Monthly energy review, March 1995 Page: 9 of 186
This report is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to UNT Digital Library by the UNT Libraries Government Documents Department.
Extracted Text
The following text was automatically extracted from the image on this page using optical character recognition software:
N
The Response Analysis Survey:
Evaluating Manufacturing Energy
Consumption Survey Methodology
by Robert K. Adler*The Manufacturing Energy Consumption Survey (MECS)
has been conducted triennially by the Energy Information
Administration (EIA). It is a major data-gathering effort that
involved more than 14 thousand manufacturing establish-
ments in 1991, the year of the most recent published survey.
Because MECS is the only comprehensive source of data on
U.S. manufacturing energy use, EIA continually seeks ways
to improve its accuracy and effectiveness. In 1985 and 1986,
before the first MECS was launched, EIA conducted a pilot
survey of 78 manufacturing establishments to pretest the
MECS format, instructions, and questions. Since then, on-
going querying of participants has led to the reshaping of
several sections of the survey.
With the 1991 MECS,1 new groups of questions were added
concerning manufacturers' allocation of fuel to specific end
uses, the square footage of manufacturing establishments,
and the use of energy-saving technologies. To evaluate the
effectiveness of those new questions and participants' ease
of response to them, EIA developed and conducted a Re-
sponse Analysis Survey (RAS) of selected MECS respon-
dents in late 1992. The RAS also provided an opportunity to
solicit open-ended suggestions for improving the MECS in
general.
This "EIA Data News" item discusses the sample of MECS
respondents included in the RAS, the RAS's design and
execution, the results of the survey, and the ways those
results contributed to the design of the 1994 MECS.
RAS Sample Design and Survey
Methodology
Unlike the MECS, which is required by law and thus can
command high response rates, the RAS was entirely volun-
tary. In order to secure adequate, representative coverage of
the MECS sample, a target of 200 RAS responses was
selected. To offset likely refusals to participate, 400 establish-
ments were selected from the 1991 MECS sample of 14,299
*Robert K. Adler is a survey statistician with the Energy Information
Administration's Office of Energy Markets and End Use (EMEU). He
gratefully acknowledges the contribution to this article of Thomas Prugh, an
energy writer on contract to EMEU.
Energy Information Administration (EIA), Manufacturing Consumption
of Energy 1991, DOE/EIA-0512(91) (Washington, DC, December 1994).and approached to take part in the RAS. The final RAS
sample numbered 199.
The 199 RAS establishments were selected not at random,
but rather to reflect the proportions of the various major
industry groups in the MECS sample. If the contact person
at an establishment declined to take part in the RAS, another
establishment from the same Standard Industrial Classification
(SIC) and size group was contacted. Of the 199 cases, 43
replaced first-round selections. This procedure ensured that
SIC groups that were more heavily represented in the MECS
sample, such as the food and kindred products industry and the
chemicals and allied products industry, also enjoyed propor-
tionally greater representation in the RAS sample.
The RAS establishments were contacted by telephone 2
months after receipt of the MECS questionnaires. Each
interview proceeded immediately or at a later time more
convenient for the respondent. Interviews normally lasted
no more than 20 minutes.
Although the RAS posed a total of 29 questions, the actual
number asked of a given respondent depended upon the
relevance of certain questions or sets of questions. Questions
were grouped as follows:
" Twelve questions, including two multipart questions,
pertained to end-use consumption. For example, re-
spondents were asked to name the major source of
information they used in preparing their estimates of
end-use consumption as a fraction of total consump-
tion and to rate their confidence in the accuracy of
those sources and the difficulty of estimating the end-
use breakdown by energy source. They were also
asked if they could have provided actual end-use con-
sumption estimates.
" Ten questions pertained to establishments' total square
footage and to square footage that was heated or cooled
or both. Again, respondents were asked to identify the
major sources of information used in arriving at their
estimates. In the MECS, respondents were asked to give
estimates in terms of ranges, and RAS respondents were
2Size was defined as large or small, depending upon whether the estab-
lishment was above or below the median for its SIC group in terms of its
energy measure of size, a composite index derived from Bureau of the
Census data on each establishment's quantity of purchased electricity and
the cost of purchased fuels other than electricity.vii
Energy Information Administration/Monthly Energy Review March 1995
4t?~. 7'
EIA Data News
Upcoming Pages
Here’s what’s next.
Search Inside
This report can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Report.
Monthly energy review, March 1995, report, March 28, 1995; Washington D.C.. (https://digital.library.unt.edu/ark:/67531/metadc683921/m1/9/: accessed April 25, 2024), University of North Texas Libraries, UNT Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.