regarded special attention paid to differences • according to gender, place of residence and to changes in time CARME Conference, Rennes, 8-11 February 2011

Gymnasium as a Part of Education System in Poland • Age of students: 14-16 • General education Analyzed • Compulsory school education level • No specialization paths Source: The Education and Assessment System in Poland ,Central Examination Board, 1999 CARME Conference, Rennes, 8-11 February 2011

Regional Aspect of the Study 8 Regional Examination Boards Source: Central Examination Board • south-western part of the country (two administrative territories) • area: 29 thousand km 2 (9,1% of total) • population: 3,9 million (10,2% of total) • number of students taking the exam: 40,9 thousand (9,2% of total) CARME Conference, Rennes, 8-11 February 2011

External Exams Idea Nationwide Standardized comparability of Procedures results External Exams Examination papers Not relevant for coded to provide graduation, relevant objective assessment for further enrolment CARME Conference, Rennes, 8-11 February 2011

Gymnasium Exam - Division • Polish language • History max 50 Humanities • Civic education points • Arts • Mathematics • Physics max 50 • Astronomy Science points • Chemistry • Biology • Geography CARME Conference, Rennes, 8-11 February 2011

Evaluated Competences Areas • Reading and interpretation of Humanities texts • Creation of own text • Finding and using information • Application of terms and procedures • Identifying and describing Science facts, relationships and dependences • Application of integrated knowledge and skills to solve problems CARME Conference, Rennes, 8-11 February 2011

General Remarks on Results (2010) 6 Median 31 5 Median 22 4 % 3 2 1 0 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 number of points science humanities CARME Conference, Rennes, 8-11 February 2011

Population Structure (2010) Gender 49,97% 50,03% Place of Residence by Commune Type Female Male 19% 21% • urban commune: limited to the administrative town/city borders • rural commune: 25% the territory does 35% not comprise a Urban (cities more than 80 th inhabitants) town • urban-rural: Urban (towns less than 80 th inhabitants) includes a town Urban-rural and some villages Rural CARME Conference, Rennes, 8-11 February 2011

Multivariate CA Based on Burt’s Matrix results obtained in points were coded as • excellent, good, moderate and weak two analyses for two administrative • territories (Lower Silesia Region and Opolskie Region) were conducted quality of display in two-dimensional space • was equal to 26% and 30%, respectively (to regions) CARME Conference, Rennes, 8-11 February 2011

Multivariate CA Based on Burt’s Matrix - Results Lower Silesia Region Opolskie Region 2,0 2,5 2,0 1,5 SCIENCE:EXCELLENT SCIENCE:EXCELLENT HUM.:WEAK HUM.:WEAK 1,5 SCIENCE:MODERATE SCIENCE:MODERATE 1,0 HUM.:GOOD HUM.:GOOD HUM.:EXCELLENT HUM.:EXCELLENT 1,0 OPOLE OPOLE Excellent results in 0,5  2 =0,24 (11,29%)  2 =0,33 (13,31%) SCIENCE:WEAK SCIENCE:WEAK RURAL RURAL URBAN URBAN SCIENCE:GOOD SCIENCE:GOOD 0,5 Weak results associated neither with gender nor with big cities GENDER:G GENDER:G URBAN/RURAL URBAN/RURAL URBAN URBAN 0,0 SCIENCE:GOOD SCIENCE:GOOD GENDER:B GENDER:B HUM.:MODERATE HUM.:MODERATE HUM.:MODERATE HUM.:MODERATE GENDER:B GENDER:B commune type C.LEG. C.LEG. 0,0 URBAN/RURAL URBAN/RURAL GENDER:G GENDER:G C.JG. C.JG. WROCLAW WROCLAW RURAL RURAL -0,5 SCIENCE:WEAK SCIENCE:WEAK -0,5 HUM.:GOOD HUM.:GOOD Girls better results than HUM.:EXCELLENT HUM.:EXCELLENT SCIENCE:MODERATE SCIENCE:MODERATE -1,0 boys in both science -1,0 SCIENCE:EXCELLENT SCIENCE:EXCELLENT and humanities HUM.:WEAK HUM.:WEAK -1,5 -1,5 -2,0 -2,0 -2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 -2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 2,5  1 =0,44 (14,78%)  1 =0,42 (16,88%) As the analysis based on Burt’s matrix did not indicated the exact diversification among the gymnasium students the data were analyzed with application of multiway contingency tables. CARME Conference, Rennes, 8-11 February 2011

Analysis Based on Multiway Contingency Tables • higher-order associations were applied between commune type and gender and between knowledge area and exam score • data table used in this analysis is 20x8, dimensionality of this table is 7 • percentage of explained inertia in the map is 96% CARME Conference, Rennes, 8-11 February 2011

Analysis Based on Multiway Contingency Tables – General Results 0,6 0,5 No particular categories Girls from big cities 0,4 associated with excellent in 0,3 weak results in humanities science SCIENCE-WEAK SCIENCE-WEAK 0,2  2 =0,01 (20,40%) Boys from rural HUM.-EXCELLENT HUM.-EXCELLENT DS-LEG-G DS-LEG-G SCIENCE-MODERATE SCIENCE-MODERATE 0,1 and urban-rural DS-JG-G DS-JG-G DS-R-B DS-R-B HUM.-GOOD HUM.-GOOD communes DS-U/R-B DS-U/R-B OP-U-B OP-U-B OP-OP-G OP-OP-G 0,0 WROCLAW-G WROCLAW-G moderate in SCIENCE-GOOD SCIENCE-GOOD DS-U-B DS-U-B HUM.-MODERATE HUM.-MODERATE OP-U-B OP-U-B DS-LEG-B DS-LEG-B humanities DS-JG-B DS-JG-B -0,1 OP-OP-B OP-OP-B WROCLAW-B WROCLAW-B HUM.-WEAK HUM.-WEAK -0,2 OP-U/R-B OP-U/R-B Boys from other SCIENCE-EXCELLENT SCIENCE-EXCELLENT No particular -0,3 big cities good categories in science Boys from the -0,4 associated with capital of the weak results in -0,5 region excellent -0,5 humanities -0,4 -0,3 -0,2 -0,1 0,0 0,1 0,2 0,3 0,4 0,5 0,6 in science  1 =0,04 (75,16%) CARME Conference, Rennes, 8-11 February 2011

Analysis Based on Multiway Contingency Tables: Lower Silesia Region – General Overview 0,6 0,5 0,4 0,3  2 =0,008 (19,21%) Humanities Weak 0,2 H. - W Science Excellent S. - EXC. BOYS BOYS 0,1 Wrocław Both Moderate H. - M S. - M Capital of 0,0 Science Big cities Urban S. - W S. - G Urban/rural the region Weak H. - G Rural Both Good -0,1 GIRLS GIRLS Humanities H. - EXC. Excellent -0,2 -0,3 -0,4 -0,4 -0,3 -0,2 -0,1 0,0 0,1 0,2 0,3 0,4 0,5 0,6  1 =0,03 (76,99%) CARME Conference, Rennes, 8-11 February 2011

Comparison of Results Obtained in 2003 and 2010 0,6 2010 From better 0,4 Wrocław results in the 2003 capital of the Wrocław 0,2 H.-EXC region to S.-G S.-EXC. weaker H.-G  2 =0,052 (26,94%) 0,0 H.-W S.-W results in Rural S.-M rural H.-M -0,2 Rural communes Various changes in science results: From better -0,4 • Cities: better results in 2010 No particular results in the • Urban-rural communes: weaker results in 2010 differences in capital of the -0,6 • Rural communes: weaker results in 2010 humanities results in region to 2003 and 2010 weaker -0,8 results in rural -1,0 -1,0 -0,8 -0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 communes  1 =0,05 (69,88%) CARME Conference, Rennes, 8-11 February 2011