Each Housing Characteristic table is presented in two ways:

a. Number of Households-Counts of millions of U.S. households.

b. Percent of Households-Percent of U.S. households (available only on our Web site).

With the exception of the Household Demographics and Usage Indicators tables, which include the suffix “u” and present both total and per-household consumption and expenditures data, each Consumption and Expenditures table is presented in two ways:

c. Consumption-Fuels in terms of Btu and physical units.

e. Expenditures-Fuels in terms of U.S. dollars.

Explanation of the Numbering Scheme

The table numbering scheme was designed to permit easy navigation through the many tables. The following example describes the meaning of each of the components of the table numbers and the hierarchical scheme.

Table CE3-2e. Electric Air-Conditioning Energy Expenditures in U.S. Households by Year of
Construction, 1997

CE

CE3

The first two letters, CE, identifies the major type of table, either Housing Characteristics (HC) or, as in this example, an Energy Consumption and Expenditures (CE) table.

The third digit, 3, identifies the topical group of the table. In this example, the 3 identifies this as a
Consumption and Expenditures (CE), Electric Air-Conditioning (3) table.

CE3-2 The fourth digit that follows the dash, 2, identifies the table heading within the topical group. In this example, the 2 identifies this as a table presenting Consumption and Expenditures (CE), Electric Air-Conditioning (3), Year of Construction (-2) data.

CE3-26 Finally, the fifth digit, e, identifies the type of data presented in the table. In this example, the e identifies this as a table presenting Energy Consumption and Expenditures (CE), Electric Air-Conditioning (3), Year of Construction (-2), Energy Expenditures (e) data. The alternative suffix would be a c, identifying the data as Energy Consumption data.

Energy Information Administration

Row and Column Factors

The tables provide row factors in the far-right column and column factors on the top line of each table. These row and column factors are a quick and easy way to measure the reliability of the data presented in this report. These factors can be used to determine the Relative Standard Error (RSE) for each estimate, which, in turn, can be used to determine the standard error of the estimate and to determine whether the difference between any two estimates is statistically significant. However, because the RSE's are only approximate, standard errors, confidence intervals, and statistical tests must also be regarded as only approximate.

The table extract below shows that the 1997 per-household expenditures for space heating in housing units constructed in the 1980s, where the main space-heating fuel was electricity, were $225. The RSE Column Factor for this estimate is 1.2 and the RSE Row Factor is 6.3.

Table CE2-2e. Space-Heating Energy Expenditures in U.S. Households by Year of Construction, 1997

To obtain the RSE factor for the $225 electricity estimate, multiply the row factor by the column factor.

Because the estimates presented in the Detailed Data Tables are based on a sample of residential housing units, they are subject to sampling error, or standard error.

To calculate the standard error for the $225 electricity estimate, multiply the estimate by the RSE factor

Energy Information Administration

For each of the estimates given in the tables, a confidence range can be determined with the estimate at the mid-point.

To determine the 95 percent confidence range for the $225 electricity estimate to approximate two standard errors, multiply the standard error calculated above by 1.96.

To determine the bottom of the confidence range around the $225 estimate, subtract $33 from the estimate.

To determine the top of the confidence range around the $225 estimate, add $33 to the estimate.

Therefore, in this example, the value for the estimate for expenditures per household for electricity for space-heating in housing units constructed in the 1980s, where the main space-heating fuel was electricity, would fall between $192 and $258, 95 percent of the time.

Statistical Significance Between Two Statistics

The difference between any two estimates given in the detailed tables may or may not be statistically significant. Statistical significance for the difference between two independent variables is computed as:

where S is the standard error, x, is the first estimate, and x, is the second estimate. The result of this computation is to be multiplied by 1.96, and if this result is less than the difference between the two estimates, the difference is statistically significant.

For example, in the 1997 RECS, in housing units constructed in the 1980s, $225 per household were spent for electricity for space heating where the main space-heating fuel was electricity, and $357 per household were spent for natural gas for space heating where the main space-heating fuel was natural gas, an estimated difference of $132. The standard error for the $225 estimate (x,) is $17.01, and the standard error for the $357 units estimate (x2) is $17.14:

Multiplying $24.15 by 1.96 yields $47. Because the estimated difference of $132 is greater than $47, the difference between the 1997 electricity and natural gas space heating expenditures estimates is statistically significant.

Energy Information Administration

1997 RECS Household Characteristics Detailed Data Tables Reference Guide

b = Percent of U.S. Household

Note: Percent of U.S. Households (b) tables and those in the shaded areas above can be accessed at http://www.eia.doe.gov/emeu/consumption.

1997 RECS Consumption and Expenditures Detailed Data Tables Reference Guide

c = Energy Consumption

e = Energy Expenditures

u = Household Demographics/Energy Usage Indicators

Note: Tables in the shaded areas above can be accessed at http://www.eia.doe.gov/emeu/consumption.