(Burkhart and Kladder 1991). In the original paper, the average indoor radon concentration of the 12 homes was not statistically different from that of the control homes, as analyzed either by area or in combination. Further analysis of 11 of the 12 homes examined the radon-resistance features of each home. Only five had passive stacks in addition to sealing and other techniques used to limit entry of radon; of these, only three had stacks passing through the heated portion of the homes. The three passive-stack homes averaged 130 Bq m-3 over the 2 days when radon was measured; one home had slightly over 150 Bq m-3. The control homes averaged 450 Bq m-3 over the same period (Kladder and others 1991).
The only other study that has attempted to estimate the average indoor radon concentration with and without radon-resistance construction features was done as part of EPA's new-house evaluation program (Murane 1998). Measurements were made in 148 houses that were built with various radon-resistance features. All but five of the houses were built near Denver and Colorado Springs, CO and did not include a passive stack as part of radon control. The other five, built near Detroit, MI, did include a passive stack. Radon measurements were made in these houses—usually in the basement and in some cases on the first floor as well—with open-face charcoal canisters with a 2-d sampling period. The results were tabulated by ZIP code and compared with measurements in the same ZIP code done as part of the Colorado and Michigan state radon surveys (also done with short-integration time-measurement techniques). The state surveys were done in occupied houses during the winter months, but some of the new-house program measurements were done in other seasons. It is not clear whether the new houses were occupied at the time the indoor radon was measured. Overall, in the 143 Colorado houses, the average basement radon concentration was 190 Bq m-3, compared with 230 Bq m-3 in 94 control houses in the same ZIP codes. In the case of the Detroit-area houses, there were only five radon-resistant houses and four controls; basement concentrations averaged 90 and 50 Bq m-3, respectively.
Comparing the two sets of houses within each ZIP code produces considerable variation in whether the average radon concentration in the radon-resistant houses is lower or higher than of the average measured in the "control" houses. Given that all the results are based on short-term measurements, with some seasonal differences in measurements between the "control" houses and the radon-resistant ones, comparison of the average radon concentrations in the two sets of houses does not provide a strong basis for evaluating the effectiveness of the radon-resistance features. Furthermore, because one is looking for small differences between the average concentrations in the two sets—perhaps a factor of 2 or 3 at most—the study and the measurement techniques will have to be carefully designed and executed to ensure statistically meaningful results.
As has been indicated in the earlier discussion, there are no data on the long-