Search

Jedan od glavnih ciljeva kurikularne reforme u Republici Hrvatskoj jest povećanje kompetencije rješavanja problema. Kako bi se pratila dosegnutost navedenoga cilja provedeno je online ispitivanje komptencije rješavanja problema na gotovo 7000 učenika. U knjizi je detaljno opisan razvoj konceptualnoga okvira za izradu ispita, razvoj ispita, provedba ispitivanja, određivanje razina postignuća te su pirkazani i interpretirani rezultati provedenoga ispitivanja.

When trying to improve the quality of education and raise students’ achievements, some policy makers suggest that students spending more time in school could help them accomplish better school results. Research differentiates between instructional and non-instructional time spent in school. Instructional time is devoted to tutoring, while non-instructional time is used for organizational or administrative activities, classroom management and discipline. There is a limited body of research that studies the association between time spent in school and students’ achievement and the results are somewhat conflicted. In the last few years, comprehensive educational reform in Croatia has started a lively debate about whether students spend enough time in school. Therefore, the goal of this research is to examine the association between time spent in school and cognitive as well as non-cognitive variables. The analysis were done by using the TIMSS 2015 and TIMSS 2019 dataset. The participants of the TIMSS 2015 and TIMSS 2019 in Croatia were the 4th grade students, hence the study was focused on this age group. As a dependent cognitive variable, students’ achievement in math and science were taken and calculated as a test score on the TIMSS 2015 and TIMSS 2019. Students’ sense of school belonging and student bullying were used as non- cognitive dependent variables. Since the TIMSS results do not include data about total time spent in school (both instructional and non- instructional time), only the instructional time was used. Data were analyzed by using regression analyses in R intsvy package made specifically for the international education surveys such as TIMSS. In both years, 2019 and 2015, multiple regression models weren’t significant meaning there is no connection between total instructional time and used measures. Since there is a very limited scientific evidence on this topic when it comes to Croatian population, the insights from this study would be very helpful in shaping educational policy and planning especially if it is taken into account that curriculum reform is going on.

Knjiga donosi rezultate OECD-ova PISA istraživanja za ciklus 2018 kojim se procjenjuju znanja i vještine petnaestogodišnjaka uključenih u obrazovni sustav. Naglasak je u ovome ciklusu bio na čitalačkoj pismenosti.

Rad je prikaz rezultata projekta Nacionalnog centra za vanjsko vrednovanje obrazovanja (NCVVO) kojim se nastojalo odgovoriti na nedostatak sustavne pomoći djeci neizvornim govornicima hrvatskog jezika i njihovim nastavnicima u osnovnom obrazovanju na hrvatskom jeziku u RH. Kao rezultat projekta oblikovani su zadaci za provjeravanje poznavanja hrvatskog jezika s detaljnim pojašnjenjima koja mogu poslužiti nastavnicima kao metodički predložak za izradu materijala u nastavi, za djecu kojima su za nesmetano praćenje sadržaja na hrvatskom jeziku potrebni dodatni sati nastave hrvatskog. Provjeravanje znanja hrvatskog jezika za djecu nečitače temelji se na procjeni vještina govorenja i razumijevanja hrvatskog jezika. Vještina govorenja može se ispitati govornom interakcijom, imenovanjem pojmova i pripovijedanjem. Vještina jezičnog razumijevanja može se ispitati razumijevanjem riječi, razumijevanjem i slijeđenjem govornih uputa te razumijevanjem slušanoga teksta u slušanju priče kroz odgovore na pitanja. U sastavljanju ispitnog materijala vodilo se računa o leksičkom i gramatičkom znaju koje može imati inojezični učenik hrvatskog jezika predškolske i rane osnovnoškolske dobi imajući u vidu jezično-kognitivne sposobnosti djece u toj dobi te posebno činjenicu da proces usvajanja jezičnih kategorija još nije dovršen u prvom jeziku.

Nacionalni centar za vanjsko vrednovanje obrazovanja je tijekom školske godine 2016./2017. proveo sveobuhvatno ispitivanje kojemu je cilj bio napraviti analizu utjecaja ispita državne mature na nastavu u srednjim školama. Korištena je metoda upitnika kako bi se moglo zahvatiti što veći udio populacije srednjoškolskih nastavnika. Kako bi upitnik sadržavao što relevantnije čestice za ispitivanje navedenoga cilja, uz stručni tim Centra, na izradi čestica sudjelovale su još i tri stručne radne skupine od kojih je svaka bila zadužena za određeno područje ispitivanja. U konačnici je sastavljen upitnik koji se sastojao od zajedničkog dijela te dijela koji su nastavnici ispunjavali ovisno o tome koji predmet predaju. Upitnik je primijenjen on-line na ukupno 4716 sudionika. Od toga je bilo: 3934 nastavnika, 373 ravnatelja te 238 stručnih suradnika. Analizama je utvrđeno da se radi o reprezentativnom uzorku te da je dobivene nalaze moguće generalizirati. Dobiveni podatci analizirani su različitim statističkim postupcima, a ovdje su prikazani deskriptivni i korelacijski parametri. Ovo istraživanje daje cjelovit uvid u razmišljanja nastavnika, ravnatelja i stručnih suradnika o brojnim pitanjima vezanim za državnu maturu. Nalazi ovih istraživanja zasigurno će biti zanimljivi populaciji na kojoj je ono provedeno, ali i nosiocima obrazovne politike. To je osobito važno u kontekstu reforme obrazovanja budući da odgovori nastavnika daju jasne smjernice o mogućim načinima djelovanja na kvalitetu nastavnog procesa.

Scoring validity of writing test implies the stable, appropriate and consensual rating criteria. This research is focused on determination and identifying elements of a text structures that can be objectively determined as a rating foundation for part of reliable writing tasks. The study was conducted on a representative sample of 300 L1 student writing papers from 8 th grade of primary school. The writing paper was a subtest of the language test administrated within the scope of Croatian national assessment. The task refers to writing and organizing short newspaper article with narrowly defined content points. Sampled student papers are recoded for the present research purposes according to four sets of objective predictor variables, namely: content relevant for newspapers article, sentence structure, morphology and orthography rules and number of words. Selected criteria were: overall test achievement, 7th grade ; achievement and overall school achievement, as well as scoring procedures based on two criteria – sentence consistency and writing style, applied by the independent graders. Analyses will aim to validate the contribution extent of these four objective predictors in assessing and scoring highly structured writing tasks.

Rad Nesročnost u hrvatskome jeziku kao vrsta morfoloških ovisnosti ima za cilj definirati i opisati gramatički proces nesročnosti u hrvatskome jeziku. Temeljna je metodološka pretpostavka da se kroz iznimke i kršenje općeg pravila slaganja otkrivaju općenite značajke procesa sročnosti. Teorijski dio rada razmatra postojeće definicije sročnosti, problem paradigmatske i sintagmatske naravi gramatičkih kategorija, a proces nesročnosti sagledava se kroz pojam morfoloških ovisnosti koje problem nesročnosti svrstavaju u teorijski kontekst ovisnosne gramatike. Hijerarhijski odnosi među jezičnim elementima u sintaktičkim domenama sročnosti utvrđeni su na temelju modela hijerarhije sročnosti. Prožimanjem teorijskih modela hijerarhije sročnosti i morfoloških ovisnosti predložena je shema X – Z1– Z2 –Y u kojoj se oznaka X odnosi na paradigmatski određenu gramatičku kategoriju, Y na sintagmatske gramatičke susljednih flektivnih kategorija, a oznaka Z na gramatičke elementa unutarnje (Z1) i elementa vanjske sročnosti (Z2). Razvijeni je teorijski model u ovome radu primijenjen je na problem nesročonosti imenica muškoga roda odnosno složenih kratica muškoga roda. Jezična analiza na odabranome jezičnome materijalu ima dva osnovna rezultata. Prvi rezultat su pravila za funkcioniranje procesa sročnosti u hrvatskome jeziku, a drugi dopuna modela hijerarhije sročnosti s obzirom na specifičnosti imenica muškoga roda kao izvora nesročnosti. Iako je model hijerarhije sročnosti primijenjen na mnoge slavenske jezike, rezultati ove analize na jezičnoj građi hrvatskoga jezika upućuju na teorijsku dopunu pravila hijerarhije.

Croatian educational system is facing the process of introducing student assessment system in order to determine information on the long term progress in the academic performance and factors that affect students’ accomplishment. The cooperation between CITO and Croatian National Centre for External Evaluation of Education (NCEEE) represents the important progress in assessment within the international project Development of Instruments in Croatian National Assessment. The main objective of that project was the education and capacity building of NCEEE experts. The most important activity within the project was the assessment of educational achievements of the 2nd grade gymnasium programme students. Along with broadly specified goal, the aim of the project was to analyze sample data in order to determine some of the factors that affect language domain constructs and possible external factors influencing students’ performance. The current research analyses the data on Croatian language performance in four tested domains: grammar, theory of literature, reading of literary and non-literary text. The results justify used division of tested domains. In addition, some demographic variables influence the students’ achievement. One of the specific influences is the one of gender on the reading performance results, which could be derived from the thematic of used texts. The results of presented analysis could be used to improve the validity of instruments and quality of reading framework for the long term assessment as well as to provide considerable information on target student population. Croatian educational system is facing the process of introducing student assessment system in order to determine information on the long term progress in the academic performance and factors that affect students’ accomplishment. The cooperation between CITO and Croatian National Centre for External Evaluation of Education (NCEEE) represents the important progress in assessment within the international project Development of Instruments in Croatian National Assessment. The main objective of that project was the education and capacity building of NCEEE experts. The most important activity within the project was the assessment of educational achievements of the 2nd grade gymnasium programme students. Along with broadly specified goal, the aim of the project was to analyze sample data in order to determine some of the factors that affect language domain constructs and possible external factors influencing students’ performance. The current research analyses the data on Croatian language performance in four tested domains: grammar, theory of literature, reading of literary and non-literary text. The results justify used division of tested domains. In addition, some demographic variables influence the students’ achievement. One of the specific influences is the one of gender on the reading performance results, which could be derived from the thematic of used texts. The results of presented analysis could be used to improve the validity of instruments and quality of reading framework for the long term assessment as well as to provide considerable information on target student population.

Izvješće o projektu Razvoj i strategija nacionalnih ispita iznosi ciljeve, aktivnosti i rezultate Projekta. Cilj je Projekta bio educirati djelatnike Nacionalnog centra za vanjsko vrednovanje obrazovanja i njegove vanjske suradnike u području vanjskoga vrjednovanja. U okviru edukacije provedeni su nacionalni ispiti – nacionalna procjena postignuća na uzorku učenika drugoga razreda škola s gimnazijskim programom. U Izvješću su opisani rezultati izradbe testova iz hrvatskoga jezika i matematike, počevši od odabira sadržajnih područja ispitivanja i obrazovnih ishoda pa do same izradbe ispitnih zadataka i oblikovanja nacrta testova. Isto je tako opisan proces izradbe popratnih upitnika od osmišljavanja ciljeva preko sastavljanja pitanja do opisa rezultata. Rezultati nacionalnih ispita – nacionalne procjene postignuća analizirani su prema klasičnoj teoriji testova i teoriji odgovora na zadatak. Ovim su Projektom prvi put provedeni formativni ispiti u Republici Hrvatskoj čime su postavljeni temelji za razvoj nacionalnih ispita s ciljem poboljšanja kvalitete hrvatskoga obrazovnog sustava.

Doktorski rad Sintaktička nesročnost u hrvatskome jeziku ima za cilj definirati i opisati gramatički proces nesročnosti. u hrvatskome jeziku. Temeljna je metodološka pretpostavka da se kroz iznimke i kršenje općeg pravila slaganja otkrivaju općenite značajke procesa sročnosti. Teorijski dio rada nakon uvoda (2. poglavlje) razmatra postojeće definicije sročnosti, problem paradigmatske i sintagmatske naravi gramatičkih kategorija i pojam sintaktisa. Hijerarhijski odnosi među jezičnim elementima u sintaktičkim domenama sročnosti utvrđeni su na temelju modela hijerarhije sročnosti. Prožimanjem ovih teorijskih modela i postavki predložena je shema X – Z1– Z2 –Y u kojoj se oznaka X odnosi na paradigmatski određenu gramatičku kategoriju, Y na sintagmatske gramatičke susljednih flektivnih kategorija, a oznaka Z na gramatičke elementa unutarnje (Z1) i elementa vanjske sročnosti (Z2). U drugom dijelu teorijskog dijela rada (3. poglavlje) proces nesročnosti sagledava se kroz pojam morfoloških ovisnosti koje problem nesročnosti svrstavaju u teorijski kontekst ovisnosne gramatike. Sukladno teorijskom dijelu rada jezična je razrada problema podijeljena u dva dijela. U prvome su dijelu analize (poglavlje 4) izdvojene su najčešće nesročne imenice muškoga ženskoga i srednjega roda i imenice s oznakom zbrojine. Analizom na računalnom korpusu suvremenih hrvatskih tekstova (HNK) utvrđeno je da se na temelju prethodno postavljene sheme nesročnost u hrvatskome jeziku može opisati sa šest uzoraka. Sustav uzoraka polazište je za analizu nestočnosti u kontekstu ovisnosne gramatike (poglavlje 5). Ovisnosni su odnosi u domenama sročnosti opisane sintaktičkim i morfološkim ovisnosnim pravilima. Na domene sročnosti odnose tri osnovna sintaktička ovisnosna pravila. Površinski se sintagmatski odnosi upotpunjuju morfološkim ovisnosnim pravilima. Morfološkim ograničenjima pripadaju i pravila koja opisuju podudaranje gramatičkih kategorija morfološke podređenice i podredbenice te ona ograničenja koja odnose na nesročnost.

Ovaj je rad prikaz knjige V. Law A History of linguistics in Europe koja daje novi sustavni pregled razvoja lingvističke misli u Europi do kraja renesanse.

Acronyms cause interest from different linguistics perception such as lexical properties up to their sociolinguistic impact. Beside these in Croatian, as language with rich inflectional system, acronyms are interesting due to peculiarities in their morphosyntactic behavior. This work is based on corpus data of most frequent acronyms from Croatian National Corpus (hnk.ffzg.hr). In this paper we will face two problems of the inflection of the acronyms in Croatian. The first one is complete absence of declension in some very frequent acronyms, and the second one are the principles on which declinable acronyms acquire their inflectional class. This analysis reveals that 1) indeclinability of acronyms is based on graphical and pronunciation inconsistencies of derivational structures ; 2) the acronyms that inflect, demonstrate regular inflectional behavior. Data disclose that acronyms belong mostly to inflection class A (masculine class) and E (feminine class) and avoid inflectional paradigm for neuter nouns. In that respect they behave like foreign words that try to preserve their original phonological structure of lexical stem. According to revealed inflectional properties of the acronyms, we will classify acronyms according to inflectional patterns that are adequate for automatic generation of acronymic word-forms and their inclusion in Croatian Morphological Lexicon (hml.ffzg.hr).

Ovim je radom i iscrpnim analizama u njemu pokazano da akronimi mogu imati morfosintaktičko ponašanje različito od svojih temeljnih oblika čime se nesumnjivo opravdava proučavanje njihovih zasebnih jezičnih osobina te ih se ne smije promatrati kao puke odraze imenica ili imeničnih skupina.

This paper questions the validity of tests considered to determine the prototypical senses of polysemous lexemes which involve linguistic frequencies and assume that the highest frequency of use corresponds to the center of the lexical category. Using a Croatian computer corpus and corpus linguistics methodology frequency data for three Croatian nouns are analyzed and compared to native speaker intuition in the establishment of prototypical meanings of these lexemes. It is pointed out that reasons for the existence of prototypicality in lexemes are cognitive in nature and that there are at least some linguistic categories in which frequency of use is not an indicator of prototypicality.

Based on the thirty million Croatian National Corpus, the paper investigates the frequency of metaphorical meanings in examples generated by two basic conceptual metaphors - People are plants and Ideas are plants - as described in Lakoff and Johnson (1980) and Lakoff and Turner (1989). Taking as the starting point the assumption that metaphorical meanings of lexemes are in language use sometimes more frequently employed than the non-metaphorical ones, several lexemes representing metaphorical concepts have been chosen, and the frequences of their metaphorical and non-metaphorical senses have been extracted from the Croatian corpus. The relationship of these frequencies will be explored, with particular emphasis on the fact that in many cases the most frequently occuring sense of a lexeme (i.e. the methaphorical one) is not its prototypical sense. For instance, although the prototypical sense of the Croatian noun plod (English fruit) is the one denoting � part of a plant� , metaphorical senses of the same lexeme in the corpus account for 72.8% percent of the total number of occurences of the lemma plod (fruit). Also, senses in particular lexical entries, as they are defined in a monolingual Croatian dictionary (Anić, 1991), will be compared with the concordances obtained from the corpora. The majority of cases will show that the dictionary does not define all metaphorical meanings found in the corpora examples. Therefore, new meanings have to be defined. Meanings will be extracted from concordances (KWIC), as has been described in Sinclair (1991) and Ooi (1998). Since results of corpora research are often applied to lexicography, data on basic conceptual metaphors from the corpus can also serve as a starting point for the creation of lexicographic entries in a monolingual Croatian dictionary, taking into account frequencies of both metaphorical and non-metaphorical senses and the distribution of grammatical categories, as well as the relationship between them.

Metaphorical meanings in language use are often more frequently employed than the non-metaphorical ones. This assumption can best be examined using a computer corpus. Based on the test version of Croatian National Corpus which includes contemporary texts of 7.67 million tokens, the article investigates the frequency of examples of some basic conceptual metaphors, as described in Lakoff and Johnson (1980) and Lakoff and Turner (1989). Some lexemes representing metaphorical concepts have been chosen, with respect to the frequencies of their metaphorical and non-metaphorical meanings in two Croatian corpora. The relationship of these frequencies will be explored. Meanings in particular lexical entries, as defined in a monolingual Croatian dictionary, will be compared with the concordances obtained from the corpora. The majority of cases will show that the dictionary does not define all metaphorical meanings found in the examples from the corpora. Therefore, new meanings have to be defined. Meanings will be extracted from concordances (KWIC), as has been described in Sinclair, (1991) and Oai (1997) Frequency data on basic conceptual metaphors acquired from the corpus can also serve as a starting point for the creation of a semantic description or lexicographic entry in a monolingual Croatian dictionary. The results of the research prove that use of large corpora is necessary in contemporary lexicography, as they are indicators of the actual state of language use.

U radu se daje pregled zastupljenosti i distribucije akronima u hrvatskome na temelju pretraživanja probne inačice Hrvatskoga nacionalnoga korpusa.

Metaforična su značenja u jezičnoj uporabi uglavnom čestotnija od nemetaforičnih. Te se pretpostavke najbolje mogu istražiti na građi računalnog jezičnog korpusa. Na građi 30 -milijunskog Hrvatskoga nacionalnog korpusa izrazit će se čestotnost primjera osnovnih konceptualnih metafora ljudi su biljke i ideje su biljke. Budući da se rezultati korpusnih istraživanja vrlo često primjenjuju u leksikografiji, podaci o čestotnosti metaforičnih različnica iz ovog korpusa poslužit će kao polazište za sastavljanje leksikografske natuknice jednojezičnog rječnika hrvatskog jezika. Rezultati ovog istraživanja pokazat će kako je u suvremenoj leksikografiji neizbježno korištenje višemilijunskim korpusima jer oni u velikoj mjeri pokazju pravo stanje u uporabi nekog jezika.