Anonymization and Privacy

• What the Find and Replace Function Does (and Why It’s Powerful for Anonymizing Data)

  Find and Replace is a built-in tool in almost every program—Word, Adobe Acrobat Pro (in limited form), Excel, Google Docs, Google Sheets—that lets you automatically search for text and swap it out with something else.

  Instead of manually scanning a document for sensitive information (names, IDs, emails), Find and Replace allows you to locate everything instantly and replace it all in one action.

  It is one of the fastest ways to “sweep” a dataset before using it in ChatGPT.

  What “Find” Does

  The Find part lets you search for:

  • A name

  • A phrase

  • A student ID

  • An email domain (e.g., @schools.nyc.gov)

  • A word you want to hide

  • Any pattern that repeats in the file

  When you press Ctrl+F:

  • A small search window appears.

  • You type a word or number.

  • Every time you click Find Next, the program jumps to the next place that text appears.

  This is how you quickly identify where sensitive data lives without reading the whole document.

  What “Replace” Does

  Replace allows you to:

  • Swap all occurrences of a word with something else

  • Remove a word entirely

  • Replace names with pseudonyms

  • Replace IDs with placeholders

  • Strip email addresses or phone numbers in seconds

  Example:

  • Find: Martinez

  • Replace with: Student A

  Or:

  • Find: @schools.nyc.gov

  • Replace with: (leave blank) → This removes all email addresses.

  Where to Use Find and Replace

  1. In Microsoft Word

  Best for:

  • Long transcripts

  • Narrative comments

  • IEP text

  • Observation notes

  Steps:

  1. Press Ctrl+H (the shortcut for Replace).

  2. Enter the text you want to find.

  3. Enter the replacement text (or leave empty to delete it).

  4. Click Replace All.

  2. In Excel

  Best for:

  • Rosters

  • Grade spreadsheets

  • Exported PDF tables

  • Columns with repeating sensitive information

  Steps:

  1. Press Ctrl+F to find data.

  2. Press Ctrl+H to open Replace.

  3. Replace all instances of a name, ID, or pattern.

  Excel Tip:
  You can limit Find and Replace to only one column by selecting that column first.

  3. In Adobe Acrobat Pro (limited but useful)

  Adobe Pro doesn’t have full Replace for scanned PDFs, but:

  • You CAN find all occurrences of a term.

  • You CAN perform a redaction that removes all occurrences of a word.

  Redaction Tool (Powerful for anonymizing PDFs):

  1. Tools → Redact

  2. Mark for Redaction → Find Text → “Search and Redact”

  3. Enter your term (e.g., a name).

  4. Apply redactions → removes every occurrence permanently.

  Why Find and Replace Speeds Up Categorization

  Before exporting or cleaning data, Find and Replace helps you:

  A. Identify patterns in the data

  • Where are all the IDs?

  • Are names always in the same format?

  • Do emails appear in a consistent column?

  B. Categorize each column or field

  Find ID → you know which column stores student IDs.
  Find @schools.nyc.gov → you know which column is emails.
  Find a student name → you know where names are stored.

  C. Remove sensitive info before even running macros

  You can:

  • Replace all names with pseudonyms

  • Replace all IDs with a placeholder

  • Delete all emails

  • Remove phone numbers or DOBs

  This creates a clean, safe dataset before using ChatGPT.

How to Convert a PDF Roster into Pseudonyms (Step-by-Step)

Using Adobe Acrobat Pro → Excel → Macro-Driven Anonymization

This workflow takes you from a raw PDF (IEP list, roster, transcript, benchmark report) all the way to a clean Excel file where every student is replaced with a safe pseudonym like JM-456.

Everything below is teacher-friendly and repeatable.

Step 1 — Open Your PDF in Adobe Acrobat Pro

1. Open Adobe Acrobat Pro.

2. Go to File → Open and load your student roster/report PDF.

3. Confirm that the PDF is text-based (not just an image).

   • Try highlighting text.

   • If it highlights, you’re good.

   • If not, run Scan & OCR → Recognize Text.

Step 2 — Export the PDF to Excel

Acrobat’s built-in converter preserves tables fairly well.

1. Go to File → Export To → Spreadsheet → Microsoft Excel Workbook.

2. Choose a location and save the file (e.g., Roster_Export.xlsx).

3. Open the Excel file.

Now your PDF table is in Excel, ready for cleaning and pseudonymizing.

Step 3 — Clean the Excel File for Macro Use

PDF exports often create messy table structures. Do these quick fixes:

A. Unmerge All Cells

Merged cells break macros. Unmerge everything:

1. Select the entire sheet:

   • Click the square at the top-left corner (above row numbers, left of column A).

2. Go to Home → Merge & Center → Unmerge Cells.

B. Auto-Fix Column Widths

This helps you see where names, IDs, and comments are.

1. Select all columns (same top-left square).

2. Double-click the right edge of any column header.

   • Excel auto-resizes everything so it’s readable.

C. Identify the Column with Names + IDs

This is where Ctrl+F helps:

1. Press Ctrl + F.

2. Search for “ID” → see which column lights up (e.g., Column F or G).

3. Search a known student last name to confirm.

This determines which macro version you’ll use:

• Column F version

• Column G version

Step 4 — Insert and Run the Anonymization Macro

This macro transforms each student in that column into a pseudonym like AB-123 (initials + last 3 digits).

A. Open the VBA Editor

1. Press Alt + F11 (Windows) or Option + F11 (Mac with Excel desktop).

2. In the left Project pane, right-click on the workbook.

3. Choose Insert → Module.

A blank code window appears.

B. Paste the Correct Macro

If your sensitive name/ID block is in Column G, paste the Column G macro.
If it’s in Column F, paste the Column F macro.

(You already have both versions—use only one at a time.)

C. Close the VBA Window and Save as Macro-Enabled Workbook

1. Close VBA.

2. Save as:
   .xlsm (Excel Macro-Enabled Workbook)

Excel will warn you about macros—that’s expected.

D. Run the Macro

1. Go to Developer → Macros

   • If you don’t see Developer, enable it in:
     File → Options → Customize Ribbon → check “Developer”

2. Select Anonymize_ReplaceNameAndRemoveID.

3. Click Run.

The macro will:

• Scan each cell in the designated column.

• Normalize messy PDF line breaks.

• Pull the first + last initials.

• Collect all digits in the cell.

• Take the last 3 digits as the pseudonym ID.

• Replace the entire block with something like:

JM-482

You will then get a popup:

“Done! Column F/G cells replaced with INITIALS-###.”

Step 5 — Verify Your Data Is Fully Anonymized

Use Ctrl+F again:

1. Search for a specific name → should be gone.

2. Search for ID → should show minimal or no results.

3. Randomly inspect rows to confirm the macro replaced all blocks.

Step 6 — Prepare the Data for ChatGPT

At this stage:

• All real names → gone

• All real IDs → gone

• No email addresses

• No DOB

• No parent/guardian info

Safe to copy just the pseudonym column + descriptive/academic columns into ChatGPT.

The safest structure is:

Pseudonym | Class | Behavior/Notes | Skills | Performance | Comments

Optional: Quick Sanitizing with Find & Replace BEFORE the Macro

If you want extra safety:

• Press Ctrl + H (Find & Replace)

• Replace @schools.nyc.gov with blank

• Replace phone numbers or DOB patterns

• Replace any teacher last names you want anonymized

Then run the macro.

Teacher-Friendly Summary

1. Open PDF in Acrobat Pro
2. Export to Excel
3. Unmerge cells & Auto-width
4. Use Ctrl+F to find the name/ID column
5. Paste the matching macro (F or G)
6. Run the macro → get pseudonyms
7. Double-check with Ctrl+F
8. Copy only pseudonym + performance fields into ChatGPT

1. Turn Off “Chat History & Training”

• Go to Settings → Data Controls → Chat History & Training.

• Toggle it off.

• This prevents your conversations from being used to improve future models and keeps your inputs out of training datasets.

• You can still use all features, but chats will not be saved in your history.

2. Use Temporary Chat (If Available)

• Temporary chats automatically disable history and training.

• Nothing from the temporary chat is saved to your account once you close it.

3. Avoid Uploading Entire Raw Documents

Even with protections on, you should avoid:

• Uploading full IEPs

• Full psychological reports

• Files containing student names, ID numbers, parent information, phone numbers, addresses, or medical info
  Instead, upload excerpts or manually paste only the sections you need, after removing identifiers (details below).

4. Verify That You’re Using a Privacy-Enhanced Tier

If possible, use:

• ChatGPT Team

• ChatGPT Enterprise
  These versions offer:

• Zero training on your data by default

• Encryption in transit and at rest

• Admin-level data retention controls

• SOC2 compliance
  For most K–12 use, Team or Pro with history disabled is sufficient, but schools or districts seeking institutional compliance should choose Team/Enterprise.

5. Use Browser Protections

• Turn off browser extensions that read webpage content.

• Use a private browser window.

• Avoid shared or public computers.