The European Union Statistics on Income and Living Conditions (EU-SILC) set of surveys are an important source of comparative statistical data. EU-SILC provides data on income, living conditions, poverty and social exclusion, material deprivation: topics of growing interest to scholars in Europe and elsewhere. EU-SILC surveys are fielded in 29 European countries and coordinated by Eurostat. Although the survey is harmonised, the individual level microdata consists of many dissimilarities across participating countries because of different national conditions, methods of data collection and/or data processing. The aim of this article is to discuss the opportunities and limitations of EU-SILC datasets. In addition to discussing the development, methodology and basic pitfalls of EU-SILC, this article focuses on (a) income variables, (b) differences in income among countries and (c) impact of income differentials on data comparability. The main problems of income data may be summarised as follows. 1) Some countries use registers to report income variables while others obtain this information from interviews, and this difference lowers their comparability. 2) The incidence of negative or zero values makes the construction of poverty and inequality measures difficult. 3) There are national differences in the net-to-gross income conversion procedure. This study shows using a four country analysis that the net-to-gross conversion procedure overestimates gross wages in two countries and underestimates it in two others. Notwithstanding these methodological issues, EU-SILC is an important resource for the comparative study of income., Martina Mysíková., and Obsahuje bibliografii a bibliografické odkazy
The European Union Statistics on Income and Living Conditions (EU-SILC) set of surveys are an important source of comparative statistical data. EU-SILC provides data on income, living conditions, poverty and social exclusion, material deprivation: topics of growing interest to scholars in Europe and elsewhere. EU-SILC surveys are fielded in 29 European countries and coordinated by Eurostat. Although the survey is harmonised, the individual level microdata consists of many dissimilarities across participating countries because of different national conditions, methods of data collection and/or data processing. The aim of this article is to discuss the opportunities and limitations of EU-SILC datasets. In addition to discussing the development, methodology and basic pitfalls of EU-SILC, this article focuses on (a) income variables, (b) differences in income among countries and (c) impact of income differentials on data comparability. The main problems of income data may be summarised as follows. 1) Some countries use registers to report income variables while others obtain this information from interviews, and this difference lowers their comparability. 2) The incidence of negative or zero values makes the construction of poverty and inequality measures difficult. 3) There are national differences in the net-to-gross income conversion procedure. This study shows using a four country analysis that the net-to-gross conversion procedure overestimates gross wages in two countries and underestimates it in two others. Notwithstanding these methodological issues, EU-SILC is an important resource for the comparative study of income.
Súčasný vývoj v oblasti psychologického hodnotenia zdôrazňuje zlepšovanie metodológie a význam zvyšovania efektívnosti. Algoritmy počítačového adaptívneho testovania (CAT) založené na teórii odpovede na položku (IRT) ponúkajú zaujímavé príležitosti pre súčasnú optimalizáciu ako presnosti, tak aj efektívnosti merania. Tento článok prezentuje zistenia 15 výskumných štúdií z oblasti testovania schopností, klinickej psychológie, testovania osobnosti a zdravotníckej starostlivosti zameraných na skúmanie reliability, užitočnosti (v zmysle úspory položiek) a validity (v zmysle korelácií s existujúcimi nástrojmi) CAT. Celkovo sú zistenia povzbudivé – CAT poskytuje efektívny prostriedok pre získanie optimálneho množstva informácie potrebnej pre zodpovedanie posudzovanej otázky, a to využitím minimálneho množstva času a/alebo počtu položiek pre získanie danej informácie. CAT skóre silno korelovalo so skóre z celej položkovej banky (rozpätie r = 0,83 – 0,99) a stredne silno so zaužívanými nástrojmi (rozpätie r = 0,58 – 0,83) poskytujúc dôkazy pre reliabilitu, validitu a porovnateľnosť adaptívnych nástrojov. Avšak tieto výsledky sú založené hlavne na CAT simulačných štúdiách a preto sú potrebné ďalšie štúdie zahŕňajúce administráciu skutočných testov živým respondentom, aby tieto zistenia potvrdili. and Computerized adaptive testing: precision, validity and efficiency
Present developments in the area of psychological assessment place emphasis on methodological improvements and the importance of increasing effectiveness. Computerized adaptive testing (CAT) algorithms based on item response theory (IRT) offer attractive opportunities for simultaneously optimizing both measurement precision and efficiency. This article presents findings of 15 research studies from field of ability testing, clinical psychology, personality testing and health care designed to explore the reliability, utility (in terms of item savings) and validity (in terms of correlations with existing tools) of CAT. Overall, the findings are encouraging – CAT provides an efective means to gain an optimal amount of information needed to answer an assessment question, while keeping time and/or number of items required to obtain that information at a minimum. CAT score correlated high with score from the full item bank (range r = 0,83 – 0,99) and moderately with established measures (range r = 0,58 – 0,83) provide the evidence for reliability, validity and comparability of adaptive tools. However, these results are based mainly on CAT simulation studies and therefore additional Live-CAT studies (involves the administration of real tests to live examinees) are needed to confirm this pattern of findings.