Full-text resources of CEJSH and other databases are now available in the new Library of Science.
Visit https://bibliotekanauki.pl

Results found: 4

first rewind previous Page / 1 next fast forward last

Search results

help Sort By:

help Limit search:
first rewind previous Page / 1 next fast forward last
1

Item analysis and evaluation using a four-parameter logistic model

100%
Świst K.
Edukacja
|
2015
|
issue 3(134)
77–97
EN
The four-parameter logistic model (4PLM) assumes that even high ability examinees can make mistakes (e.g. due to carelessness). This phenomenon was reflected by the non-zero upper asymptote (d-parameter) of the IRT logistic curve. Research on 4PLM has been hampered, since the model has been considered conceptually and computationally complicated – and its usefulness has been questioned. After 25 years, following introduction of appropriate software, the psychometric characteristics of 4PLM and the model’s usefulness can be assessed more reliably. The aim of this article is to show whether 4PLM can be used to detect item-writing flaws (which introduce construct-irrelevant variance to the measurement). Analysis was conducted in two steps: (a) qualitative – assessment of compliance of items with the chosen item-writing guidelines, (b) quantitative – fitting 4PLM to compare the results with qualitative analysis – to determine whether the same items were detected as flawed. Other IRT models (3PLM and 2PLM) were also fitted to check the validity of results. Flawed items can be detected by the means of qualitative analysis as well as by 4PLM and simpler IRT models. This model is discussed from the perspective of practical use in educational research.
2

Konsekwencje błędnego określenia rodzaju zadania testowego

51%
Skórska P., Świst K., Szaleniec H.
Edukacja
|
2014
|
issue 2(127)
67–84
PL
Przedmiotem artykułu jest wpływ niepoprawnego określenia rodzaju zadania na trafność wyników egzaminu. W wyniku przeglądu zadań części matematyczno-przyrodniczej egzaminu gimnazjalnego w latach 2002–11 zidentyfikowano 9 zadań uznanych przez CKE za otwarte, mimo że treściowo i psychometrycznie funkcjonowały one jak zamknięte. Dla jednego z tych zadań przeprowadzono studium przypadku z wykorzystaniem modelowania IRT. Omówiony przypadek dowodzi zgadywania poprawnej odpowiedzi w zadaniach błędnie zakwalifikowanych jako zadania otwarte.
EN
The article demonstrates the consequences on validity when test items are misclassified as constructed-response but in fact function as selected-response. From the review of items included in the scientific part of the national exam (2002–11) that concludes 3-year lower secondary school, 9 items, which were classified as constructed-response but functioning in content and psychometrically as selected-response, were identified. One such item was examined in a case study using IRT modelling. The study showed how guessing of correct responses to items might influence test validity when using an incorrectly classified format.
3

Sex differences in guessing and item omission

51%
Świst K., Skórska P., Koniewski M., Jasińska-Maciążek A.
Edukacja
|
2015
|
issue 3(134)
48–62
EN
Guessing and item omission may be regarded as risk-taking or risk-avoidance strategies – sex specific adaptations to testing situations. In this article, these phenomena were analysed by (a) percentage of omissions by sex, (b) negative binomial regression to asses sex differences in the number of omissions, (c) c-DIF analysis using IRT-LR test and (d) linear regression using item attributes, to assess whether the c-parameter is sex differentiated by the percentage of omits (controlling item difficulty). The data set analysed were from the 2012–2014 Polish lower-secondary schools final exams, comprising tests in maths, language, science and humanities. Contrary to the vast body of literature, boys omitted items slightly more frequently than girls. Possible explanations of this finding – specific to the Polish examination system – were provided. The hypothesis of a higher c-parameter for boys did not find strong support from this study. It was shown that the c-parameter should not only be interpreted as resulting from item non-omission. This supports the modern concept of the c-parameter as a consequence not only of random guessing, but also problem solving, creative guessing or cheating.
4

Szacowanie trafności predykcyjnej ocen szkolnych z wykorzystaniem hierarchicznego modelowania liniowego

51%
Skórska P., Świst K., Szaleniec H.
Edukacja
|
2014
|
issue 3(128)
75–94
PL
Artykuł omawia zagadnienie trafności predykcyjnej ocen szkolnych dla wyników uzyskanych w egzaminach zewnętrznych. Opierając się na ocenach szkolnych i ich średniej, oszacowano stopień, w jakim można było przewidzieć uzyskany przez ucznia wynik na egzaminie gimnazjalnym w latach 2012 i 2013. Ze względu na hierarchiczną strukturę danych, w analizach wykorzystano dwupoziomowy model regresji. Do oszacowania stopnia selekcyjności oddziałów szkolnych w gimnazjum użyto modelowania IRT. Wyniki świadczą o wysokiej mocy predykcyjnej ocen szkolnych i ich średniej, zarówno na poziomie indywidualnym, jak i na poziomie oddziału szkolnego. Współczynniki korelacji wewnątrzklasowej świadczą o wysokim stopniu selekcyjności oddziałów szkolnych. Przedstawione wyniki mają znaczenie w kontekście procesów rekrutacyjnych do szkół ponadgimnazjalnych, zwłaszcza wyboru przez ucznia konkretnego oddziału szkolnego.
EN
The issue of predictive validity of school grades relative to large-scale standardized tests is described. The extent to which the results from standardized external exams at the end of lower secondary school from 2012–2013 can be predicted on the basis of school grading (both single and average grades) was estimated. Due to the hierarchical structure of analysed data, two-level linear regression models were used. In order to estimate the level of class selectivity, IRT modelling was used. The results indicated that single school grades and their averages are very good predictors for standardized test results, both at individual and class levels. Further, intra-class correlation coefficients demonstrated a high degree of class selectivity. The results are practically applicable to high school recruitment, particularly to student choice of class.
first rewind previous Page / 1 next fast forward last
JavaScript is turned off in your web browser. Turn it on to take full advantage of this site, then refresh the page.