Bu çalışmanın temel amacı, Ergenler İçin Sosyal Medya Bağımlılığı Ölçeği'nin cinsiyet ve okul kademesi açısından ölçme değişmezliğine sahip olup olmadığını incelemektir. Çalışma betimsel modeldedir. Çalışmada örneklem, Niğde Ölçme Değerlendirme Merkezi tarafından yürütülen eğitim araştırmaları kapsamında yer alan 939 ortaokul ve lise öğrencisinden oluşmaktadır. Örneklem iki düzeyli tabakalama yöntemi kullanılarak oluşturulmuştur. Ölçme değişmezliği analizi için Çok Gruplu Doğrulayıcı Faktör Analizi (ÇG-DFA) yaklaşımı tercih edilmiştir Analiz öncesinde kayıp veri analizi, uç değer analizi, örneklem büyüklüğü, çok değişkenli uç değer analizi, normallik için çarpıklık ve basıklık değerlerinin incelenmesi, doğrusallık incelemesi, eşvaryanslılık incelemesi ve çoklu bağlantı durum incelemesi gerçekleştirilmiştir. Ön inceleme doğrultusunda WLSMV tahminleyicisi kullanarak analizler gerçekleştirilmiştir. Analizde; doğrulayıcı faktör analizi ayrı ayrı incelenmiş, özdeş faktör yapısının eş zamanlı testi uygulanmış, faktör yüklerinin eşitliği test edilmiş, gösterge sabitlerinin eşitliği test edilmiş ve gösterge artık varyanslarının eşitliği test edilmiştir. Gerçekleştirilen analizler sonucunda ölçeğin hem cinsiyet hem okul kademesi açısından skaler düzeyde ölçme değişmezliğine sahip olduğu bulgusuna ulaşılmıştır. Elde edilen bu bulgular ile cinsiyet ve okul kademesi açısından gruplar arasında örtük veya gözlenen ortalamalar karşılaştırılabileceği ve farklı örneklemlerden toplanan verilerin daha güvenli bir şekilde kullanılabileceği ifade edilebilir. Bu çalışma, Niğde örnekleminde gerçekleştirilmiştir. Araştırmada kullanılan ölçeğin ölçme değişmezliği çalışmalarının Türkiye’nin çeşitli bölgelerinde ve farklı gelişmişlik düzeylerindeki illerde tekrarlanmasının sonuçları farklılaştırabileceği düşünülmektedir. Bundan sonraki çalışmalarda farklı alt gruplar açısından ölçme değişmezliği araştırmaları yürütülebilir. Bu çalışmada ÇG-DFA kapsamında ölçme değişmezliği incelenmiş olup önümüzdeki çalışmalarda farklı teoriler ve teknikler ile ölçme değişmezliği incelemesi gerçekleştirilebilir.
The general purpose of this study is to examine whether the Social Media Addiction Scale for Adolescents has measurement invariance in terms of gender and grade level. The study is a descriptive model. The sample in the study consists of 939 middle and high school students within the scope of educational research conducted by Niğde Measurement and Evaluation Center. The sample was formed using the two-level stratification method. Multigroup Confirmatory Factor Analysis approach was preferred for measurement invariance analysis. Prior to the analysis, missing data analysis, outlier analysis, sample size, multivariate outlier analysis, examination of skewness and kurtosis values for normality, linearity examination, equivariance examination and multicollinearity examination were performed. In line with the preliminary examination, the analyses were conducted using the WLSMV estimator. In the analysis; confirmatory factor analysis was examined separately, simultaneous testing of identical factor structure was applied, equality of factor loadings was tested, equality of indicator constants was tested and equality of indicator residual variances was tested. As a result of the analysis, it was found that the scale had measurement invariance at the scalar level in terms of both gender and grade level. With these findings, it can be stated that latent or observed averages can be compared between groups in terms of gender and school level, and data collected from different samples can be used more reliably. This study was conducted in Niğde sample. New studies can be conducted in different samples. In future studies, measurement invariance studies can be conducted in terms of different subgroups. In this study, measurement invariance was examined within the scope of Multigroup Confirmatory Factor Analysis, and measurement invariance can be examined with different theories and techniques in future studies.
Primary Language | Turkish |
---|---|
Subjects | Studies on Education |
Journal Section | Educational Sciences and Sciences of Field Education |
Authors | |
Publication Date | June 30, 2023 |
Published in Issue | Year 2023 Volume: 14 Issue: 3 |
This journal uses a CC BY-NC-SA license.
editor@e-ijer.com http://www.e-ijer.com Address: Ege University Faculty of Education İzmir/Türkiye