# 🧠 Psikotest Engine β€” Panduan Teknis (Production Guide) ## 1. 🎯 Tujuan Sistem Sistem psikotest bertujuan untuk: - Mengukur **kemampuan kognitif (IQ)** - Mengukur **kepribadian (personality)** - Mengukur **ketahanan & konsistensi kerja** Output utama: - Skor numerik - Profil kepribadian - Interpretasi HR --- # 2. 🧱 Arsitektur Data ## Core Tables ``` test_types questions question_items test_sessions candidate_answers scoring_rules candidate_results ``` --- # 3. πŸ”„ FLOW END-TO-END ## Step 1 β€” Setup Test Admin: - Buat `test_types` - Input `questions` - Input `question_items` - Define `scoring_rules` --- ## Step 2 β€” User Start Test Create: ``` test_sessions ``` Contoh: ```json { "candidate_id": "uuid", "test_type_id": 1, "started_at": "timestamp" } ``` --- ## Step 3 β€” Render Soal Backend: - Ambil `questions` by `test_type_id` - Ambil `question_items` Frontend: - Render berdasarkan `question_type` --- ## Step 4 β€” Submit Jawaban Masuk ke: ``` candidate_answers ``` Contoh: ```json { "question_id": "...", "answer": { "selected": "B" } } ``` --- ## Step 5 β€” Scoring Engine (πŸ”₯ CORE) Backend: - Ambil `candidate_answers` - Ambil `scoring_rules` - Hitung skor --- ## Step 6 β€” Simpan Hasil Masuk ke: ``` candidate_results ``` --- # 4. 🧠 SCORING ENGINE DETAIL --- # 🧩 A. CFIT / IQ TEST ## Tujuan Mengukur kemampuan logika non-verbal --- ## Data - Jawaban benar / salah - Total soal --- ## Rumus ### 1. Raw Score ``` raw_score = jumlah_jawaban_benar ``` --- ### 2. Accuracy ``` accuracy = correct / total_questions ``` --- ### 3. IQ Mapping (Simplified) Contoh: ```json { "min": 0, "max": 20, "iq": 80 } ``` --- ### Implementasi ``` for range mapping: if raw_score in range: iq = mapping.iq ``` --- ## Output ```json { "correct": 42, "wrong": 8, "iq": 110 } ``` --- # 🧠 B. DISC (PERSONALITY) ## Tujuan Menentukan dominasi kepribadian: - D (Dominance) - I (Influence) - S (Steadiness) - C (Compliance) --- ## Input Setiap soal: - Most - Least --- ## Rumus ### 1. Inisialisasi ``` D = 0 I = 0 S = 0 C = 0 ``` --- ### 2. Perhitungan ``` for setiap jawaban: MOST β†’ +1 LEAST β†’ -1 ``` --- ### 3. Dominan ``` dominant = max(D, I, S, C) ``` --- ## Output ```json { "D": 12, "I": 8, "S": 15, "C": 5, "dominant": "S" } ``` --- # 🧠 C. MBTI ## Dimensi: - E / I - S / N - T / F - J / P --- ## Rumus ``` for each answer: add +1 to selected trait ``` --- ## Result ``` EI β†’ E or I SN β†’ S or N TF β†’ T or F JP β†’ J or P ``` --- ## Output ``` INTJ ``` --- # πŸ”’ D. KRAEPELIN ## Tujuan - Kecepatan - Ketelitian - Konsistensi --- ## Data - jumlah jawaban - jumlah benar - pola per baris --- ## Rumus ### 1. Total ``` total = jumlah_jawaban ``` --- ### 2. Accuracy ``` accuracy = benar / total ``` --- ### 3. Konsistensi ``` std_dev per row ``` --- ### Interpretasi - stabil β†’ bagus - fluktuatif β†’ kurang stabil --- ## Output ```json { "total": 120, "accuracy": 0.92, "consistency": "high" } ``` --- # 5. βš™οΈ SCORING_RULES STRUCTURE Contoh: ```json { "method": "correct_count", "score_per_correct": 1 } ``` --- ## Jenis method: | method | kegunaan | | ------------- | -------- | | correct_count | CFIT | | trait_count | DISC | | sum | Likert | | mapping | IQ | --- # 6. 🧠 INTERPRETATION ENGINE Contoh: ### IQ ``` >120 β†’ sangat tinggi 100-120 β†’ normal <100 β†’ rendah ``` --- ### DISC ``` D tinggi β†’ leader S tinggi β†’ stabil ``` --- # 7. ⚠️ BEST PRACTICE ## ❗ Jangan expose answer_key ## ❗ scoring di backend ## ❗ gunakan JSON untuk fleksibilitas ## ❗ versioning scoring_rules --- # 8. πŸš€ FLOW SUMMARY ``` User β†’ Answer β†’ candidate_answers ↓ scoring_engine ↓ candidate_results ↓ HR Dashboard ``` --- # 9. πŸ”₯ VALUE BISNIS Kalau sistem kamu punya: - scoring engine - interpretation - analytics πŸ‘‰ kamu bukan jual soal πŸ‘‰ kamu jual **decision system HR** --- # 10. πŸ“ˆ NEXT IMPROVEMENT - percentile ranking - benchmarking kandidat - AI recommendation --- # 🧠 FINAL INSIGHT > β€œSoal itu murah. Interpretasi itu mahal.” ---