Kısa cevaplı ve çoktan seçmeli maddelerden oluşan testlerin psikometrik özelliklerinin karşılıklı olarak incelenmesi
Başlık çevirisi mevcut değil.
- Tez No: 21140
- Danışmanlar: PROF.DR. İLHAN AKHUN
- Tez Türü: Doktora
- Konular: Eğitim ve Öğretim, Education and Training
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1992
- Dil: Türkçe
- Üniversite: Hacettepe Üniversitesi
- Enstitü: Sosyal Bilimler Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 189
Özet
İİİ SUMMARY In this study, psychometric charecteristics of dif ferent types of items were studied. The measuring powers of short-answer and multiple-choice items were compared in terms of native language abilities of primary school children. The influence of variation of the item structure on the item and test statistics has also been investi gated. While comparing the measuring power of language ability of the two tests, the reliability and validity coefficients of the tests and sub-tests relating to the sub-problems were also calculated. The internal consis tency coefficient of KR-20 was used in comparing reliabi lity of the two tests. Item-test and test-criterion cor relations and analysis factors in comparing of validity were used. The sample was 240 fifth grade students. They were equally taken from three primary schools, which are at three different social-economical and culturel levels. 80 student from each school were divided into two groups. Each consisiiag of 40 students, in terms of grades received in their Turkish lesson and their sexes* A short- answer test of language ability was given to the first group, multiple-choice test of language ability was given to the second group. 120 students took the multiple choise test of language ability. The short-answer test and the multiple-choice test of ability, were constructed by a common specification tablo. Each test consisted of 72 items measuring the same at titudes with the equally items. While preparing the language ability tests in a greativ number of- items were written and the ones which were on the last test were chosen by opinion of the classroom teachers. Factor analysis were being done by using“SPSS”packet programme and the other analyses were being done by hand using a calculator. In data analysing tests of sig nificant were used for testing the differences between the means, standart deviations, KR-20 internal consistency coefficients, the correlation of item by item and between attitudes and item-test. According to group received in Turkish courses grades whether the test scores showed dif ferences or not were tested with the analysis of variance. As a result of factor analysis it was found that the average difficulties, the correlation between attitudes, factoral structure and the average of applying duration of short-answer and multiple-choice language ability tests were found to be significant. However, KR-20 internal con sistency coefficient, the median values of item dis crimination indexes and the differences between variation of these tests weren't significant. There were structural differences between the paral lel attitudes which showed significant differences from point of variations, item discriminations and the correla tions between the other attitutes. Although, the two tests were seen to be the same in terms of validity and rel iabi 1 i ty, in structurel analysis, it was found that, the multiple-choice test was more homogen and simple dimensional than short-answer tests. On the other hand, it was found that the short-answer test was more difficult, its applying duration was more longer and was more powerful from the point of predicting the success Turkish courses than multiple choice test. It was
Özet (Çeviri)
İİİ SUMMARY In this study, psychometric charecteristics of dif ferent types of items were studied. The measuring powers of short-answer and multiple-choice items were compared in terms of native language abilities of primary school children. The influence of variation of the item structure on the item and test statistics has also been investi gated. While comparing the measuring power of language ability of the two tests, the reliability and validity coefficients of the tests and sub-tests relating to the sub-problems were also calculated. The internal consis tency coefficient of KR-20 was used in comparing reliabi lity of the two tests. Item-test and test-criterion cor relations and analysis factors in comparing of validity were used. The sample was 240 fifth grade students. They were equally taken from three primary schools, which are at three different social-economical and culturel levels. 80 student from each school were divided into two groups. Each consisiiag of 40 students, in terms of grades received in their Turkish lesson and their sexes* A short- answer test of language ability was given to the first group, multiple-choice test of language ability was given to the second group. 120 students took the multiple choise test of language ability. The short-answer test and the multiple-choice test of ability, were constructed by a common specification tablo. Each test consisted of 72 items measuring the same at titudes with the equally items. While preparing the language ability tests in a greativ number of- items were written and the ones which were on the last test were chosen by opinion of the classroom teachers. Factor analysis were being done by using“SPSS”packet programme and the other analyses were being done by hand using a calculator. In data analysing tests of sig nificant were used for testing the differences between the means, standart deviations, KR-20 internal consistency coefficients, the correlation of item by item and between attitudes and item-test. According to group received in Turkish courses grades whether the test scores showed dif ferences or not were tested with the analysis of variance. As a result of factor analysis it was found that the average difficulties, the correlation between attitudes, factoral structure and the average of applying duration of short-answer and multiple-choice language ability tests were found to be significant. However, KR-20 internal con sistency coefficient, the median values of item dis crimination indexes and the differences between variation of these tests weren't significant. There were structural differences between the paral lel attitudes which showed significant differences from point of variations, item discriminations and the correla tions between the other attitutes. Although, the two tests were seen to be the same in terms of validity and rel iabi 1 i ty, in structurel analysis, it was found that, the multiple-choice test was more homogen and simple dimensional than short-answer tests. On the other hand, it was found that the short-answer test was more difficult, its applying duration was more longer and was more powerful from the point of predicting the success Turkish courses than multiple choice test. It was
Benzer Tezler
- Aynı davranışı ölçmeye yönelik kısa cevaplı ve çoktan seçmeli maddelerden oluşan testlerin psikometrik özelliklerinin incelenmesi
In This study by comparing the psychometric properties of two different types of items one reguiring selection and the other reguiring short answers measuring thesome behaviour
NİMET UĞURLU
- Kısa cevaplı, çoktan seçmeli ve kompozisyon tipi maddelerden oluşan testlerin psikometrik özelliklerinin karşılaştırmalı olarak incelenmesi
Constrative analysis of psychometric properties of short answer, multiple - choice and essay type tests
DEVRİM ÖZDEMİR
Yüksek Lisans
Türkçe
1997
Eğitim ve ÖğretimHacettepe ÜniversitesiEğitim Bilimleri Ana Bilim Dalı
DOÇ. DR. A. ATA TEZBAŞARAN
- Kısa cevaplı ve çoktan seçmeli testlerin madde ve test istatistiklerinin karşılaştırılması
The Comparison of item and test statistics of short-answer and multiple choice tests
MUSTAFA AKDENİZ
Yüksek Lisans
Türkçe
2003
Eğitim ve ÖğretimHacettepe ÜniversitesiEğitim Bilimleri Ana Bilim Dalı
YRD. DOÇ. DR. HÜLYA KELECİOĞLU
- Kısa cevaplı ve çoktan seçmeli maddelerin '1-0' ve ağırlıklı puanlama yöntemleriyle puanlanmasının testin psikometrik özellikleri açısından incelenmesi
Analysis of short-answered and multiple-choice items scored via '1-0' and weighted scoring methods according to psychometric characteristics of tests
GÖKSU GÖZEN
Yüksek Lisans
Türkçe
2002
Eğitim ve ÖğretimHacettepe ÜniversitesiEğitimde Ölçme ve Değerlendirme Ana Bilim Dalı
DOÇ. DR. SELAHATTİN GELBAL
- Students' and teachers' perceptions of the test techniques used to assess language skills at university level
Öğrenci ve öğretmenlerin üniversite seviyesinde dil becerilerini değerlendirmede kullanılan test tekniklerine ilişkin algılamaları
KÜRŞAT CESUR
Yüksek Lisans
İngilizce
2008
DilbilimÇanakkale Onsekiz Mart ÜniversitesiYabancı Diller Eğitimi Ana Bilim Dalı
PROF. DR. DİNÇAY KÖKSAL