| Ordinal Alpha | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| raw_alpha | std.alpha | G6(smc) | average_r | SN | median_r | ||||||
| 0.794 | 0.794 | 0.873 | 0.099 | 3.849 | 0.109 | ||||||
| Ordinal Omega | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| omega_h | omega.lim | alpha | omega.tot | G6 | |||||
| 0.600 | 0.638 | 0.921 | 0.939 | 0.942 | |||||
^ Berikut ini adalah analisis reliabilitas MHI dengan mengasumsikan bahwa aitem-aitem dalam MHI merupakan data ordinal. Ada dua analisis yang dilakukan, yakni Cronbach's α dan McDonald's ω. Yang terakhir lebih cocok digunakan karena alat ukur melanggar asumsi τ equivalence, yaitu masing-masing aitem berkontribusi tidak setara dalam menjelaskan konstruk.
Cara melaporkan reliabilitas:
Dengan menggunakan teknik analisis reliabilitas McDonald's ω, reliabilitas Mental Health Inventory secara umum, diketahui cukup memuaskan (ωh=0,60).
Procedure Notes
The TestROC optimal cutpoint analysis has been completed using the following specifications:
Measure Variable(s): MentalHealthIndex
Class Variable: srq_diagnostik
Positive Class: 1
Method: maximize_metric
All Observed Cutpoints: FALSE
Metric: sum_sens_spec
Direction (relative to cutpoint): >=
Tie Breakers: c
Metric Tolerance: 0.05
For more information on how calculations are performed and interpretting results, please see the documentation
| Scale: MentalHealthIndex | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Cutpoint | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | Youden's index | AUC | Metric Score | ||||||||
| 121 | 72.96% | 51.18% | 34.3% | 84.42% | 0.241 | 0.691 | 1.241 | ||||||||
| 122 | 70.35% | 55.42% | 35.53% | 84.26% | 0.258 | 0.691 | 1.258 | ||||||||
| 123 | 68.2% | 58.48% | 36.45% | 84.04% | 0.267 | 0.691 | 1.267 | ||||||||
| 124 | 64.67% | 61.91% | 37.22% | 83.38% | 0.266 | 0.691 | 1.266 | ||||||||
| 125 | 62.06% | 65.13% | 38.33% | 83.09% | 0.272 | 0.691 | 1.272 | ||||||||
| 126 | 58.83% | 68.24% | 39.28% | 82.6% | 0.271 | 0.691 | 1.271 | ||||||||
| 127 | 57.14% | 71.62% | 41.29% | 82.71% | 0.288 | 0.691 | 1.288 | ||||||||
| 128 | 53.3% | 73.87% | 41.61% | 81.92% | 0.272 | 0.691 | 1.272 | ||||||||
| 129 | 50.08% | 76.56% | 42.73% | 81.45% | 0.266 | 0.691 | 1.266 | ||||||||
| 130 | 46.85% | 79.56% | 44.46% | 81.08% | 0.264 | 0.691 | 1.264 | ||||||||
| 131 | 43.47% | 81.71% | 45.35% | 80.54% | 0.252 | 0.691 | 1.252 | ||||||||
[3] [4]
^ Dari tabel di atas, cut-off score dengan Youden's index yang paling optimal (0.288) adalah 127. Artinya, responden yang skornya di bawah 127 dapat didiagnosis menderita gangguan mental-emosional, sedangkan >= 128 dianggap memiliki kondisi mental yang optimal.
Dengan nilai cut-off tersebut diatas, sensitivity diketahui sebesar 57.14%, sehingga 6 dari 10 pasien dengan gangguan mental emosional akan mendapatkan diagnosis yang tepat atas kondisinya. Selain itu, specificity diketahui sebesar 71.62%, yaitu 7 dari 10 pasien tanpa gangguan mental emosional akan mendapati diagnosis negatif dari MHI. Dapat disimpulkan bahwa MHI dapat merule out pasien tanpa gangguan mental emosional secara lebih baik daripada menemukan kasus positif.
Dari indikator positive predictive value (PPV), diketahui bahwa 4 dari 10 (41.29%) pasien yang menerima diagnosis positif (menderita gangguan emosional), benar-benar memiliki gangguan tersebut. Sedangkan 8 dari 10 pasien yang didiagnosis negatif, memang benar-benar tidak memiliki gangguan mental emosional.
Diketahui pula Area under Curve (AUC) sebesar 0,691, artinya MHI dapat mengklasifikasikan pasien dengan akurasi 69.1 persen.
^ Berikut di atas adalah visualisasi dari kurva ROC.
| Scale: MentalHealthIndex | Score: 131 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 1523 (TN) | 341 (FP) | |
| Positive | 368 (FN) | 283 (TP) | |
| Scale: MentalHealthIndex | Score: 130 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 1483 (TN) | 381 (FP) | |
| Positive | 346 (FN) | 305 (TP) | |
| Scale: MentalHealthIndex | Score: 129 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 1427 (TN) | 437 (FP) | |
| Positive | 325 (FN) | 326 (TP) | |
| Scale: MentalHealthIndex | Score: 128 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 1377 (TN) | 487 (FP) | |
| Positive | 304 (FN) | 347 (TP) | |
| Scale: MentalHealthIndex | Score: 127 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 1335 (TN) | 529 (FP) | |
| Positive | 279 (FN) | 372 (TP) | |
| Scale: MentalHealthIndex | Score: 126 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 1272 (TN) | 592 (FP) | |
| Positive | 268 (FN) | 383 (TP) | |
| Scale: MentalHealthIndex | Score: 125 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 1214 (TN) | 650 (FP) | |
| Positive | 247 (FN) | 404 (TP) | |
| Scale: MentalHealthIndex | Score: 124 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 1154 (TN) | 710 (FP) | |
| Positive | 230 (FN) | 421 (TP) | |
| Scale: MentalHealthIndex | Score: 123 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 1090 (TN) | 774 (FP) | |
| Positive | 207 (FN) | 444 (TP) | |
| Scale: MentalHealthIndex | Score: 122 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 1033 (TN) | 831 (FP) | |
| Positive | 193 (FN) | 458 (TP) | |
| Scale: MentalHealthIndex | Score: 121 | |||
|---|---|---|---|
| DECISION BASED ON MEASURE | |||
| CRITERION | Negative | Positive | |
| Negative | 954 (TN) | 910 (FP) | |
| Positive | 176 (FN) | 475 (TP) | |
^ Dari tabel di atas dapat diketahui bahwa dengan menerapkan cut off point pada 127, maka klasifikasi TN, TP, FN, dan FP mencapai jumlah yang paling optimal.
| Ordinal Alpha | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| raw_alpha | std.alpha | G6(smc) | average_r | SN | median_r | ||||||
| 0.794 | 0.794 | 0.873 | 0.099 | 3.849 | 0.109 | ||||||
| Ordinal Omega | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| omega_h | omega.lim | alpha | omega.tot | G6 | |||||
| 0.600 | 0.638 | 0.921 | 0.939 | 0.942 | |||||
[1] The jamovi project (2020). jamovi. (Version 1.6) [Computer Software]. Retrieved from https://www.jamovi.org.
[2] R Core Team (2020). R: A Language and environment for statistical computing. (Version 4.0) [Computer software]. Retrieved from https://cran.r-project.org. (R packages retrieved from MRAN snapshot 2020-08-24).
[3] Thiele, C. (2019). cutpointr: Determine and Evaluate Optimal Cutpoints in Binary Classification Tasks. [R package]. Retrieved from https://cran.r-project.org/package=cutpointr.
[4] Friesen, L., Kroc, E., Zumbo, B. D. (2019). Psychometrics & Post-Data Analysis: Test ROC. [jamovi module]. Retrieved from https://github.com/lucasjfriesen/jamoviPsychoPDA.