You are here:
POWER ANALYSIS OF LINEAR RELATIONSHIPS BETWEEN PERSONAL EXPOSURE AND INDOOR/OUTDOOR PM 2.5 CONCENTRATIONS AT BALTIMORE AND FRESNO
Citation:
Suggs, J C., R W. Williams, AND J P. Creason. POWER ANALYSIS OF LINEAR RELATIONSHIPS BETWEEN PERSONAL EXPOSURE AND INDOOR/OUTDOOR PM 2.5 CONCENTRATIONS AT BALTIMORE AND FRESNO. Presented at PM 2000 AWMA Conference, Charleston, SC, January 24-28, 2000.
Description:
The National Exposure Research Laboratory is currently in the process of conducting panel studies to investigate personal exposure to particulate matter(PM). One of the primary goals of PM exposure studies is to establish mathematical relationships between personal exposure and indoor/outdoor concentrations of particle matter in the 0 - 2.5 um size range (PM2.5). Linear regression relationships were fitted using data from studies conducted at sites in Baltimore MD, and Fresno CA. The coefficient of determination R2 was calculated from the log likelihood ratio of restricted and unrestricted maximum likelihood estimations using the SAS mixed models procedure and allowing for first order auto-correlation in the residuals. The power of a statistical test is the probability of seeing a statistically significant relationship when it exists and is inextricably tied to the the significance level (a or Type T Error), and the sample size which, in this case, is the number of days of sampling. By using R2 to estimate the effect size and applying power analysis, we were able to determine the required sample sizes to achieve statistically significant results for the a = .05 and .10 significance levels for tests of the null
hypothesis Ha: R2 >0 versus several alternative hypotheses between zero and 1.0. This paper examines the results with a retrospective look using power calculations (post hoc) to see how well we did and what adjustments should be made to sample sizes for future exposure studies. Power curves are plotted showing the interrelationship between R2, N, and the power for
the a=.05 and. 10 significance levels.
This paper has been reviewed in accordance with the U. S. Environmental Protection Agency's peer and administrative review policies and approved for presentation and publication. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.