Naslov
Attitudes Towards Future Employment and Acquired Essential Teaching Skills in Predicting Students' Achievement

Autori
Đurđević Babić, Ivana

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), ostalo

Izvornik
Education and training as basis for future employment / Matanović, Damir ; Uemura, Arata - Osijek : Fakultet za odgojne i obrazovne znanosti Sveučilišta Josipa Jurja Strossmayera u Osijeku, 2019, 81-91

ISBN
978-9536365-85-4

Skup
International conference Education and Training as Basis for Future Development

Mjesto i datum
Wakayama, Japan, 21.09.2017. - 22.09.2017

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
future employment, teaching skills, achievement, neural network, students’ attitudes

Sažetak
It is understandable why educational policies and educational institutions put primary focus on students’ achievement. Bearing in mind that students’ achievement is influenced by many factors, this research explores the idea of predicting students’ academic achievement when students’ attitudes toward their future employment and acquired teaching skills are considered as predictor variables. 229 students of class teacher studies participated in this research. Besides revealing students’ viewpoints regarding these important issues, this research, using neural network approach, aims to increase the knowledge of non-obvious factors that could be useful in estimating students’ achievement, such as students’ attitudes to their employment in labour market, in their profession, after university education. In addition, it aims to test the ability of this data mining method to build an effective classification model with acceptable generalization ability for predicting students’ achievement based on their attitudes. In order to fulfil these aims, statistical relationships between variables were determined and Multi-Layer Perceptron (MLP) and Radial Basis Function (RBF) neural network models were used. MLP neural network model with satisfactory overall classification accuracy on test sample and good generalization ability on validation sample was chosen as the most suitable model and variables with relevant impact on this model’s performance were presented.

Izvorni jezik
Engleski

Znanstvena područja
Informacijske i komunikacijske znanosti

POVEZANOST RADA