CHANGE FORTHE
* ~ BETTER WITH
ENERGY STAR
Products that earn the ENERGY STAR prevent
greenhouse gas emissions by meeting strict energy
efficiency guidelines set by the U.S. Environmental
Protection Agency and the U.S. Department of Energy.
www.energystar.gov
NATIONAL AWARENESS OF
ENERGY STAR® FOR 2003
ANALYSIS OF CEE
HOUSEHOLD SURVEY
9
M
V
-------
Table of Contents
ACKNOWLEDGMENTS i
EXECUTIVE SUMMARY ES-1
INTRODUCTION 1
METHODOLOGY 1
KEY FINDINGS 4
RECOGNITION 4
UNDERSTANDING 8
INFLUENCE 12
INFORMATION SOURCES 16
APPENDICES:
APPENDIX A: DETAILED METHODOLOGY A-l
APPENDIX B: DEMOGRAPHICS B-l
APPENDIX C: 2003 CEE WebTV/Internet QUESTIONNAIRE C-l
-------
ACKNOWLEDGMENTS
The authors of this study, Miriam Goldberg (KEMA-XENERGY Inc.), Valy Goepfrich (KEMA-
XENERGY Inc.), and Jocelyn Spielman (The Cadmus Group, Inc.) would like to thank the
Consortium for Energy Efficiency (CEE) for making its survey data available for this analysis. The
following CEE member organizations sponsored the 2003 survey:
• Alliant Energy
• Bonneville Power Administration
Key Span
Long Island Power Authority
• MidAmerican Energy Company
• Midwest Energy Efficiency Alliance
• National Grid USA
• Nevada Power
• Northwest Energy Efficiency Alliance
• NSTAR Electric & Gas
Sacramento Municipal Utility District
Southern California Edison
Unitil Corporation
Western Massachusetts Electric Company
Wisconsin Dept. of Administration
• Xcel Energy
In addition, we would like to extend special thanks to the following individuals for their
contributions:
• Marc Hoffman, Executive Director of the Consortium for Energy Efficiency (CEE), and Monica
Nevius, CEE's Research and Evaluation Manager, for making the study data available to EPA for
the national analysis.
• Maureen McNamara of the U.S. Environmental Protection Agency (EPA) ENERGY STAR
Program for project management and oversight of this report.
Tim Pettit of Nexus Market Research for support developing the initial methodology and
analysis.
1
-------
EXECUTIVE SUMMARY
In the fall of 2003, members of the Consortium for Energy Efficiency (CEE) sponsored the fourth
national household survey of consumer awareness of ENERGY STAR. Each year, the survey
objectives have largely been the same: to collect national data on consumer recognition,
understanding, and purchasing influence of the ENERGY STAR label, as well as data on messaging,
product purchases, and information sources used by consumers in their purchasing decisions. Some
CEE members also chose to supplement the national sample in order to better gauge awareness in
their local service territories. Additional survey cases were collected in Massachusetts, the Midwest,
and the Pacific Northwest. As in the three previous years, CEE and the sponsoring members made
the survey data publicly available.
This report discusses the results of the CEE 2003 ENERGY STAR Household Survey, building on
prior years' survey results and focusing on the extent to which consumers recognize the ENERGY
STAR label, understand its intended messages, and use (or are influenced by) the label in their
energy-related purchasing decisions. Research questions of interest included:
Where do consumers see or hear about the ENERGY STAR label?
How does increased publicity impact ENERGY STAR label recognition, understanding, and
influence?
Which key messages about the ENERGY STAR label are consumers retaining?
Do consumers demonstrate loyalty to the ENERGY STAR label?
Key Findings at the National Level
Fifty-six percent of households recognize the ENERGY STAR label (with a visual aid).
Sixty-three percent of households have a "high" or "general" understanding of the label.
Furthermore, the proportion of households that demonstrate a general understanding is small
compared with the percent of households that demonstrate a high understanding, 13 percent
versus 50 percent.
Of households that recognized the ENERGY STAR label and purchased a product in the past
twelve months, 58 percent purchased an ENERGY STAR-labeled product.
Considering households that recognized the label and those that did not (i.e., all households), 22
percent of households knowingly purchased an ENERGY STAR-labeled product in the past
twelve months.
For 52 percent of households that knowingly purchased an ENERGY STAR-labeled product, the
presence of the label influenced their purchasing decision "very much" or "somewhat." For
another 14 percent of households, the presence of the label influenced their purchasing decision
"slightly."
ES-1
-------
Fifteen percent of households that knowingly purchased an ENERGY STAR-labeled product
received a financial incentive. Sixty-five percent of these households report they would have
been "very likely" (51 percent) or "somewhat likely" (14 percent) to purchase the labeled product
without the financial incentive.
Sixty-six percent of households that knowingly purchased an ENERGY STAR-labeled product
report they were "very likely" or "somewhat likely" to recommend labeled products to a friend,
and another 16 percent report they were "slightly likely."
Key Findings from Publicity-level Analyses
A larger proportion of households in high- than in low-publicity areas recognize the ENERGY
STAR label, both with and without a visual aid. (High-publicity areas are areas with active
ENERGY STAR promotions by a regional/local energy efficiency program sponsor for two or
more continuous years.)
Considering only households that recognized the label (with a visual aid), a larger proportion of
these households in high- than in low-publicity areas associate the ENERGY STAR label with
products heavily promoted by regional program sponsors. From September 2002 to September
2003, many regional programs participated in product-specific seasonal campaigns focusing on
lighting, home electronics, appliances, and heating and cooling products.
A larger proportion of households in high- than in low-publicity areas have at least a general
understanding of the ENERGY STAR label.
For three messages of the ENERGY STAR label, a larger proportion of households in high- than
in low-publicity areas associate the message with the label. These three messages are "energy
efficiency or energy savings," "associating specific products with the ENERGY STAR label,"
and "energy product or environmental standards."
A larger proportion of households in high- than in low-publicity areas knowingly purchased an
ENERGY STAR product within the past 12 months.
Considering households that recognized the label (with a visual aid), a larger proportion of these
households in high- than in low-publicity areas have seen or heard something about ENERGY
STAR via store displays, utility mailings or bill inserts, TV commercials, newspaper or magazine
advertisements, and radio commercials.
Conclusions and Future Directions
This fourth national study of household awareness of the ENERGY STAR label confirms key
findings from the previous years' surveys:
Substantial portions of the U.S. households in the surveyed population recognize, understand,
and are influenced by the ENERGY STAR label.
ES-2
-------
The proportion of households that exhibit only a general understanding of the label is small (13
percent) compared with the proportion that exhibit a high degree of understanding (50 percent).
Publicity from active regional/local energy efficiency program sponsors increases recognition
and understanding of the label.
Furthermore, between 2003 and 2002, consumer recognition of the ENERGY STAR label increased
15 percentage points, from 41 to 56 percent.
The results of the CEE 2003 ENERGY STAR survey indicate that activities to promote the
ENERGY STAR label carried out by EPA, DOE, survey sponsors, and ENERGY STAR partners in
2003 were effective. In addition, a comparison of the sources that households recognizing the
ENERGY STAR label consult for product information against the venues in which they reported
seeing or hearing something about ENERGY STAR, suggests future ENERGY STAR activities
should aim to:
Increase positive exposure of ENERGY STAR in newspapers and magazines, and on TV.
Expand cooperative efforts to include ENERGY STAR materials in utility mailings or bill
inserts.
• Improve the availability of ENERGY STAR information on the Internet.
Enhance efforts to train salespersons and contractors to actively and accurately deliver
information about ENERGY STAR.
ES-3
-------
INTRODUCTION
In the fall of 2003, members of the Consortium for Energy Efficiency (CEE) sponsored the fourth
national household survey of consumer awareness of ENERGY STAR. Each year, the survey
objectives have largely been the same: to collect national data on consumer recognition,
understanding, and purchasing influence of the ENERGY STAR label, as well as data on messaging,
product purchases, and information sources used by consumers in their purchasing decisions. Some
CEE members also chose to supplement the national sample in order to better gauge label awareness
in their local service territories. Additional survey cases were collected in Massachusetts, the
Midwest, and the Pacific Northwest. As in the three previous years, CEE and sponsoring members
made the survey data publicly available.
This report discusses the results of the CEE 2003 ENERGY STAR Household Survey, building on
prior years' survey results and focusing on the extent to which consumers recognize the ENERGY
STAR label, understand its intended messages, and use (or are influenced by) the label in their
energy-related purchasing decisions. Research questions of interest included:
Where do consumers see or hear about the ENERGY STAR label?
How does increased publicity impact ENERGY STAR label recognition, understanding, and
influence?
Which key messages about the ENERGY STAR label are consumers retaining?
Do consumers demonstrate loyalty to the ENERGY STAR label?
This report has two parts. Part I includes an Executive Summary, this introduction, a summary of
methods, key findings in four sections, and three appendices. Appendix A is the detailed
methodology, Appendix B considers demographic information from the 2003 WebTV/Internet
survey, and Appendix C provides a copy of the 2003 WebTV/Internet questionnaire. Part II presents
the 2003 WebTV/Internet survey results by publicity category. In all cases, the results presented are
properly weighted to obtain national estimates.
METHODOLOGY
From September through October 2003, CEE fielded a household survey to obtain information at the
national level on consumer awareness of the ENERGY STAR label. The survey was fielded to a
random sample of households that are part of a WebTV/Internet panel that is selected by random
digit dial and recruited by telephone.
The survey was similar to the 2002 (and 2001) WebTV/Internet surveys. As in the previous three
years, CEE and its sponsoring members made the survey data publicly available.
1
-------
The survey was a national survey. The sampling frame for the survey is all households in the 57
largest Nielsen Designated Market Areas® (DMAs). In 2003, the 57 largest DMAs accounted for
approximately 70 percent of U.S. television households. In addition, some CEE members sponsored
more intensive sampling (i.e., an oversample) for various states and metropolitan areas, which are
referred to here as "sponsor areas." The sponsor areas were:
Massachusetts
• Chicago DMA
Minneapolis-St. Paul DMA
• Iowa
Idaho
Montana
Oregon
Washington
Both Chicago, IL and Minneapolis-St. Paul, MN are among the 57 largest DMAs.
For each of the sponsor areas that is a state, the frame was not limited to the large DMAs, but
included the entire state. Thus, the complete frame for the study was the combination of the largest
DMAs and any portion of the sponsor areas that fell outside these DMAs.
To facilitate comparisons across years, the national results are based only on data collected from all
respondents in the top 57 DMAs. Data collected from respondents not in the top 57 DMAs, but in a
sponsor area, are not included in the analysis. Some of the top 57 DMAs are also included in the
sponsor areas and, therefore, were oversampled. The data from these respondents, as well as from
the other respondents in the top 57 DMAs, received the appropriate weight in the analysis to generate
valid national results and comparisons against data from other years.
As in previous years' studies, to consider the effect of publicity on national awareness, the DMAs in
the complete frame were classified by publicity category. The same publicity classification
procedure used the past two years was used this year.1 A DMA was classified as high publicity, low
publicity, or other using the following criteria:
• High publicity: At least two recent years of sustained promotions and publicity from
non-federal activities.
Low publicity: Federal campaign activities only and no significant regional program sponsor
activities.
• Other: All other DMAs.
Between September 2002 and 2003, only one DMA changed publicity category. Orlando, FL changed from "Low" to
"Other."
2
-------
This classification procedure identifies three publicity categories and provides clear and verifiable
definitions. The key working definitions are:
• Recent: The two years of activity must include the time of the survey fielding.
• Sustained: The two years of activity must be continuous.
• Significant: In addition to any direct federal publicity efforts,2 publicity efforts must include a
deliberate and multifaceted regional program sponsor investment in ENERGY STAR
programming, such as direct marketing and promotional efforts.
These definitions are sufficiently operational to be applicable to future survey efforts, and can be
modified by simply increasing the duration of sustained high publicity.
The sample is stratified by area and within an area by publicity category. The areas consist of each
sponsor area and the largest or parts of the largest DMAs that were not in a sponsor area as a single
area. Further stratifying each area by publicity category results in 15 strata.3 The CEE members who
funded the oversample for a sponsor area determined the number of sampling points allocated to the
area as a whole. This total number of sampling points was then allocated across publicity categories
present in a sponsor area proportional to population. In the single area consisting of the largest or
parts of the largest DMAs that were not in a sponsor area, each publicity category was allocated
approximately 333 sampling points.
This report presents the 2003 survey results at the national level and often by publicity category.
The publicity category results provide evidence on the effectiveness of EPA's model to increase
awareness of ENERGY STAR by supporting regional energy efficiency program sponsors. Results
are presented on consumer recognition, understanding, and purchasing influence of the ENERGY
STAR label, as well as on messaging, product purchases, and information sources used by
consumers in their purchasing decisions.
During the September 2001 to September 2003 period, EPA launched and maintained three television national Public
Service Announcements (PSAs) as part of its Change campaign. The Change campaign also included continued
distribution of several radio and print PSAs as a component of its overall outreach strategy.
3
Montana was not further stratified by publicity category because 95 percent of Montana households are in low-publicity
DMAs and the number of sampling points to be allocated across all of Montana was only about 50.
3
-------
KEY FINDINGS
RECOGNITION
In 2003, 56 percent of households recognized the ENERGY STAR mark when shown the label (i.e., aided
recognition). Approximately 33 percent of households correctly assessed whether or not they had seen or
heard of the ENERGY STAR label without first being shown the label (i.e., unaided recognition).
For purposes of this analysis, respondents are said to recognize the ENERGY STAR label if they have
seen or heard of the label before the survey. Recognition of the ENERGY STAR label was explored two
different ways. "Aided" recognition was measured by showing the label and asking if the respondent had
heard of or seen it before. Delivering the survey by WebTV/Internet also made it possible to measure
"unaided" recognition. Unaided recognition was measured by asking this same question, but without
showing the label. Both methods are useful measurements of label recognition, although unaided
recognition is more conservative.
Recognition results for both this year's and last year's WebTV/Internet surveys are summarized in the next
table. Aided recognition of the ENERGY STAR label is clearly higher this year than last year (the
difference is statistically significant at the 1-percent level, p-value < 0.0001). Between 2003 and 2002,
aided recognition increased 15 percentage points, from 41 to 56 percent. Unaided recognition is 5
percentage points higher this year than last year, 33 versus 28 percent (the difference is statistically
significant at the 5-percent level, p-value = 0.047).
Recognition of the ENERGY STAR Label
(Base = All respondents)
Recognize ENERGY
STAR label
2003
2002
Aided
(n=2,098)
Unaided
(n=1,853)
Aided
(n=1,091)
Unaided
(n=991)
Yes
Standard error
56%
1.7%
33%
1.7%
41%
2.2%
28°/c
2.1 °/c
Note: The unaided recognition results for both years are based on the question ESI: "Have you ever seen or heard of
the ENERGY STAR label?" The 2003 aided recognition results are based on five questions. (1) ES3A and (2) ES3B
are asked if ESI = "yes." ES3A: "Is this the label you have seen or heard of before?," whether the old or new label is
shown is randomly determined. ES3B: "Have you seen or heard of this version of the ENERGY STAR label?," where
the label shown is the one not shown previously. (3) ES3C and (4) ES3D are asked if ESI = "no." ES3C: "Please
look at the ENERGY STAR label on the left. Have you ever seen or heard of this label?," whether the old or new label
is shown is randomly determined. ES3D: "Have you seen or heard of this version of the ENERGY STAR label?,"
where the label shown is the one not shown previously. (5) ES6 is asked if either ES1 = "no" or both ES3 A and ES3B
= "no." ES6: "Now that you have had the opportunity to see the ENERGY STAR label, do you recall seeing or
hearing anything about it before this survey?," where both the old and new labels are shown. The 2002 aided
recognition results are based on two questions. (1) ES3: "Is this the label you have seen or heard of before?," which is
asked if ESl="yes." (2) ES6: "Now that you have had the opportunity to see the ENERGY STAR label, do you recall
seeing or hearing anything about it before this survey?," which is asked if either ESl="no" or ES3="no."
4
-------
Recognition by Publicity Category
Both aided and unaided recognition were higher in high-publicity areas (areas with an active local
ENERGY STAR program sponsored by a utility, state agency, or other organization for two or more
continuous years) than in low-publicity areas. Aided households in high-publicity areas recognized
the ENERGY STAR label at 67 percent versus 51 percent in low-publicity areas. Unaided
recognition was 50 percent in high-publicity areas compared with 21 percent in low-publicity areas.
Recognition of the ENERGY STAR Label by Publicity Category
(Base = All respondents)
80% n
¦ High Publicity
~ Low Publicity
•"Aided (n=2,098) •"Unaided (n=1,853)
~ ~~High- and low-publicity areas proportions are statistically different from each other at the 1-percent level of
significance (p-value<=0.01).
5
-------
Product Associations
Sixty-three percent of households have seen the ENERGY STAR label on refrigerators. Washing
machines, dishwashers, and computers were the next most commonly associated products with the
label, at 49 to 44 percent. Windows, room air conditioners, and central air conditioners were in the
35- to 30-percent range. Products supported by regional programs, such as refrigerators,
dishwashers, washing machines, and air conditioning equipment, show strong association with the
ENERGY STAR label. The strong association of the label with computers is probably the combined
effect of manufacturer labeling and the prevalence of these products in daily life. Twenty-four
percent of households associate the ENERGY STAR label with microwave ovens, which do not in
fact have an ENERGY STAR specification. However, microwave ovens were the least recognized
of all the appliances. Products that showed an increase in association with the ENERGY STAR
label from 2002 to 2003 were refrigerators, washing machines, newly built homes, furnaces/boilers,
and audio products.
Product Association With the ENERGY STAR Label
(Base = Recognize label aided, n=811)
"Refrigerator
'"Washing machine
Dishwasher
Computer or monitor
Window
Room air conditioner
Central A/C
Microwave oven
Television
"Newly built home
"Furnace/boiler
Compact fluorescent light bulb
Door
Lighting fixture
VCR
Insulation
Computer printer
"Audio product
Copying machine
Fax machine
Heat pump
Thermostat
Scanner
Skylight
Roofing material
163%
2 49%
Z3 46%
2 44%
] 35%
134%
30%
~ 24%
2 24%
2 20%
120%
2 17%
~ 15%
] 14%
113%
~ 13%
2 12%
]10%
110%
~ 6%
H 6%
5%
^]3%
0%
10%
20%
30%
40%
50%
60%
70%
Note: Q5(a, b, and c): "Now we're going to ask you about several groups of products. As you review the list,
please select each of the products, product literature, or packaging on which you have seen the ENERGY
STAR label.
***2003 and 2002 proportions are statistically different from each other at the 1-percent level of significance (p-
value<0.01). Proportion of households in 2003 is larger than in 2002.
**2003 and 2002 proportions are statistically different from each other at the 5-percent level of significance (p-
value<0.05). Proportion of households in 2003 is larger than in 2002.
6
-------
Product Associations by Publicity Category
For many products, a larger proportion of households in high- than in low-publicity areas associated
the product with the ENERGY STAR label. This was the case for compact fluorescent light bulbs
and newly built homes, as well as two or more products in each of the following categories: heating
and cooling products, home appliances, home electronics, and building materials. Regional energy
efficiency program sponsors promoted refrigerators and washing machines heavily, along with
various audio, lighting, and heating and cooling products.
Product Association With the ENERGY STAR Label by Publicity Category
(Base = Recognize label aided, n=811)
74%
*** Refrigerator
Washing machine
Dishwasher
54%
60%
44%
50%
43%
47%
Room air conditioner
14%
43%
Computer or monitor
Window
39%
42%
33%
Central A/C
25%
30%
Microwave o\«n
14%
28%
Television
25%
Furnace/boiler
7%
24%
'Compact fluorescent light bulb
"Newly built home
"VCR
12%
22%
¦ High Publicity
~ Low Publicity
12%
18%
9%
18%
Door
7%
18%
Lighting fixture
"Insulation
10%
17%
9%
13%
"Audio product
Computer printer
Copying machine
Fax machine
5%
3%
14%
12%
10%
9%
Scanner
6%
Heat pump
9%
¦ 7%
~7%
| 6%
Thermostat
"'Skylight
Roofing material
3%
4%
0%
10%
20%
30%
40%
50%
60%
70%
80%
~ ~~High- and low-publicity areas proportions are statistically different from each other at the 1-percent level of
significance (p-value<=0.01).
"High- and low-publicity areas proportions are statistically different from each other at the 5-percent level of
significance (p-value<=0.05).
*High- and low-publicity areas proportions are statistically different from each other at the 10-percent level of
significance (p-value<=0.10).
7
-------
UNDERSTANDING
In 2003, 63 percent of households have at least a general understanding of the ENERGY STAR
label. Furthermore, the proportion of households that exhibits only a general understanding is small
compared with the proportion that exhibits a high degree of understanding, 13 versus 50 percent.
Understanding was probed by asking respondents what messages came to mind when they saw the
ENERGY STAR label. Based on these messages, a respondent's understanding was classified as
high, general, or no understanding.
The results on understanding of the ENERGY STAR label for both this year's and last year's
WebTV/Internet surveys are provided in the next table. The proportions of households with at least
a general understanding of the ENERGY STAR label were similar between 2003 and 2002, 63 and
58 percent, respectively (the difference is not statistically significant).4
Level of Understanding of the ENERGY STAR Label
(Base = All respondents)
Level of Understanding
of the Label
2003
(n=2,206)
2002
(n=1,168)
High understanding
50%
46%
General understanding
13%
12%
No understanding
38%
42%
Total
100%
100%
Note: The level of understanding of the label is based on two questions. (1) If respondent
recognized the label unaided, ES2: "What does the ENERGY STAR label mean to you?" (2)
If respondent did not recognize the label unaided, ES4A1: "Please look at the ENERGY
STAR label(s) on the left. Type the messages that come to mind when you see the ENERGY
STAR label(s)." (In 2003, "labels" and in 2002 "label.")
Throughout the report, "not statistically significant" refers to not significant at the 10-percent level.
8
-------
Understanding by Publicity Category
Understanding of the ENERGY STAR label was greater in high- than in low-publicity areas. Sixty-
nine percent of households in high-publicity areas had at least a general understanding of the label
compared with 61 percent of households in low-publicity areas. (The difference is statistically
significant at the 10-percent level, p-value = 0.073.)
In both publicity categories, among those households with at least a general understanding of the
ENERGY STAR label, more respondents exhibited a high degree of understanding.
Understanding of the ENERGY STAR Label by Publicity Category
(Base = All respondents)
80% n
70% -
60%
50%
10.3%
16.4%
40% -
30%
20%
58.3%
44.9%
10%
0%
High
Low
~ General Understanding
¦ High Understanding
9
-------
Label Messaging
Open-ended responses used to measure understanding are also an indicator of how effectively EPA
communicates its messages through the ENERGY STAR label. By far, the most common message
associated with the label is "energy efficiency or energy savings," which is considered high
understanding of the label. Forty-seven percent of households associate the ENERGY STAR label
with this message. The second most common message is "associating specific products with the
ENERGY STAR label," at 13 percent of households. "Associating specific products with the
ENERGY STAR label" is considered general understanding of the label.
Messages of the ENERGY STAR Label
(Base = All respondents)
'Energy efficient/savings I I 47%
'"Environmental benefit
Energy conservation
-cr
-Save money on operation 5% High Understanding
'Energy/environmental product standards
Savings (not linked to operation)
Mentions specific products 13%
Energy no link to efficiency
***Bectricity I 14% General Understanding
Confuses w ith Energy Guide
Environmental no link to benefit
Quality ¦ 1 %
Product standards no environmental link I 1%
'Government backing 11%
Save money on purchase | <1 %
1 I I I I I I I I I I
0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50%
***2003 and 2002 proportions are statistically different from each other at the 1-percent level of significance (p-
value<=0.01). For "Save money on operation" and "Electricity," proportion of households in 2003 is larger
than in 2002. For "Environmental benefit," proportion of households in 2003 is smaller than in 2002.
**2003 and 2002 proportions are statistically different from each other at the 5-percent level of significance (p-
value<=0.05). Proportion of households in 2003 is larger than in 2002.
*2003 and 2002 proportions are statistically different from each other at the 10-percent level of significance (p-
value<=0.10). Proportion of households in 2003 is larger than in 2002.
10
-------
Messaging by Publicity Category
For most messages, the proportion of households that associated the message with the ENERGY
STAR label was similar for high- and low-publicity areas. For three messages, however, a larger
proportion of households in high- than in low-publicity areas associated the messages with the label.
These three messages are "energy efficiency or energy savings," "associating specific products with
the ENERGY STAR label," and "energy product standards or environmental product standards." In
addition, a smaller proportion of households in high- than in low-publicity areas associated the
message "energy no link to efficiency" with the ENERGY STAR label. Associating either "energy
efficiency or energy savings" or "energy product standards or environmental product standards" with
the ENERGY STAR label is considered high understanding of the label. Associating either specific
products or "energy no link to efficiency" with the ENERGY STAR label is considered general
understanding of the label.
Messages of the ENERGY STAR Label by Publicity Category
(Base = All respondents)
"Energy efficient/savings
Energy conservation
Environmental benefit
Save money on operation
'Energy/environmental product standards
Savings (not linked to operation)
155%
145%
XT
Hgh Understanding
| 3%
12%
"Mentions specific products
'Energy no link to efficiency
Electricity
Confuses w ith Energy Guide
Environmental no link to benefit
Product standards no environmental link
Government backing
Save money on purchase
Quality
116%
111%
I 4%
^7%
General Understanding
~
~
10% 20%
30%
40% 50%
60%
~ ~~High- and low-publicity areas proportions are statistically different from each other at the 1-percent level of
significance (p-value<=0.01).
"High- and low-publicity areas proportions are statistically different from each other at the 5-percent level of
significance (p-value<=0.05).
*High- and low-publicity areas proportions are statistically different from each other at the 10-percent level of
significance (p-value<=0.10).
11
-------
Understanding by Aided Recognition
Households that recognize the ENERGY STAR label with a visual aid are more likely to have at
least a general understanding of the label than those who do not recognize the label. Among
households that recognize the label, 73 percent have at least a general understanding of the label,
compared with households that do not recognize the label at 50 percent. (The difference is
statistically significant at the 1-percent level, p-value < 0.0001.)
Understanding of the ENERGY STAR Label
by Aided Recognition of the Label
(Base = All respondents)
Recognize ENERGY
At Least General
STAR Label Aided
Understanding of Label
Yes
73%
No
50%
Yes-No
23%
p-value
<0.0001
INFLUENCE
The survey provided some information on consumers' decisions to purchase ENERGY STAR-
labeled products, including the following:
The proportion of households, nationally, that recognize the ENERGY STAR label and report
actually purchasing a labeled product.
The influence of the label on purchasing decisions.
The role of rebates or financing in decisions to buy ENERGY STAR products.
The loyalty of ENERGY STAR purchasers.
12
-------
Purchases of ENERGY STAR
In order to estimate the proportion of all households that knowingly purchased an ENERGY STAR
product, the following three proportions were multiplied:
The proportion of all households that recognized the ENERGY STAR label (aided).
Of the households that recognized the label, the proportion that purchased a product.
Of the households that recognized the label and reported purchasing a product, the proportion
that purchased an ENERGY STAR product.
The result of this question is that 22 percent of all households knowingly purchased at least one
qualifying ENERGY STAR product in the past twelve months.
Considering only those households that recognized the label in 2003, 58 percent purchased at least
one qualifying ENERGY STAR product in the past twelve months. This result is very similar to last
year's, 60 percent (the difference is not statistically significant).
Purchased ENERGY STAR
(Base = Recognize label aided and purchaser)
2003
(n=565)
2002
(n=228)
Purchased ENERGY STAR product
58%
60%
Note: Q7: "For any of the products you purchased, did you see the ENERGY STAR label (on
the product itself, on the packaging, or on the instructions)?"
Purchases of ENERGY STAR by Publicity Category
A larger proportion of all households knowingly purchased an ENERGY STAR product in high-
publicity areas than in low-publicity areas. Thirty percent of all households in high-publicity areas
knowingly purchased an ENERGY STAR product compared with 16 percent of all households in
low-publicity areas. (The difference is statistically significant at the 5 percent level, p-value = 0.012).
National Household Market Penetration
of ENERGY STAR Products by Publicity Category
(Base = All respondents)
Publicity Category
% Households
High
Low
30%
16%
High-Low
15%
p-value
0.012
13
-------
Influence of the ENERGY STAR Label
In 2003, for 52 percent of households that reported purchasing an ENERGY STAR-labeled product,
the presence of the label influenced their purchasing decision "very much" or "somewhat." For
another 14 percent of households, the presence of the label influenced their purchasing decision
"slightly."
The results on influence of the ENERGY STAR label for this year's and last year's surveys are
provided in the following table. The proportions of households for which the ENERGY STAR label
was at least somewhat influential in their purchasing decision were similar between 2003 and 2002,
52 and 46 percent, respectively (the difference is not statistically significant).
Influence of the ENERGY STAR Label on Purchasing Decisions
(Base = Recognize label aided and ENERGY STAR purchasers)
Influence of the Label on
Purchasing Decisions
2003
(n=319)
2002
(n=141)
Very much
22%
21%
Somewhat
30%
25%
Slightly
14%
21%
Not at all
33%
33%
Total
100%
100%
Note: Q8: "For any ENERGY STAR-labeled product(s) you purchased, how much did the
presence or absence of the ENERGY STAR label influence your purchasing decision?"
Influence of the ENERGY STAR Label by Publicity Category
For two of the three levels of influence of the ENERGY STAR label on purchasing decisions, the
proportion of households was similar for high- and low-publicity areas. However, a smaller
proportion of the purchasing decisions of households in high- than in low-publicity areas was very
much influenced by the ENERGY STAR label, 23 percent compared to 44 percent. This result was
unexpected. In none of the previous years was a smaller proportion of the purchasing decisions of
households in high- than in low-publicity areas influenced by the ENERGY STAR label. This year's
result may be an anomaly. If next year's results are similar, further examination will be warranted.
Influence of the ENERGY STAR Label on Purchasing Decisions by Publicity Category
(Base = Recognize label aided and ENERGY STAR purchasers, n=319)
Publicity Category
Very much
Very much
or somewhat
Very much,
somewhat, or
slightly
High
23%
61%
76%
Low
44%
51%
57%
High-Low
-21%
11%
18%
p-value
0.096
0.435
0.183
14
-------
Rebate and Financing Influence
Fifteen percent of the households that knowingly purchased an ENERGY STAR-labeled product
received rebates or reduced-rate financing. A large proportion of these households, 65 percent,
reported that they would have been "very likely" or "somewhat likely" to purchase the labeled
product if financial incentives had not been available.
Influence of Rebates and Financing on Purchasing Decisions
(Base = Recognize label aided, ENERGY STAR purchaser, and received an incentive, n=52)
Likelihood Purchase ENERGY
STAR Product Without
Financial Incentive
% Households
Very likely
51%
Somewhat likely
14%
Slightly likely
24%
Not at all likely
11%
Total
100%
Note: Q10: "If rebates or reduced-rate financing had not been available, how likely is it that
you would have purchased the ENERGY STAR-labeled product?"
Loyalty to ENERGY STAR
In 2003, 66 percent of households that reported purchasing an ENERGY STAR-labeled product say
they would be "very likely" or "somewhat likely" to recommend labeled products to a friend.
Furthermore, only 18 percent of households would be "not at all likely" to recommend ENERGY
STAR products to a friend.
The results on loyalty to the ENERGY STAR label for both this year's and last year's surveys are
shown in the next table. The proportions of households at least somewhat likely to recommend
labeled products to a friend were similar between 2003 and 2002, 66 and 63 percent, respectively
(the difference is not statistically significant).
Loyalty to ENERGY STAR
(Base = Recognize label aided and ENERGY STAR purchasers)
Likelihood Recommend
ENERGY STAR Products
2003
(n=292)
2002
(n=121)
Very likely
35%
39%
Somewhat likely
31%
24%
Slightly likely
16%
23%
Not at all likely
18%
14%
Total
100%
100%
Note: Q11: "How likely are you to recommend ENERGY STAR-labeled products to a
friend?"
15
-------
INFORMATION SOURCES
Sources Seen
Sixty-nine percent of households have seen something about ENERGY STAR on appliance or
electronic equipment labels, followed by store displays at 40 percent. Next, about 30 percent of
households have heard or seen something about ENERGY STAR on TV commercials or on utility
mailings or bill inserts. After these four sources, about 20 percent of households have seen
something about ENERGY STAR on EnergyGuide labels or in newspaper or magazine
advertisements. A larger proportion of households in 2003 than 2002 saw something about
ENERGY STAR on appliances or electronic equipment labels or in utility mailings or bill inserts, or
heard something about ENERGY STAR from contractors.
Sources Saw or Heard Something About ENERGY STAR
(Base = Recognize label aided, n=779)
Labels on appliances or electronic equipment
69%
Displays in stores
40%
TV commercial
30%
Utility mailing or bill insert
Yellow EnergyGuide label
23%
New spaper or magazine advertisement
19%
Internet
9%
Salesperson
7%
7%
New spaper or magazine article
Direct mail or circular advertisement
6%
Friend, neighbor, relative, or co-worker
4%
Billboard
3%
Radio commercial
3%
3%
Contractor
TV news feature story
2%
0%
Note: SOI: "Where did you see or hear something about ENERGY STAR? Please mark all that apply."
***2003 and 2002 proportions are statistically different from each other at the 1-percent level of significance (p-
value<=0.01). Proportion of households in 2003 is larger than in 2002.
**2003 and 2002 proportions are statistically different from each other at the 5-percent level of significance (p-
value<=0.05). For utility mailings or bill inserts, proportion of households in 2003 is larger than in 2002.
For radio commercials, proportion of households in 2003 is smaller than in 2002.
16
-------
Sources Seen by Publicity Category
For most sources, the proportion of households that have heard or seen something about ENERGY
STAR was similar for high- and low-publicity areas. For several sources, however, a larger
proportion of households in high- than in low-publicity areas have heard or seen something about
ENERGY STAR via the source. These sources are store displays, utility mailings or bill inserts, TV
commercials, newspaper or magazine advertisements, and radio commercials.
Sources Saw or Heard Something About ENERGY STAR by Publicity Category
(Base = Recognize label aided, n=779)
Labels on appliances or electronic equipment
"Displays in stores
***Utility mailing or bill insert
***TV commercial
Yellow Energy Guide label
*Newspaper or magazine advertisement
Salesperson
Internet
Newspaper or magazine article
Direct mail or circular advertisement
**Radio commercial
Billboard
Friend, neighbor, relative, or co-worker
Contractor
TV news feature story
I 73%
I 68%
I 52%
¦ High Publicity
~ Low Publicity
0% 10% 20% 30% 40% 50% 60% 70% 80%
~ ~~High- and low-publicity areas proportions are statistically different from each other at the 1 percent level of
significance (p-value<=0.01).
"High- and low-publicity areas proportions are statistically different from each other at the 5 percent level of
significance (p-value<=0.05).
*High- and low-publicity areas proportions are statistically different from each other at the 10 percent level of
significance (p-value<=0.10).
17
-------
Sources Consumers Consult for Product Information
The survey asked about the sources consumers are most likely to use to obtain information about
products covered by the ENERGY STAR program. The question was asked separately for two
product groups: (1) heating and cooling products, and (2) home appliances, lighting, and home
electronics. The results for the two product groups are similar. The top source was personal
acquaintances at around 60 percent, followed by consumer magazines, retailers, and the Internet. For
heating and cooling products, the proportion of households consulting each of these latter three
sources ranges from 36 to 41 percent. For appliances, home electronics, and lighting, this range is a
little broader, from 34 to 46 percent.
Product Information Sources Consulted
(Base = All respondents)
Friend/neighbor/etc.
Consumer magazines
Retailer
60%
Internet
Contractor
Utility program
Television
Newspapers
Radio
Other magazines
] 64%
41%
] 46%
38%
] 42%
¦ Heating and Cooling
Products(a) (n=1960)
~ Appliances, Home
Electronics, Lighting(b)
(n=1987)
50%
60%
70%
(a)Q13_l: "Now, please think only about Heating and Cooling Products. Please select the source(s) of
information you are most likely to use to obtain information about this product type. Please mark all that
apply."
(b)Q13_2: "Now, please think only about Home Appliances/Lighting/Home Electronics. Please select the
source(s) of information you are most likely to use to obtain information about this product type. Please
mark all that apply."
18
-------
Considering only households that recognized the ENERGY STAR label, there were clear differences
between the sources they consult for product information and where they saw or heard something
about ENERGY STAR. Most of these differences were not surprising and will persist due to the
nature of the sources. Nonetheless, the comparison of the sources households consult for product
information and where they saw or heard something about ENERGY STAR is informative. This
comparison suggests future ENERGY STAR activities should aim to:
Increase positive exposure of ENERGY STAR in newspapers, magazines, and on TV.
Expand cooperative efforts to include ENERGY STAR materials in utility mailings or bill
inserts.
Improve the availability and accessibility of ENERGY STAR information on the Internet.
Enhance efforts to train salespersons and contractors to actively and accurately deliver
information about ENERGY STAR.
ENERGY STAR Sources Compared With Sources Consulted
(Base = Recognized aided)
Sources
ENERGY
STAR
Sources
(n=779)
Sources Consulted
Heating and Cooling
Products
(n=1067)
Home Appliance/Lighting/
Home Electronics
(n=1079)
Newspaper or magazine
advertisement
19%
Consumer Reports,
other product-oriented
magazines
48%
Consumer Reports,
other product-oriented
magazines
51%
Newspaper or magazine article
7%
Newspaper
13%
Newspaper
16%
Other magazines
8%
Other magazines
12%
TV commercial
31%
19%
24%
TV news feature story
2%
Radio commercial
3%
8%
8%
Utility mailing or bill insert
30%
26%
15%
Internet
9%
42%
42%
Salesperson
7%
37%
45%
Contractor
3%
27%
18%
Friend,neighbor,relative,or co-
worker
4%
56%
61%
19
-------
APPENDIX A
DETAILED METHODOLOGY
From September through October 2003, the Consortium for Energy Efficiency (CEE) fielded a
household survey to obtain information on consumer awareness of the ENERGY STAR label. The
survey was fielded to a random sample of households that are part of a WebTV/Internet panel that is
selected by random digit dial and recruited by telephone. The survey was similar to the 2002 (and
2001) WebTV/Internet survey. As in the previous three years, CEE and its sponsoring members
made the survey data publicly available. In 2001, a rigorous comparative analysis of mail survey and
WebTV/Internet survey results was conducted. The results from both techniques were comparable
for most major indicators.5 Results in that time frame were also analogous to telephone surveys with
aided recognition.
This report discusses the results of the CEE 2003 ENERGY STAR Household Survey, building on
prior years' survey results and focusing on the extent to which consumers recognized the ENERGY
STAR label, understood its intended messages, and used (or were influenced by) the label in their
energy-related purchasing decisions. Research questions of interest included:
Where do consumers see or hear about the ENERGY STAR label?
How does increased publicity impact ENERGY STAR label recognition, understanding, and
influence?
Which key messages about the ENERGY STAR label are consumers retaining?
Do consumers demonstrate loyalty to the ENERGY STAR label?
The survey was fielded from September 17 through October 13, 2003.
The remainder of Appendix A discusses the questionnaire design, sampling and weighting
methodologies, data collection, and the national analysis.
1 Questionnaire Design
In 2003, CEE conducted the ENERGY STAR survey using a questionnaire designed to be delivered
by WebTV/Internet. The 2003 WebTV/Internet questionnaire was used in a survey conducted via an
interactive WebTV/Internet format in the homes of members of a WebTV/Internet panel. People on
the panel were originally selected to participate in the panel by random digit dial and recruited by
telephone. The panel is designed to be representative of the U.S. population. Panel members are
provided with an Internet appliance (WebTV) and Internet service connection, and surveys are
fielded to them via Internet and Web TV. Panel members who already have Internet service receive
other incentives to participate in the panel. Panel members receive three to four short surveys each
month, and are expected to respond to a percentage of these.
National Analysis of CEE 2001 ENERGY STAR® Household Surveys.
A-l
-------
The data from this survey may be compared with data collected using the 2002 and 2001
WebTV/Internet questionnaires, for which CEE was also responsible. Sampling for the survey is
discussed in Section 2, data collection is discussed in Section 3, and the national analysis is
discussed in Section 4.
The committee had several broad objectives in designing the 2003 questionnaire including:
• To maintain consistency with the CEE 2000 and 2001 mail surveys and the 2001 and 2002
WebTV/Internet surveys.
To fine-tune the questionnaire based on lessons learned from the analysis of the CEE 2000
survey, focusing on achieving the greatest value from the analysis of the CEE 2001 and 2002
surveys.
The 2003 WebTV/Internet questionnaire addressed the following:
Respondent recognition of the ENERGY STAR label.
Understanding of, and key messages communicated by, the ENERGY STAR label.
Sources of information consulted about ENERGY STAR.
Products on which respondents have seen the label.
Products that respondents have purchased in the past year.
Products that respondents have purchased on which they have seen the label (or on whose
packaging or instructions they have seen the label).
Influence of the presence or absence of the label on the purchase decision.
Whether purchases of ENERGY STAR-labeled products involved rebates or reduced-rate
financing.
Likelihood of having purchased ENERGY STAR-labeled products in the absence of rebates or
reduced-rate financing.
Likely sources of information about product categories.
Demographic questions. (Most of the demographic questions were not asked in the
WebTV/Internet survey, because demographic characteristics of the respondents were already on
file.)
Likelihood of recommending ENERGY STAR-labeled products to a friend.
Recognition and understanding of the yellow EnergyGuide labels.
A-2
-------
The 2003 WebTV/Internet questionnaire is very similar to the 2002 questionnaire. The aided
recognition questions were changed to accommodate the new ENERGY STAR label and
experimental questions about how households view companies that produce or sell ENERGY
STAR-labeled products was added.
To assess awareness of both the old and new ENERGY STAR labels, in the 2003, survey
respondents were asked if they recognized both labels. To control for question order effects, the
order in which the labels were shown was randomized. The 2003 aided recognition questions were
presented as follows. (For the complete questionnaire, see Appendix C.)
Those respondents who answered "yes" to question ES1 ("Have you ever seen or heard of the
ENERGY STAR label?") were shown one of the labels in random order and asked ES3 A, and
then the other label and asked ES3B:
ES3A: "Is this the label you have seen or heard of before?"
ES3B: "Have you seen or heard of this version of the ENERGY STAR label?"
Those respondents who answered "no" to ESI were shown one of the labels in random order and
asked ES3C, and then the other label and asked ES3D:
ES3C: "Please look at the ENERGY STAR label on the left. Have you ever seen or
heard of this label?"
ES3D: "Have you seen or heard of this version of the ENERGY STAR label?"
Those respondents who answered "no" to either ESI or both ES3B and ES4B were shown both
labels and asked ES6:
ES6: "Now that you have had the opportunity to see the ENERGY STAR label, do you
recall seeing or hearing anything about it before this survey?"
A-3
-------
The experimental question (Q14) about how households view companies that produce or sell
ENERGY STAR-labeled products was asked as follows: "We're interested in how you view
companies that produce or sell ENERGY STAR-labeled products. Below are phrases that some
people have used to describe these companies. On the scale by each word or phrase, please indicate
how strongly you agree or disagree with each." The response scale was 1 = strongly agree, 2 =
somewhat disagree, 3 = neither agree nor disagree, 4 = somewhat agree, and 5 = strongly agree. The
order in which the words or phrases were presented was random. The words or phrases were:
Responsible • Uncaring
Trustworthy • Impractical
Industry leader • Innovative
• Behind the times
Agreement with positive characteristics (and disagreement with negative ones) can be used as a lever
to encourage vendor involvement with ENERGY STAR. The answers can also help identify the
positive features that are most effectively identified with ENERGY STAR.
Results were favorable for this question, with 61% of households who recognize the label having a
favorable impression of companies that manufacture or sell ENERGY STAR. That is, they agreed
with at least one of four positive statements. By contrast, very few, less than 9%, gave a negative
response.
The question series as worded on the 2003 survey had some limitations with regard to these
interpretations. One is that several of the adjectives respondents were asked to agree or disagree
with are not part of the ENERGY STAR brand messaging, and do not necessarily have any
association with the label one way or another. As a result, many respondents had difficulty either
agreeing or disagreeing, and chose the neutral response.
Another limitation was that the questions were asked of all respondents, whether or not they
recognized the label. The rate of neutral responses was much higher among non-recognizers.
On the other hand, the difference between responses for recognizers and non-recognizers is an
indication of the level of meaningful associations over and above systematic response tendencies
regardless of having any direct experience or association with the label. For one such comparison,
the question responses were coded from -2 (strongly disagree with positive, or strongly agree with
negative) to +2 (strongly agree with positive or strongly disagree with negative). On this basis, both
recognizers and non-recognizers tended toward positive ratings on all questions, but the recognizer
average score tended to be higher. The average score for the non-recognizers was around 0.2 across
the questions, while recognizers' average scores were usually 0.5 or higher.
This comparison does indicate that the label carries positive associations for those who recognize it.
Nonetheless, benchmarking against other kinds of brands would be more useful in understanding the
effectiveness of the ENERGY STAR label.
A-4
-------
In light of these considerations, CEE may wish to reassess the wording and intent of these questions.
Possibly changes could be:
Ask a single question, rating vendors on a 5-point scale from unfavorable to favorable.
Ask this question in parallel with other questions that ask for similar ratings of other brands or
institutions.
Ask about "bundles" of vendor characteristics, some of which do and some of which do not
include selling ENERGY STAR.
The interactive format of a WebTV/Internet questionnaire allows questions to be asked in a way that
is not possible with a printed questionnaire. On printed questionnaires, respondents can see
questions in advance. For example, while the 2000 and 2001 mail questionnaires begin by showing
the ENERGY STAR label and asking about understanding and whether they recognize it before
asking other questions, respondents can still potentially educate themselves in a limited way about
the ENERGY STAR label by reading the survey before completing it, affecting their responses. The
WebTV/Internet questionnaires (after questions about the yellow EnergyGuide label), however, ask
respondents—without showing the label—whether they have ever seen or heard of the ENERGY
STAR label. Responses to this question should thus be comparable to those obtained through a
telephone survey.
The WebTV/Internet questionnaires then show the ENERGY STAR label(s) (which is obviously not
possible with the telephone questionnaire) about understanding and recognition. Responses to this
question (or these questions) should thus be comparable to those obtained through the mail survey
where respondents are shown the label. Other differences between the mail questionnaires and the
WebTV/Internet questionnaires are that the latter—much like a telephone questionnaire using
computer-assisted telephone interviewing (CATI)—can program lines of questions based on
responses to earlier questions. For example, WebTV/Internet respondents who say they have bought
a given product in the past year can then be asked whether that specific product (or its packaging or
instructions) had the ENERGY STAR label.
A-5
-------
2 Sampling
2.1 Designated Marketing Areas Publicity Categories
The same publicity classification procedure used the past two years was used this year. A Nielsen
Designated Marketing Area® (DMA) was classified as high publicity, low publicity, or other using
the following criteria:
• High publicity: At least two recent years of sustained promotions and publicity from non-
federal activities.
Low publicity: Federal campaign activities only and no significant regional program sponsor
activities.
• Other: All other DMAs.
This classification procedure identifies three publicity categories and provides clear and verifiable
definitions. The key working definitions are:
• Recent: The two years of activity must include the time of the survey fielding.
• Sustained: The two years of activity must be continuous.
• Significant: In addition to any direct federal publicity efforts,6 publicity efforts must include a
deliberate and multifaceted regional program sponsor investment in ENERGY STAR
programming, such as direct marketing and promotional efforts.
These definitions are sufficiently operational to be applicable to future survey efforts, and can be
modified by simply increasing the duration of sustained high publicity. The publicity-level
assignments are detailed in the table below, followed by a table of supplemental CEE member
sponsor areas.
During the September 2001 to September 2003 period, EPA launched and maintained three national television Public
Service Announcements (PSAs) as part of its Change campaign. The Change campaign also included continued
distribution of several radio and print PSAs as a component of its overall outreach strategy.
A-6
-------
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
Top 57 Designated Market Areas
# TV Households
% of US TV
Publicity
Designated Market Area (DMA)
2002-2003
Households
Category
New York
7,282,320
6.8%
High
Los Angeles
5,318,040
5.0%
High
Chicago
3,351,330
3.1%
Other
Philadelphia
2,830,470
2.7%
Other
San Francisco-Oak-San Jose
2,436,220
2.3%
High
Boston (Manchester)
2,353,500
2.2%
High
Dallas-Ft. Worth
2,195,540
2.1%
Other
Washington, DC (Hagrstwn)
2,169,230
2.0%
Other
Atlanta
1,971,180
1.8%
Low
Detroit
1,899,910
1.8%
Other
Houston
1,814,140
1.7%
Other
Seattle-Tacoma
1,659,100
1.6%
High
Tampa-St. Pete (Sarasota)
1,620,110
1.5%
Low
Minneapolis-St. Paul
1,594,740
1.5%
Other
Cleveland-Akron (Canton)
1,528,840
1.4%
Other
Phoenix
1,524,130
1.4%
Other
Miami-Ft. Lauderdale
1,486,860
1.4%
Other
Denver
1,366,250
1.3%
Other
Sacramnto-Stktn-Modesto
1,227,600
1.2%
High
Orlando-Daytona Bch-Melbrn
1,224,470
1.1%
Other
Pittsburgh
1,165,660
1.1%
Other
St. Louis
1,156,370
1.1%
Other
Portland, OR
1,061,080
1.0%
High
Baltimore
1,060,450
1.0%
Other
Indianapolis
1,019,870
1.0%
Other
San Diego
1,004,220
0.9%
High
Hartford & New Haven
980,410
0.9%
High
Charlotte
962,540
0.9%
Low
Raleigh-Durham (Fayetvlle)
929,460
0.9%
Low
Nashville
880,670
0.8%
Low
Milwaukee
860,350
0.8%
High
Cincinnati
854,250
0.8%
Low
Kansas City
852,510
0.8%
Other
Columbus, OH
835,780
0.8%
Other
Greenvll-Spart-Ashevll-And
792,110
0.7%
Low
Salt Lake City
769,230
0.7%
Other
San Antonio
718,730
0.7%
Low
Grand Rapids-Kalmzoo-B.Crk
713,800
0.7%
Other
West Palm Beach-Ft. Pierce
700,850
0.7%
Low
Birmingham (Ann and Tusc)
690,030
0.6%
Low
Norfolk-Portsmth-Newpt Nws
677,610
0.6%
Low
New Orleans
658,830
0.6%
Low
Memphis
653,840
0.6%
Low
Buffalo
639,190
0.6%
High
Oklahoma City
636,970
0.6%
Low
Greensboro-H.Point-W. Salem
634,140
0.6%
Low
Harrisburg-Lncstr-Leb-York
626,660
0.6%
Other
Providence-New Bedford
624,020
0.6%
High
Albuquerque-Santa Fe
620,230
0.6%
Low
Louisville
612,300
0.6%
Other
Jacksonville, Brunswick
587,200
0.6%
Low
Las Vegas
585,440
0.5%
Other
Wilkes Barre-Scranton
580,290
0.5%
Low
Austin
552,060
0.5%
Other
Albany-Schenectady-Troy
532,520
0.5%
High
Little Rock-Pine Bluff
523,810
0.5%
Low
Fresno-Visalia
519,330
0.5%
High
Total
75,126,790
70.4%
A-7
-------
Sponsor Areas
Sponsor Area
Publicity
Category
Top 57/Other DMAs
Chicago
Other
Top 57: Chicago DMA (Rank 3)
Idaho
Other
Top 57: parts of Salt Lake City DMA (Rank 36)
Other: parts of Spokane DMA (Rank 79)
Low
Other:
*parts of Boise DMA (Rank 124)
*parts of Idaho Falls-Pocatello DMA (Rank 165)
Twin Falls DMA (Rank 191)
Iowa
Low
Other: parts of
*Des Moines-Ames DMA (Rank 72)
"Omaha DMA (Rank 78)
"Cedar Rapids-Wtrlo-IWC&Dub DMA (Rank 88)
"Davenport-R.lsland-Moline DMA (Rank 92)
"Sioux Falls(Mitchell) DMA (Rank 112)
"Sioux City DMA (Rank 140)
"Rochestr-Mason City-Austin DMA (Rank 152)
"Quincy-Hannibal-Keokuk DMA (Rank 164)
"Ottumwa-Kirksville DMA (Rank 198)
"Mankato DMA (Rank 199)
Massachusetts
High
Top 57: parts of
"Boston DMA (Rank 6)
"Providence-New Bedford (Rank 48)
"Albany-Schenectady-Troy DMA (Rank 55)
Other: Springfield-Holyoke DMA (Rank 106)
Minneapolis St. Paul
Other
Top 57: Minneapolis-St. Paul DMA (Rank 14)
Montana
Other
Other: parts of Spokane DMA (Rank 79)
Low
Other:
"parts of Minot-Bismarck-Dickinson DMA (Rank 155)
"parts of Billings DMA (Rank 170);
"Missoula DMA (Rank 169)
"parts of Rapid City DMA (Rank 175)
"Great Falls DMA (Rank 187)
"Butte-Bozeman, MT DMA (Rank 195)
"Helena DMA (Rank 207)
"Glendive DMA (Rank 210)
Oregon
High
Top 57: parts of Portland, OR DMA (Rank 23)
Other
Other:
"parts of Spokane DMA (Rank 79)
"Eugene DMA (Rank 121)
"parts of Yakima-Pasco-Rchlnd-Knnwck DMA (Rank 127)
Low
Other:
"parts of Boise DMA (Rank 124)
"parts of Medford-Klamath Falls DMA (Rank 141)
"Bend, OR DMA (Rank 201)
Washington
High
Top 57:
"Seattle-Tacoma DMA (Rank 12)
"parts of Portland, OR DMA (Rank 23)
Other
Other:
"parts of Spokane DMA (Rank 79)
"parts of Yakima-Pasco-Rchlnd-Knnwck DMA (Rank 127)
A-8
-------
Map of Top 57 DMAs and Sponsor Areas
by Publicity Category7
H "High" publicity category
L "Low" publicity category
O "Other" publicity category
~ CEE sponsor area (ranking in Top 57 DMAs)
~ CEE sponsor area (not ranking in Top 57 DMAs)
Neither Alaska nor Hawaii contained DMAs ranking in the Top 57 DMAs.
A-9
-------
2.2 Sample Design
The sample is a national sample. The sampling frame is all households in the 57 largest DMAs. In
2003, the 57 largest DMAs accounted for approximately 70 percent of U.S. television households.
In addition, some CEE members sponsored more intensive sampling (i.e., oversample) for various
states and metropolitan areas, which are referred to as "sponsor areas." Two of the sponsor areas
were among the largest DMAs and the remaining sponsor areas were states. For each sponsor area
that is a state, the frame was not limited to the large DMAs, but included the entire state. Thus, the
complete frame for the study was the combination of the 57 largest DMAs and any portion of the
sponsor areas that fell outside these DMAs.
The sample is stratified by area and within an area by publicity category. There are nine areas,
comprising eight sponsor areas plus one area consisting of the largest or parts of the largest DMAs
that were not in a sponsor area. Further stratifying each eight sponsor area by publicity category
results in 12 strata.8 Further stratifying the remaining area by publicity category results in 3 strata,
for a total of 15 strata.
The CEE members who funded the oversample for a sponsor area determined the number of
sampling points allocated to the area as a whole. This total number of sampling points was then
allocated across publicity categories present in a sponsor area proportional to population. In the
single area consisting of the largest or parts of the largest DMAs that were not in a sponsor area, each
publicity category was allocated approximately 333 sampling points. For each stratum, a larger
sample was selected to receive the survey to allow for nonresponse.
2.3 Weighting Procedures
The weights used in the analysis are the weights developed by Knowledge Networks, the company
that provides the WebTV/Internet survey service. Knowledge Networks begins with a typical
sampling weight that also accounts for differences between the WebTV/Internet panel and the U.S.
population of households. This adjustment is based on geographic and demographic characteristics
known for both the panel and the population. It is designed to scale up the groups that are under-
represented in the panel and scale down the groups that are over-represented in the panel so that they
are more closely aligned with the basic demographic characteristics of the U.S. population of
households.
The typical sampling weight is then corrected for survey nonresponse. The correction for survey
nonresponse is analogous to the adjustment for differences in the WebTV/Internet panel from the
U.S. population of households. The correction for survey nonresponse is based on geographic and
demographic characteristics known for both the sample of panel completes and the entire frame for
the study. It scales up the under-represented groups and scales down the over-represented groups in
the sample of panel completes.
Montana was not further stratified by publicity category because 95 percent of Montana households are in low-publicity
DMAs and the number of sampling points to be allocated across all of Montana was only about 50.
A-10
-------
3 Data Collection
3.1 Survey Implementation
The survey began on September 17 and closed on October 13, 2003.
3.2 Response Rates
For WebTV/Internet, the return rate is the ratio of the number of survey questionnaires completed to
the number of panel members who were asked to complete the survey. For the CEE 2003 ENERGY
STAR household survey, the return rate was 80 percent. While this number is quite high, it must be
adjusted by the recruitment rate, that is, the number of households that agreed to participate in the
WebTV/Internet panel as a proportion of the number of households asked to participate. Thus, the
WebTV/Internet response rate is the product of the survey-specific return rate and the recruitment
rate of 37 percent. This product is equivalent to the ratio of the number of surveys completed to the
number of households that were offered the opportunity to be in the study. For the CEE 2003
ENERGY STAR household survey the response rate was 30 percent. This level of response is
typical for a WebTV/Internet survey fielded to the Knowledge Networks panel.
Survey Response Rate
Sendout/requested
2,745
Completed
2,206
Return rate
80%
Recruitment rate
37%
Response rate
30%
4 National Analysis
To facilitate comparisons across years, the national results are based only on data collected from all
respondents in the 57 largest DMAs. Data collected from respondents not in the 57 largest DMAs,
but in a sponsor area, are not included in the national analysis. Some of the 57 largest DMAs are
also included in the sponsor areas and, therefore, were oversampled. The data from these
respondents, as well as from the other respondents in the 57 largest DMAs, received the appropriate
weight in the analysis to generate valid national results.
A-ll
-------
APPENDIX B
DEMOGRAPHICS
The analysis presented in this appendix suggests the weighted survey results are a reasonable
representation of the study population, which is all U.S. households. Professional survey and data
collection firms make significant efforts to ensure the rigor of their methods and to produce the
highest quality results. However, in any survey effort, the persons who respond to the survey tend to
be different from those who do not respond. While Knowledge Networks, the company that
maintains the WebTV/Internet panel, strives to create a representative panel for its WebTV/Internet
frame, the respondent base will contain subjects and their associated biases that are receptive to the
WebTV/Internet incentive-for-service tradeoff.
The weights used in the analysis attempt to account for differences between the WebTV/Internet
panel and the U.S. population of households and for survey nonresponse. To the extent this effort is
successful, the distribution of various demographic characteristics based on the weighted survey data
will be similar to the distribution based on national Census data. For most demographic
characteristics, the two distributions are similar. This suggests the weighted survey results are a
reasonable representation of the study population. A summary of the demographic characteristics
compared is provided in the table below, and the detailed comparisons are provided in the tables at
the end of this appendix.
Summary of Distribution Comparisons
Demographic Characteristic
Largest Difference (Absolute Value):
Survey Estimate Minus Census %
Number of persons in household
One
-7.3%
Householder/respondent age
65 or older
-6.7%
Householder/respondent gender
Gender
+/-1.0%
Dwelling type
Apt. bldg.
-6.7%
Own/rent
Own/rent
+/-3.5%
Household annual income
$25,000-$49,000
4.5%
The largest differences (in absolute value) between the weighted survey data and the national Census
data are all about 7 percentage points and concern the number of persons in a household,
householder/respondent age, and dwelling type. The difference between the weighted survey data
and the national Census data is about 7 percentage points for: one-person households, 19 versus 26
percent; householders 65 years or older, 14 versus 20 percent; and households living in apartment
buildings, 15 versus 22 percent. Neither the under-representation of one-person households,
householders 65 years or older, nor households living in apartment buildings is expected to bias the
survey results in a particular direction. For the remaining demographic characteristics, the largest
differences between the weighted survey data and the national Census data range between 1 and 5
percent.
B-l
-------
Household Size Distribution
Number of Persons in
Household
Census
% Dwelling Units3
Survey Estimate
Minus Census
% Dwelling Units
One
26%
-7.3%
Two
33%
4.8%
Three
16%
1.7%
Four
15%
0.0%
Five or more
10%
0.7%
Total (%)
100%
Total (1,000s)
106,261
a U.S. Census Bureau, American Housing Survey: 2001, Table 2-9.
Age Distribution
Householder/
Respondent Age
Census
% Householders3
Survey Estimate
Minus Census
% Householders
18-24b
6%
4.5%
25-34
17%
4.4%
35-44
22%
-0.8%
45-54
20%
-0.9%
55-64
15%
-0.4%
65 or older
20%
-6.7%
Total (%)
100%
Total (1,000s)
111,278
a U.S. Census Bureau, Current Population Reports, Income in the United
States: 2002, Table 3.
b Census, 15-24 years; WebTV/Internet 2003, 18-24 years.
Gender Distribution
Householder/
Respondent
Gender
Census
% Population3
Survey Estimate
Minus Census
% Population
Female
51%
1.0%
Male
49%
-1.0%
Total (%)
100%
a U.S. Census Bureau, Census 2000.
B-2
-------
Dwelling Type Distribution
Dwelling Type
Census
% Dwelling Units3
Survey Estimate
Minus Census
% Dwelling Units
Single-family, unattached
60%
3.4%
Single-family, attached
7%
5.7%
Apt. bldg. (>=2 units)b
22%
-6.7%
Mobile home
6%
-1.4%
Other
5%
-0.9%
Total (%)
100%
Total (1,000s)
111,577
a U.S. Census Bureau, American Housing Survey: 2001, Table 2-1.
b Census, 2 or more units; WebTV/Internet 2002,4 or more units.
Own/Rent Distribution
Own/Rent
Census
% Households3
Survey Estimate
Minus Census
% Households
Own
68%
-3.5%
Rent
32%
3.5%
Total (%)
100%
Total (1,000s)
106,261
a U.S. Census Bureau, American Housing Survey: 2001, Table 2-1.
Income Distribution
Total Household Annual
Income (before taxes)
Census
% Households3
Survey Estimate
Minus Census
% Households
Less than $15,000
16%
-3.0%
$15,000-$24,999
13%
-2.6%
$25,000-$49,999
27%
4.5%
$50,000-$74,999
18%
3.7%
$75,000 and over
25%
-2.7%
Total (%)
100%
Total (1,000s)
111,278
a U.S. Census Bureau, Current Population Reports, Income in the United
States: 2002, Table A-l.
B-3
-------
APPENDIX C
2003 WebTV/Internet QUESTIONNAIRE, 8/11/03
Yes
Don't
Know
No
Yes
No or Don't Know
C-l
ES3A.
Is this the label you have seen or heard of
before? [SHOW OLD OR NEW LABEL, IN
RANDOM ORDER]
ESI.
Have you ever seen or heard of the
ENERGY STAR label?
EG1.
Have you ever seen or heard of yellow
stickers called EnergyGuide labels?
ES2.
What does the ENERGY STAR label
mean to you?
EG2.
What information does the Energy Guide
label provide?
-------
^ Yes/No
I /Don't Know J
ES3C.
Please look at the ENERGY STAR label
on the left. Have you ever seen or heard
of this label? [SHOW OLD OR NEW
LABEL, IN RANDOM ORDER]
'
^Yes/No/
I Don't Know J
ES3B.
Have you seen or heard
of this version of the
ENERGY STAR label?
[SHOW LABEL NOT
PREVIOUSLY SEEN]
r
ES3D.
Have you seen or heard
of this version of the
ENERGY STAR label?
[SHOW LABEL NOT
PREVIOUSLY SEEN]
'
Yes to EITHER
or BOTH ES3 A
or ES3B
Yes/No/
Don't Know
No/Don t Know (or
combo of the two) to
both ES3 A and
ES3B
C-2
-------
SOI.
Where did you see or hear something
about ENERGY STAR? Please mark all
that apply.
[checkbox]
• Newspaper or magazine
advertisement
Newspaper or magazine article
TV commercial
TV news feature story
Radio commercial
Billboard
Utility mailing or bill insert
Direct mail or circular
advertisement
Labels on appliances or electronic
equipment
Yellow EnergyGuide label
Displays in stores
Internet
Salesperson
Contractor
Friend, neighbor, relative, or co-
worker
Other (please specify) [text box]
Don't know
ES4al.
Please look at the ENERGY STAR labels
on the left. Type the messages that
come to mind when you see the
ENERGY STAR labels.
[SHOW LABEL]
ES6.
Now that you have had the opportunity to
see the ENERGY STAR label, do you recall
seeing or hearing anything about it before
this survey?
rNo or
Don't Know
S02.
What did you see or hear about
ENERGY STAR? Please be specific.
i
r
Skip to Q6a
C-3
-------
Q5(c). Finally, please review the last of the product lists
below and select each of the products, product literature,
or packaging on which you have seen the ENERGY
STAR label.
Building Materials:
Buildings
Door
Roofing material
Window
Skylight
Newly built home
Insulation
None of these products
Home Appliances/Lighting:
Home Electronics Dishwasher
Television Refrigerator
VCR Lighting fixture
Audio product Washing machine
Compact fluorescent light bulb Microwave oven
None of these products
Q5(b). Please continue reviewing the lists of products
below, and select each of the products, product literature,
or packaging on which you have seen the ENERGY
STAR label.
Q5(a). Now we're going to ask you about several groups
of products. As you review the list, please select each of
the products, product literature, or packaging on which
you have seen the ENERGY STAR label.
Heating and Cooling Products:
Central air conditioner
Computer or monitor
Computer printer
Copying machine
Fax machine
Scanner
Home Office Equipment
Furnace or boiler
Heat pump
Thermostat
Room air conditioner
None of these products
-------
Have you or someone else in your
household been shopping in a store in the
last 12 months for any of the products listed
below?
~ Yes
~ No
~ Don't know
Heating and Cooling Products
• Central air conditioner
• Furnace or boiler
• Heat pump
• Thermostat
• Room air conditioner
Home Office Equipment
• Computer or monitor
• Computer printer
• Copying machine
• Fax machine
• Scanner
Home Appliances/Lighting
• Dishwasher
• Refrigerator
• Lighting fixture
• Washing machine
• Compact fluorescent light bulb
• Microwave oven
Home Electronics
• Television
• VCR
• Audio product
Building Materials
• Window
• Door
• Skylight
• Insulation
• Roofing material
Q6b
Have you or someone else in your
household been shopping for a newly built
home in the last 12 months?"
~ Yes
~ No
~ Don't know
-------
Q 12(c). Finally, please review the last of the product lists
below. Which of these products have you purchased in the
last 12 months? Please check all that apply.
Building Materials:
Buildings
Door
Roofing material
Window
Skylight
None of these products
Newly built home
Insulation
Q 12(b). Please continue reviewing the lists of products below.
Which of these products have you purchased in the last 12
months? Please check all that apply.
Home Appliances/Lighting:
Home Electronics Dishwasher
Television Refrigerator
VCR Lighting fixture
Audio product Washing machine
Compact fluorescent light bulb Microwave oven
None of these products
Q 12(a). Please look at each of the groups of products again.
Which of these products have you purchased in the last 12
months? Please check all that apply.
Heating and Cooling Products:
Central air conditioner
Computer or monitor
Computer printer
Copying machine
Fax machine
Scanner
Home Office Equipment
Furnace or boiler
Heat pump
Thermostat
Room air conditioner
None of these products
-------
No products purchased OR
ES6="No" or "Don't know":
Skip to Q 13a
Products
purchased
Q7: For any of the products you purchased, did you see the
ENERGY STAR label (on the product itself, on the
packaging, or on the instructions)?
"No" or "Don't
Know"
(Skip to Q13a)
Yes
Q7a_l thru Q7a_3: On which products did you see the
ENERGY STAR label?
(show only the products they checked off in Q12,
with options to check for each -> "Saw label" "Did
not see label" "Don't know")
C-7
-------
Q8. For any ENERGY STAR-labeled product(s) you purchased,
how much did the presence or absence of the ENERGY STAR
label influence your purchasing decision?
Very much
Somewhat
Slightly
Not at all
Don't know
Q9. Did you receive rebates or reduced-rate financing for any
ENERGY STAR-labeled product(s) you purchased?
"No" or "Don't
Know"
(SkiptoQll)
Q10. If rebates or reduced-rate financing had not been available,
how likely is it that you would have purchased the ENERGY
STAR-labeled product?
Very likely
Somewhat likely
Slightly likely
Not at all likely
Don't know
-------
Qll. How likely are you to recommend ENERGY STAR-
labeled products to a friend?
Very likely
Somewhat likely
Slightly likely
Not at all likely
Don't know
Q13a. Now, please think only about Heating and Cooling Products. Please select
the source(s) of information you are most likely to use to obtain information about
this product type. Please mark all that apply.
Heating and Cooling Products:
Consumer Reports and other product-oriented
Magazines
Advice from retailers or salespersons
Advice from a friend, neighbor, relative, or
co-worker
Television
Electric or gas utility
Other magazines
Newspapers
Advice from contractors
Radio
Internet
Other
Don't know
Q13b. Now, please think only about Home Appliances / Lighting / Home
Electronics. Please select the source(s) of information you are most likely to use
to obtain information about this product type. Please mark all that apply.
Home Appliances / Lighting / Home Electronics:
Consumer Reports and other product-oriented
Magazines
Advice from retailers or salespersons
Advice from a friend, neighbor, relative, or
co-worker
Television
Electric or gas utility
Other magazines
Newspapers
Advice from contractors
Radio
Internet
Other
Don't know
-------
I
We're interested in how you view companies that produce or sell ENERGY STAR-labeled products. Below are
phrases that some people have used to describe these companies. On the scale by each word or phrase, please
indicate how strongly you agree or disagree with each.
Strongly
Disagree
Somewhat
Disagree
Neither Agree
Nor Disagree
Somewhat
Agree
Strongly
Agree
Q14a. Responsible
1
2
3
4
5
Q14b. Uncaring
1
2
3
4
5
Q14c. Trustworthy
1
2
3
4
5
Q14d. Industry leader
1
2
3
4
5
Q14e. Impractical
1
2
3
4
5
Q14f. Innovative
1
2
3
4
5
Q14g. Behind the times
1
2
3
4
5
Go to demographic and closing questions.
c-io
------- |