1,721,011 research outputs found
Investigating gender differences in mathematics by performance levels in the Italian school system
Full text linkAmongst the extensive research conducted about girls’ lack of interest in STEM subjects, an issue that is frequently examined is the learning of mathematics. This research investigates the gender gap in mathematics among Italian students based on performance levels, using standardised large-scale data from INVALSI tests. We performed a quantile regression to better understand the differences along the entire score distribution. A latent class model was then estimated to identify groups of students with similar performance levels, taking into account the gender covariate. The results indicate that boys are already ahead from primary education and that there is a general decrease in the performance as students progress through the education stages
The Hellinger Distance within Posterior Predictive Assessment for Investigating Multidimensionality in IRT Models
Full text linkUnder the Bayesian approach, posterior predictive model checking (PPMC) has become a popular tool for fit assessment of item response theory (IRT) models. In this study, we propose
the use of the Hellinger distance within PPMC to quantify the distance between the realized and the predictive distribution of the model-based covariance for item pairs.
Specifically, the case of multidimensional data analyzed with a unidimensional approach is taken into account. The results of the simulation study show the effectiveness of the
method in detecting model misfit and the sensitivity to the trait correlations. An application to real data on tourism perceptions shows the feasibility of the method in practice and
especially the capability of detecting potential misfit attributed to specific items
LA STRUTTURA DELLA PROVA INVALSI DI ITALIANO NEI DIVERSI LIVELLI SCOLASTICI: UN APPROCCIO MULTIDIMENSIONALE
No full textAbstrac
Item response theory model fit assessment via posterior predictive checking: Two case studies
No full text
Modelling subjective well-being dimensions through an IRT bifactor model: Evidences from an Italian study
Full text linkThe investigation of individual and community well-being has acquired a particular relevance over time for governments to develop strategies and identify resources for improving standards of living. To this aim, it is necessary to analyse changes at the overall level and examine how subjective well-being differs between different sub-groups of the population as well as across local areas. Using data measuring the well-being of residents in the Romagna area (Italy), we propose a multidimensional approach within the item response theory (IRT) framework to estimate an overall score of community Subjective Well-Being (SWB) and individual scores re ecting specific dimensions, taking into account for the ordinal polytomous nature of the items. The results show that aspects dealing with Life Evaluation mainly affect the overall SWB, while issues pertaining to Community and Environment are less important. The proposed approach is effective in developing an indicator which takes into account the multidimensionality of SWB and estimating individual scores reflecting the heterogeneity among residents
La performance in matematica degli studenti del V anno di scuola secondaria di secondo grado: un’analisi multilivello sui dati dell’anno scolastico 2018-2019
No full textThe performance in mathematics of students in the fifth year of upper secondary school : a multilevel analysis on the data of the school year 2018-2019 · The main aim of this paper was the analysis of the Italian students’ Mathematics achievements at the end of upper secondary school. By applying a hierarchical linear model to a sample of 36.589 thirteenth-graders nested within 990 schools who participated in the INVALSI tests in 2019, it was possible to find out the determinants of pupils’ performance both at individual and at class and school level. The results revealed that a three-level Random Slope Model with cross-level interactions between covariates was the best-fitting model. It assumes that the effects of gender, average class ESCS (Economic, Social and Cultural Status) and oral assessment in Mathematics on students’ score vary from class to class and from school to school, meaning that the influence of said factors on students’ performance depends on the specific group to which they belong. All the other independent variables which contribute to pupils’ score are assumed to have the same effect independently of the specific class or school examined. Finally, the results from the investigation of variance decomposition showed that the proportion of the total variation in Mathematics achievement explained by school and class-level covariates, which represent the context effects, is larger than the amount of variability accounted for by individual characteristics
Automated Test Assembly for Large-Scale Standardized Assessments: Practical Issues and Possible Solutions
Full text linkIn testing situations, automated test assembly (ATA) is used to assemble single or multiple test forms that share the same psychometric characteristics, given a set of specific constraints, by means of specific solvers. However, in complex situations, which are typical of large-scale assessments, ATA models may be infeasible due to the large number of decision variables and constraints involved in the problem. The purpose of this paper is to formalize a standard procedure and two different strategies—namely, additive and subtractive—for overcoming practical ATA concerns with large-scale assessments and to show their effectiveness in two case studies. The MAXIMIN and MINIMAX ATA methods are used to assemble multiple test forms based on item response theory models for binary data. The main results show that the additive strategy is able to identify the specific constraints that make the model infeasible, while the subtractive strategy is a faster but less accurate process, which may not always be optimal. Overall, the procedures are able to produce parallel test forms with similar measurement precision and contents, and they minimize the number of items shared among the test forms. Further research could be done to investigate the properties of the proposed approaches under more complex testing conditions, such as multi-stage testing, and to blend the proposed approaches in order to obtain the solution that satisfies the largest set of constraint
Posterior Predictive Assessment for Item Response Theory Models: A Proposal Based on the Hellinger Distance
No full textBayesian posterior predictive assessment has received considerable attention for investigating specific aspects of fit of item response theory models. In fact, this approach is easy to apply within Markov chain Monte Carlo estimation, it is flexible and free from distributional assumptions. In its classical implementation, the method is based on graphical analysis and the estimation of posterior predictive p-values to investigate the degree to which observed data are expected under the model, given a discrepancy measure. In this work, we propose to quantify the distance between the realized and the predictive distributions of the discrepancy measure based on the Hellinger distance. The results show that this measure is able to provide clear recommendations about the investigated aspects of model fit
Posterior Predictive Assessment of IRT Models via the Hellinger Distance: A Simulation Study
No full textFit assessment of item response theory models is a crucial issue. In recent years, posterior predictive model checking has become a popular tool for investigating overall model fit and potential misfit due to specific items. Different approaches rely on graphical analysis, posterior predictive p-values, the relative entropy and, more recently, the Hellinger distance. In this study, we focus on the performance of the Hellinger distance in the case multidimensional data are analyzed with a unidimensional approach. In particular, we consider the case of three
latent dimensions. A simulation study is conducted to show the effectiveness
of the method. Finally, the results of an empirical application to potential three-dimensional data are discussed
- …
