ABOUT THE DATABASE
WHAT IT IS AND HOW IT WAS BUILT
The database was built with much labor and eyestrain over
a four-month period from October 2000 to January 2001. Microsoft
Access is the database engine. Take a look at some of the
register page images (for example, this
one) and you'll see what I had to read, interpret, and
move through my fingertips to my computer and into Access!
The primary table is a flat file of 3,704 records, one each
for every landing recorded in the Register. Each record consists
of 45 fields. This yields a database with over 165,000 individual
data points. Some of the fields are blank (as when a pilot
did not enter a departure time or destination), but most are
filled. The database is being updated constantly as I discover
new information, or as more clear information enables me to
correct errors of spelling or content. You may review the most recent updates to the site by clicking "What's
New on the Site?" at the bottom of any page.
The database is quality controlled, in that I did edit pilot's
names or other information for uniformity and searchability.
For example, I added certain information that I knew. An example
would be if Ira Eaker signed in as "I. Eaker," I
made that entry in the database read "Ira C. Eaker".
Likewise, I cross checked handwriting so that if the handwriting
of "J. Smith" matched that of "John Smith"
I uniformly entered the most complete name so I didn't count
them as two different people. I also added city and state
information to locations.
I added some fields to make the data more tractable.
I added a key field, which assigns each record a sequential
number in the order that it appears chronologically in the
Register. This makes it easy to sort the database back to
its original chronological sequence. I added internal URLs
for pilots, places and airplane data, which enables intrasite
links to information I provide for those entities.
I added a "VIP?" field in order to flag pilots,
aircraft or passengers identified as famous for one reason
or another. I added a "MilorCiv?" field to identify
pilots as military or civilian. The "Civ" ranking
was further broken down to identify female pilots with a "Fem"
notation. There are 1,718 civilian records, including 58 landings
by 42 female pilots. There are 1,913 military records. I separated
passenger listings into fields with individual passenger names.
Passenger lists were studied as a separate entity, especially
helpful in analyzing air transport operations.
The Register contains a "Remarks" column, which
I included as a separate field in the database. Surprisingly,
very few pilots used that space to write anything. Where they
did, I enclosed their words in quotes when I transcribed them
to the database. I also used this field for my own remarks,
for example, if I wanted to add information about the pilot
or aircraft. Where I included my own remarks, I did not enclose
them in quotation marks.
POSSIBLE SOURCES OF ERROR
The Register is handwritten with nib pens or pencils, and
it is pushing 85 years old. As such, the information transcribed
to the database was subject to my interpretations of handwriting,
confounded by inkblots and smudges. As I read various books
and articles and cross-reference the information with my database,
I frequently find errors that I made during my original transcriptions.
These errors are corrected immediately. Again, check "What's
New on the Site?".
Although I made good effort to cross check multiple entries
by the same pilot to make sure I didn't confuse letters or
numbers, some records were still open to interpretation. Where
I couldn't decipher the text, I simply entered it as "Unreadable".
I was able, in September 2002, to visit with the original
Register maintained at the Operations Office at the Davis-Monthan
Air Force Base in Tucson. I reduced significantly the numbers
of "Unreadable" entries in my database by having
access to clearer, high-quality images of the original (photo,
Here are some other possible sources of error in the database.
Civilian and military pilots were generally identified and
entered into the database by the type or registration number
of airplane flown, or by military rank notations written in
the log. There may be some military pilots who did not include
rank, and who were flying civilian aircraft, that were counted
as civilian. I have no way of determining that kind of error,
and I encourage others who may analyze the database to correct
Likewise, female pilots were identified by given name. If
a woman signed in using only first initials, I had no way
of identifying her as female. In the absence of evidence to
the contrary, I entered as "Civ" (male) any pilots
identified with only initials for their given name. Again,
I encourage others who analyze the database to correct that
type of error.