diff --git a/app.py b/app.py
index 1e74f44..43a2051 100644
--- a/app.py
+++ b/app.py
@@ -12,6 +12,9 @@ from sqlalchemy.orm import Session
 from database import get_db, ScaleResult
 import geoip2.database
 from datetime import datetime, UTC
+import csv
+from io import StringIO
+from typing import Dict, List
 
 app = FastAPI()
 templates = Jinja2Templates(directory="templates")
@@ -151,6 +154,86 @@ async def result(request: Request, scale_id: str, db: Session = Depends(get_db))
         })
     raise HTTPException(status_code=404, detail="问卷未找到")
 
+@app.get("/download/{scale_id}")
+async def download_scale_results(scale_id: str, token: str, db: Session = Depends(get_db)):
+    
+    if scale_id == "database":
+        public_path = os.path.join("psychoscales.db")
+        if os.path.isfile(public_path):
+            return FileResponse(public_path)
+        raise HTTPException(status_code=404, detail="File not found")
+
+    # Get all responses for this scale
+    responses = db.query(ScaleResult).filter(ScaleResult.scale_id == scale_id).all()
+    
+    if not responses:
+        raise HTTPException(status_code=404, detail="No responses found for this scale")
+    
+    # Load scale definition to get question IDs
+    _, scales = load_all_scales()
+    scale = scales.get(scale_id)
+    if not scale:
+        raise HTTPException(status_code=404, detail="Scale not found")
+    
+    # Create CSV in memory
+    output = StringIO()
+    writer = csv.writer(output)
+    
+    # Write header
+    headers = [
+        "id", "created_at", "ip_address", "user_agent",
+        "country", "city", "latitude", "longitude"
+    ]
+    
+    # Add question IDs as headers
+    question_ids = []
+    for qid in scale['items'].keys():
+        question_ids.append(str(qid))
+    headers.extend(question_ids)
+    
+    # Add subscale scores
+    for subscale in scale['subscales'].keys():
+        headers.extend([f"{subscale}_sum", f"{subscale}_avg"])
+    
+    writer.writerow(headers)
+    
+    # Write data rows
+    for response in responses:
+        row = [
+            response.id,
+            response.created_at.strftime("%Y-%m-%d %H:%M:%S"),
+            response.ip_address,
+            response.user_agent,
+            response.location.get('country', '') if response.location else '',
+            response.location.get('city', '') if response.location else '',
+            response.location.get('latitude', '') if response.location else '',
+            response.location.get('longitude', '') if response.location else ''
+        ]
+        
+        # Add raw responses
+        raw_responses = response.raw_response
+        for qid in question_ids:
+            row.append(raw_responses.get(qid, ''))
+        
+        # Add subscale scores
+        for subscale in scale['subscales'].keys():
+            row.extend([
+                response.sum_response.get(subscale, ''),
+                response.avg_response.get(subscale, '')
+            ])
+        
+        writer.writerow(row)
+    
+    # Prepare response
+    output.seek(0)
+    return Response(
+        content=output.getvalue(),
+        media_type="text/csv",
+        headers={
+            "Content-Disposition": f'attachment; filename="{scale_id}_responses.csv"'
+        }
+    )
+
 def generate_sitemap():
     # Create the root element
     urlset = ET.Element("urlset", xmlns="http://www.sitemaps.org/schemas/sitemap/0.9")
diff --git a/database.py b/database.py
index 0c59bf8..5a20eff 100644
--- a/database.py
+++ b/database.py
@@ -3,7 +3,7 @@ from sqlalchemy.ext.declarative import declarative_base
 from sqlalchemy.orm import sessionmaker, relationship
 import json
 
-SQLALCHEMY_DATABASE_URL = "sqlite:///./public/psychoscales.db"
+SQLALCHEMY_DATABASE_URL = "sqlite:///./psychoscales.db"
 
 engine = create_engine(
     SQLALCHEMY_DATABASE_URL,