APG-L Archives

Archiver > APG > 2008-06 > 1213595823

Subject: [APG] Describing Degrees of Confidence
Date: Mon, 16 Jun 2008 01:57:03 EDT

The following request is being crossposted to APG, BCG and TGF
mail lists. If you are inclined to take part, please submit only one
response, but name the list where you first saw it.

Elizabeth Shown Mills, in Evidence Explained, page 19, presented a list of
terms that can be used to describe and qualify levels of confidence we have in
information, evidence and conclusions. They were arranged in what she called
a logical hierarchy, ranging from the highest, certainly, to the lowest,
perhaps. Each is accompanied by a word description, and you may find it helpful
to read over them before going on to the short survey I am asking you to

The purpose of the survey is to determine how uniformly people perceive
these and some other similar terms, and the extent of any differences that may
exist. In order to do so, you will be asked to assign numbers to these
qualitative terms, to provide some quantitative basis for estimating the extent of
any differences that may exist, even though the statistical validity of the
results may be limited.

To respond, cut and paste the material between the horizontal lines below
into an email addressed to me: (mailto:)
rather than use the "reply" button and flood the list, digest and archives
with individual replies. If you prefer to respond anonymously, you may mail
your answers to me at 2004 Kentmere Pkwy, Wilmington DE 19806.

I'll let you know generally through this list what the answers show, and
may, if warranted, offer more detail to a publication for consideration.
Please note beside each of the terms that follow the lowest percentage of
probability you associate with the term, or if you prefer, the range of
probabilities for which you find the term appropriate. (A 90% probability would mean
chances are 90 out of 100 that an assertion is true, correct, or in accord
with what happened.)

The first group contains the terms from Evidence Explained, arranged from
highest to lowest level of confidence. The second group, in no particular
order, are other descriptive terms that have been used in an attempt to
communicate levels of confidence more precisely, including negative ones.

Thank you for your cooperation.


Mail list where first seen:

Please place beside each term either (1) the lowest numerical probability
you associate with the term, or (2) the range of probabilities, from lowest to
highest, that you associate with the term. You may use either alternative, as
seems most appropriate for a particular term.

Terms from Evidence Explained







Other terms describing confidence levels

Almost certainly


Highly likely

Very probably


Very unlikely

Hardly possible

Quite impossible


Caution: The statistical limitations of this survey are recognized. The
sample will be self-selected rather than random, little consideration has been
given to the order in which the items are presented or its possible effect, and
there has been no pre-testing or other validation. However, I believe the
results may be of interest to the extent they show differences in how we may
view these terms, and allow some analysis of whatever variation appears.

Donn Devine, CG, CGL
Wilmington, Delaware, USA

CG, Certified Genealogist, CGL, and Certified Genealogical Lecturer are
service marks of the Board for Certification of Genealogists, used under license
by board certificants after periodic evaluation, and the board name is
registered in the US Patent & Trademark Office.

**************Vote for your city's best dining and nightlife. City's Best
2008. (http://citysbest.aol.com?ncid=aolacg00050000000102)

This thread: