The Analyses were conducted via LIS

The Analyses were conducted via LISREL8.8 structural equation modeling computer program (Jo¨ reskog and So¨ rbom, 1996, 1998). The following indexes were used as measures of goodness-of-fit: First the comparative fit index (CFI) was used as a focal index, since it has advantageous statistical properties: it has standardized range, small sample variability, and stability with various sample sizes (Jo¨ reskog and So¨rbom, 1981; Bentler 1990). A value of CFI greater than 0.95 indicates an adequate model fit (Hu and Bentler, 1999). In addition, the goodness-of-fit w2, the Standardized Root Mean-square Residual (SRMR), the Root Mean-Square Error of Approximation (RMSEA), the Non- Normed Fit Index (NNFI) and the Adjusted Goodness of Fit Index (AGFI), were also used. Note that the goodness-of-fit w2 indicates the difference between the observed and implied by the proposed theoretical model variance-covariance matrices. Thus, the non-significant values of w2 are desired indicating that the proposed theoretical model significantly reproduced the sample variance-covariance relationships in the matrix (Schumacker and Lomax, 2010).
A confirmatory factor analysis. The confirmatory factor model was used to demonstrate the existence of two latent variables, the Particulate and the Collective, which have been proposed as the dimensions of structure understanding (Johnson, 1998a; Tsitsipis et al., 2010).
Fig. 1 shows the confirmatory factor model of structure understanding. The observable variables load the two latent variables, the Particulate and Collective. Three of the questions (S1, S2, and S3) load the Particulate and the other two (S4 and S5) load the Collective.
The value of CFI is 0.999; the goodness-of-fit w2 = 3.65, df = 4, p = 0.46; the Standardized Root Mean-square Residual SRMR is 0.016; the Root Mean-Square Error of Approximation RMSEA is 0.0; the Non-Normed Fit Index NNFI is 1.0 and the Adjusted Goodness of Fit Index AGFI is 0.983. They indicate an adequate model fit.
A multiple-indicator multiple-cause (MIMIC) model. The Structural equation modeling involved the twelve observed variables (Table 1) and three latent variables. Particulate and Collective were measured as indicated by the above factor model and the latent variable understanding changes of state (UnChSt) measured by S6, S7, S8 and S9 item. The model is a multiple-indicator multiple-cause model (MIMIC), where latent

Hasil (Bahasa Indonesia) 1:

The Analyses were conducted via LISREL8.8 structural equation modeling computer program (Jo¨ reskog and So¨ rbom, 1996, 1998). The following indexes were used as measures of goodness-of-fit: First the comparative fit index (CFI) was used as a focal index, since it has advantageous statistical properties: it has standardized range, small sample variability, and stability with various sample sizes (Jo¨ reskog and So¨rbom, 1981; Bentler 1990). A value of CFI greater than 0.95 indicates an adequate model fit (Hu and Bentler, 1999). In addition, the goodness-of-fit w2, the Standardized Root Mean-square Residual (SRMR), the Root Mean-Square Error of Approximation (RMSEA), the Non- Normed Fit Index (NNFI) and the Adjusted Goodness of Fit Index (AGFI), were also used. Note that the goodness-of-fit w2 indicates the difference between the observed and implied by the proposed theoretical model variance-covariance matrices. Thus, the non-significant values of w2 are desired indicating that the proposed theoretical model significantly reproduced the sample variance-covariance relationships in the matrix (Schumacker and Lomax, 2010).A confirmatory factor analysis. The confirmatory factor model was used to demonstrate the existence of two latent variables, the Particulate and the Collective, which have been proposed as the dimensions of structure understanding (Johnson, 1998a; Tsitsipis et al., 2010).Gambar 1 menunjukkan model konfirmasi faktor pemahaman struktur. Variabel yang diamati memuat dua variabel laten, partikulat dan kolektif. Tiga pertanyaan (S1, S2 dan S3) beban partikulat dan lain dua (S4 dan S5) beban kolektif.Nilai CFI adalah 0.999; kebaikan dari cocok w2 = 3,65, df = 4, p = 0,46; Standar Root Mean square sisa SRMR adalah 0.016; Root Mean Square kesalahan dari pendekatan RMSEA adalah 0.0; Normed bebas sesuai indeks NNFI 1.0 dan disesuaikan kebaikan dari sesuai indeks AGFI adalah 0.983. Mereka menunjukkan sesuai model yang memadai.Model multi-indikator multiple-penyebab (MENIRU). Model persamaan struktural melibatkan dua belas diamati variabel (Tabel 1) dan ketiga variabel laten. Partikulat dan kolektif diukur seperti yang ditunjukkan oleh faktor di atas model dan perubahan laten variabel pemahaman negara (UnChSt) diukur oleh S6, S7, S8 dan S9 item. Model adalah model multi-penyebab beberapa-indikator (MENIRU), di mana laten

Hasil (Bahasa Indonesia) 2:

The Analisis dilakukan melalui LISREL8.8 program pemodelan persamaan struktural komputer (JO reskog dan begitu rbom, 1996, 1998). Indeks berikut digunakan sebagai ukuran kebaikan-of-fit: Pertama indeks fit komparatif (CFI) digunakan sebagai indeks fokus, karena memiliki sifat statistik yang menguntungkan: ia telah dibakukan range, variabilitas sampel kecil, dan stabilitas dengan berbagai sampel ukuran (JO reskog dan So¨rbom, 1981; Bentler 1990). Sebuah nilai CFI lebih besar dari 0,95 menunjukkan model fit yang memadai (Hu dan Bentler, 1999). Selain itu, w2 kebaikan-of-fit, yang Standar Root Mean persegi Residual (SRMR), Error Root Mean-Square dari Approximation (RMSEA), yang Non bernorma Fit Index (NNFI) dan Goodness of Fit Index Disesuaikan (AGFI), juga digunakan. Perhatikan bahwa kebaikan-of-fit w2 menunjukkan perbedaan antara diamati dan tersirat oleh model teoritis matriks varians-kovarians yang diusulkan. Dengan demikian, nilai-nilai non-signifikan w2 yang diinginkan menunjukkan bahwa model teoritis yang diusulkan secara signifikan direproduksi sampel hubungan varians-kovarians dalam matriks (Schumacker dan Lomax, 2010).
Sebuah analisis faktor konfirmatori. Model faktor konfirmatori digunakan untuk menunjukkan adanya dua variabel laten, yang Partikulat dan Kolektif, yang telah diusulkan sebagai dimensi pemahaman struktur (Johnson, 1998a;. Tsitsipis et al, 2010).
Gambar. 1 menunjukkan model faktor konfirmatori dari pemahaman struktur. Variabel teramati memuat dua variabel laten, yang Partikulat dan Kolektif. Tiga dari pertanyaan (S1, S2, dan S3) memuat Particulate dan dua lainnya (S4 dan S5) memuat Kolektif.
Nilai CFI adalah 0,999; kebaikan-of-fit w2 = 3,65, df = 4, p = 0,46; yang SRMR Residual Standar Root Mean-square adalah 0,016; Kesalahan Root Mean-Square dari Aproksimasi RMSEA adalah 0,0; Indeks Fit Non-bernorma NNFI adalah 1.0 dan Goodness of Fit Index Disesuaikan AGFI adalah 0,983. Mereka menunjukkan model fit yang memadai.
Sebuah beberapa penyebab (MIMIC) Model multiple-indikator. Pemodelan persamaan struktural melibatkan dua belas variabel yang diamati (Tabel 1) dan tiga variabel laten. Partikulat dan Kolektif diukur seperti yang ditunjukkan oleh model faktor di atas dan laten perubahan variabel pemahaman negara (UnChSt) diukur dengan S6, S7, S8 dan S9 item. Model merupakan kelipatan-indikator beberapa penyebab Model (MIMIC), di mana laten