Abstract
In the present paper, we deal with the mapping of the issue of written expression among fi fth-grade pupils. In the introduction of this paper we defi ne the term "written expression" and place it in the framework of both our and foreign literature, including pointing out problems in written expression and the possibilities of its diagnosis.
The focus of the research report is to mediate the results of research project. During this project a diagnostic test was assigned to the fi fth-grade pupils of primary school and monitored the production of children in written expression.
During the creation of the diagnostic test we were inspired by some foreign diagnostic instruments and one diagnostic instrument from the Czech Republic, which is used by university students (Cimlerová, et al., 2014). Our next goal was to verify the predictive value of the proposed written expression test among fi fth grade pupils and to compare pupils' performance in terms of gender, font pattern (bound font, Comenia Script font) and the presence of specific difficulties.
Data collection was conducted by a group administration. A total of six elementary schools took part in research studies of the written expression.
We have collected data from 125 fi fthgrade students. We randomly selected 48 pupils without specifi c diffi culties and all 8 pupils with specifi c diffi culties were randomly assigned to the sample in which we performed a static data evaluation.
In the test of written expression, a statistically signifi cant difference was found in the results of the subtest Number of words, Number of sentences, Number of conjectures, Composition of text. Text content and overall score.
Girls scored statistically signifi cantly better than boys in these categories. There were no statistically signifi cant differences between the subgroups divided by the term of font pattern in any subtests nor the overall score.
There were statistically signifi cant differences between the group of pupils with specifi c diffi culties and the group of pupils without specifi c diffi culties only in the subtest Number of conjectures. Differences between these groups have not been confi rmed in any other subtests or overall score.
The reliability of the test is 0.752 (Cronbach's alpha). Correlation relationships have been established between the established evaluation criteria, with the exception of the Spelling and morphology, the Content of the text and the Paraphrase of sentences.
The results obtained in this study will be used to construct a defi nitive tool for evaluation of written expression.