Manual Code Editing and Management
After completing AI text analysis and creating a coded column, you can manually review, add, edit, and delete codes on individual text records. This allows you to refine the AI's work and ensure accuracy in your analysis.
Accessing the Manual Editing Interface
From Chart View
- After AI coding is complete, you'll see your coded column in chart view
- Click "View text and codes" button on the left side of the screen
- This switches you to table view where you can see both the original text and applied codes
Understanding the Table View
In table view, you can see:
- Original text column: The source text data
- Coded column: Shows which codes have been applied
- Highlighted text: Color-coded sections showing exactly where each code was applied
- Code tags: Visual indicators showing which codes are assigned to each text segment
Adding Codes to Text
Adding Codes to Specific Text Segments
- Select text: Click and drag to highlight the relevant portion of text in any record
- Choose action: A popup will appear with options to:
- Apply existing code: Choose from your existing codes
- Create new code: Enter a new code name
- Confirm: Click "Add Code" to apply your selection
Tips for Adding Codes
- Select the most relevant text snippet that represents the code
- You can apply multiple codes to overlapping or different parts of the same text record
- New codes you create will be available for the entire column, not just the current record
Removing Codes
Removing Codes from Individual Records
- Find the code tag: Locate the code you want to remove in the table view
- Hover over the code: A trash icon will appear
- Click the trash icon: The code will be removed from that specific record only
Managing Codes Globally
To remove or merge codes across your entire dataset:
- Return to chart view by clicking the chart icon
- Use the legend management tools to:
- Delete codes from all records
- Merge similar codes together
- Rename codes for clarity
Applying New Codes with AI
If you've manually added new codes and want AI to apply them to the rest of your dataset:
- Return to chart view from the table view
- Click "Use AI to apply new codes"
- Select codes to apply: Choose which of your new manually-created codes should be applied
- Let AI work: AddMaple will analyze your data and apply the new codes where relevant
Important: This process only adds new codes where relevant - your existing manual coding work won't be changed.
Best Practices for Manual Coding
Quality Control
- Review AI suggestions: Check that automatically applied codes make sense in context
- Be consistent: Apply similar codes to similar concepts across your dataset
- Use specific text selection: Highlight the exact words or phrases that represent each code
Workflow Efficiency
- Start with AI: Let AI do the bulk of the coding work first
- Focus on edge cases: Manually review records where AI might have struggled
- Iterate gradually: Make small corrections rather than wholesale changes
- Use AI re-application: When you add new codes, let AI help apply them broadly
Code Management
- Keep codes focused: Each code should represent a distinct concept or theme
- Use clear names: Make code names descriptive and unambiguous
- Merge similar codes: Combine codes that represent the same concept
- Document your process: Keep notes about your coding decisions for consistency
Troubleshooting
If you can't see the manual editing interface:
- Make sure you're viewing a coded column (created through AI analysis)
- Click "View text and codes" from the chart view
- Ensure you have the necessary permissions to edit the data
If codes aren't applying correctly:
- Check that you've selected the right text segment
- Verify the code name doesn't conflict with existing codes
- Try refreshing the page if the interface becomes unresponsive
If AI re-application isn't working:
- Ensure you've created at least one new code manually
- Check that you have sufficient credits for AI processing
- Return to chart view before attempting to use AI features
By combining AI efficiency with manual precision, you can create highly accurate and nuanced text analysis that meets your specific research needs.