RESOURCES:
Using Technology in Organizing for Civic Participation in the Democratic Process

What’s “healthy data”?

• You can mail merge without fear. Addresses are complete and formatted correctly.
• There are no duplicate records.
• It's easy to search, sort or mass-change data. Fields are used consistently so you can find what you're looking for quickly and be sure you know what you have.
• Data can be interpreted by computer programs. Data is entered consistently.
• List-enhancement works!

Obstacles to healthy data

• You inherited the database from someone else who messed it up.
• Data is entered by nineteen different volunteers who haven't had any training.
• Don’t have time to deal with it because you’re busy trying to get out this mailing which is hard to do because the data is so mixed up.

Principles of healthy data

• Separate fields for different pieces of information:
  • Up to six fields for a name - title, First name, middle name, last name, suffix, nickname;
  • area code and phone number separate;
  • zip and zip4 separate;
  • organization, address, suite number, separate.
• Field names reflect data they contain.
• Information goes into the field for which it was intended:
• Data entry discipline and training - especially for volunteers.
• Data entered consistently.
• One list, not five:
• Use relational databases to avoid duplicating name records.
• Unique identifier for every record.

Common data illnesses:

Name cramming.

Two fields are not enough. If all you have are fields for the first name and last name, you're going to end up cramming something when you get a contribution from Dr. Aloyisius M. Jones, Jr. (and you know he uses the nickname Al). You'll want to address letters to "Dear Dr. Jones", or "Dear Al".

• Correct name storage involves using separate fields for different pieces of info.
• English names have six parts: Title, First Name, Middle Name, Last Name, Suffix, and Salutation.
• You can address letters to “Dear <<Salutation>>”.
• First Name and Last Name will be clear of clutter like Jr., etc. so that list matching, duplicate searches, etc., will work.

Address cramming.

Need more than Addr, City, State, ZIP;

Correct address storage:

• Separate Organization field - provides a consistent place to store the info, and doesn't mix it with names or addresses.
• Address is in 2 parts to satisfy post office format.
• Separate 5-digit ZIP from ZIP+4 so you can sort by zip code only for bulk mailing.
• Home vs. business info - so you can mail to the business address if requested, but have the residence address for the voter file match.

Various techniques to automate mailing to the preferred address.

• Manual check box.
• Date trigger for vacation or student addresses.

Cryptic Field naming.

Simple principle: your successor should be able to figure out what the field is without any explanation just by its name.

• Phone1, Phone2 doesn't say anything except that there are two phone numbers. HPhone, WPhone says which is which, and encourages correct use of the fields.
• Date could be anything. LastDonationDate, JoinDate says what's in the field.

Misuse of existing fields

• Organizations are not last names. Use a separate field for the organization name. If it's an individual record, you'll leave it blank. And if you want to find everyone for whom you have a business address, you can do it.
• New info, new field. If you create a new action alert newsletter, create a new field to record subscribers. If you use an existing field the name will not reflect the use, and the two uses will eventually, inevitably, become confused.

Inconsistent data entry.

Inconsistent entry is main source of duplicates (besides importing & never de-duping)

• St, St., or Street?
• Env Fdn or Environmental Foundation?
• Before entering a record, look for it first.
• Write a short “style guide” for your organization and keep a printout near the computer.

Too many lists

• Keeping separate lists for your members, board members, prospects, action alert rosters, etc., guarantees duplication because most of your constituents will appear on more than one list.
• Data operations must be repeated - someone changes address and you have to search all the lists to be sure all your communications go to the new address.
• Keep a single list and use fields or related files to distinguish entry types:
  • Newsletter (Y/N);
  • Member (Y/N);
  • Board (Y/N);
  • Elected official (Y/N);
  • etc.

Missing unique identifier

• Essential for matching your list to another list based on yours.
• Auto-generated (not user-entered).
• Usually easy to create:
  • Access: AutoNumber field;
  • FileMaker: Serial Number field;
  • ebase: Already there for you!
• Preserve when you migrate to a new db.

Final thoughts

Imagine how a computer would see your database ­ not just for printing labels. If you need to sort and search on data it needs to be in a separate field.

Garbage in, garbage out.

It’s great to add in new fields to help coordinate a campaign, but be sure to clean-up once the campaign’s over. 

Originally developed by Washington Environmental Alliance for Voter Education (WEAVE). 
Modifications and additions by Progressive Technology Project

11/12/2003