Sample selection and weighting

These data are drawn from 2019 through 2024 Claritas Energy Behavior Track annual surveys, conducted in partnership with E Source, in which approximately 32,500 residential customers in the US were surveyed on a variety of energy-related topics. The households were chosen from Survey Sampling International (SSI) and Critical Mix (now Dynata), which are leading providers of survey sampling solutions. Recent years included SaberEsPoder/Welcome Technologies and InnovateMR to recruit harder to reach populations. The survey has also been conducted annually since 2009, but this tool focuses only on results from the past five years of data (the study was paused in 2020 due to the COVID-19 pandemic).

The sample is nationally representative across the nine US census divisions, five US age groups, and three ethnicity and race groups. The data also includes between 500 and 1,500 Spanish-language responses per year. A postfieldwork weighting scheme was applied using geographic and demographic data. Note that some sample sizes are small; we cannot guarantee that analyzing data within a smaller area will elicit accurate results that can be generalized to that area’s population.

The utility-level data are based on respondents’ self-reported utility providers. Only utilities with at least 70 respondents in three out of the five survey years appear in the drop-down list. When base sizes fall below this threshold for a subset of respondents, results should be interpreted with caution.

Analysis

All statistical analysis was performed on the weighted sample to ensure that the results were as representative of the US population as possible. Percentages shown in the charts reflect weighted data; sample sizes (n) are based on unweighted data.

Statistical significance testing is available where differences in proportions for column variables are calculated at the 95% confidence level using a Z-test. Statistical significance is represented with the formula p < 0.05. The p-value allows us to assess whether the findings are significantly different from some reference value. For this study, a p-value that is less than 0.05 is the cutoff beyond which we assert that the findings are statistically significant. The statistical significance table shows statistically significant differences from other groups within the same row. Cells with dashes indicate that there are no statistically significant differences from other groups in that row. An asterisk (*) indicates that statistical significance testing was not conducted due to a small sample size.

Trends over time can be accessed within the tool when questions are similar within each survey iteration.

Understanding residential customers

Who are your residential customers? How do they feel about energy efficiency, renewable energy, emerging technologies, smart meters, products and services, and their utility’s brand image? This large-scale survey from Claritas includes more than 300 variables and allows us to provide detailed data and analysis across US Census Regions (figure 1), demographics, household characteristics, energy profiles, and utility program participation.

Figure 1: US Census regions and divisions

The US Census Bureau has organized the country into four regions: West, Midwest, Northeast, and South. Each region is further separated into multiple divisions.

Note: This tool is supported by many browsers, including Internet Explorer version 10 and above, Google Chrome, Mozilla Firefox, and Apple Safari. If you’re having trouble viewing the data visualizations, please check that your browser is one of these.

Choosing a question

In the left sidebar, the View questions tab is selected by default; this view shows a list of all questions in the database, divided into topical categories. You can search by question text or number, or browse the questions by category. Selecting a question will redraw the chart.

Customizing the data view

The “View filters” tab in the left sidebar lists the year, demographic categories, US Census regions, states, and fuel types available to filter by. Once you’ve selected a filter, if you change questions, the filters will persist until you remove them, or until you leave the tool. Each filter you add will reload the chart, adding a new version that shows the data you’ve selected.

By default, we show data for the entire base; you can filter to smaller sets if you prefer. If you don’t want to see the All US chart, uncheck the All US box directly beneath the year buttons on the Filters panel.

Choose as many years as you like; if data is not available for a selected year on the chart you&rquo;re viewing, the button for that year will not be shown. You will have to reselect missing years if you move to another chart.

You can apply one of our sixteen demographic category crosstabs to the data at a time. These defining characteristics of the person who answered the questions include age, income level, or home ownership, among others. Whichever category you select will persist until you remove it or until you leave the tool.

In addition to year and demographic category, you can show the data by geographic area or a specific entity (choose as many as you like):

• Region or state. Choose the US Census regions or states you’d like to see reflected in your data. Click within the box to see a list; you can also begin typing to narrow the options. Select the ones you’d like to see from the drop-down list. Delete an item you no longer want to see by clicking the X beside its name.
• Electric or dual-fuel and gas utilities. The data is segregated by electric or dual-fuel utilities and gas-only utilities; use these two fields to choose specific utilities that you’re interested in. Click within the box to see a list; you can also begin typing to narrow the options. Select the utilities you’d like to see from the drop-down list. Delete an item you no longer want to see by clicking the X beside its name.

Show statistical significance

When a filter is selected, the statistical significance button will appear beneath the chart. Click it to show a table that compares differences between columns; each cell in that table lists the groups that show a statistically significant difference from the selected answer. Statistical significance is based on an Independent Z-test of proportions at the 95% confidence level.

When exporting to Microsoft Excel, the statistical significance table will be included whenever you have a demographic category selected, even if it isn’t showing on screen; it will only be exported to PDF if the table’s open.

Export your results

You can export your filtered data chart to Excel for further analysis, to PDF to share, or to a PNG image file for use in internal presentations. Click the download button at the top right of any chart to select which format to export; only individual pages are available to export, though all filters applied will be included.

• Plot (pdf). Creates a PDF of the chart, including the question text and any notes. Does not include statistical signficance data tables.
• Plot (png). Creates a PNG image file of the chart, including the question text, and any notes. Does not include the statistical significance table.
• Data (xlsx). Creates an Excel spreadsheet of the data used to build that chart, including the question text, any notes, and the statistical significance data table.

If you need further assistance, call us at 1-800-ESOURCE (1-800-376-8723) Monday through Friday, 9:00 a.m. to 5:00 p.m. MT, or contact customer service.