Spitzer's contribution to the AGN population

Infrared selection is a potentially powerful way to identify heavily obscured AGNs missed in even the deepest X-ray surveys. Using a 24 μm-selected sample in GOODS-S, we test the reliability and completeness of three infrared AGN selection methods: (1) IRAC color-color selection, (2) IRAC power-law selection, and (3) IR-excess selection; we also evaluate a number of IR-excess approaches. We find that the vast majority of non-power-law IRAC color-selected AGN candidates in GOODS-S have colors consistent with those of star-forming galaxies. Contamination by star-forming galaxies is most prevalent at low 24 μm flux densities (~100 μJy) and high redshifts (z ~ 2), but the fraction of potential contaminants is still high (~50%) at 500 μJy, the highest flux density probed reliably by our survey. AGN candidates selected via a simple, physically motivated power-law criterion ("power-law galaxies," or PLGs), however, appear to be reliable. We confirm that the IR-excess methods successfully identify a number of AGNs, but we also find that such samples may be significantly contaminated by star-forming galaxies. Adding only the secure Spitzer-selected PLG, color-selected, IR-excess, and radio/IR-selected AGN candidates to the deepest X-ray-selected AGN samples directly increases the number of known X-ray AGNs (84) by 54%-77%, and implies an increase to the number of 24 μm-detected AGNs of 71%-94%. Finally, we show that the fraction of MIR sources dominated by an AGN decreases with decreasing MIR flux density, but only down to f_24 μ m = 300 μJy. Below this limit, the AGN fraction levels out, indicating that a nonnegligible fraction (~10%) of faint 24 μm sources (the majority of which are missed in the X-ray) are powered not by star formation, but by the central engine. The fraction of all AGNs (regardless of their MIR properties) exceeds 15% at all 24 μm flux densities.

Bayesian adaptive estimation of arbitrary points on a psychometric function

Bayesian adaptive methods have been extensively used in psychophysics to estimate the point at which performance on a task attains arbitrary percentage levels, although the statistical properties of these estimators have never been assessed. We used simulation techniques to determine the small-sample properties of Bayesian estimators of arbitrary performance points, specifically addressing the issues of bias and precision as a function of the target percentage level. The study covered three major types of psychophysical task (yes-no detection, 2AFC discrimination and 2AFC detection) and explored the entire range of target performance levels allowed for by each task. Other factors included in the study were the form and parameters of the actual psychometric function Psi, the form and parameters of the model function M assumed in the Bayesian method, and the location of Psi within the parameter space. Our results indicate that Bayesian adaptive methods render unbiased estimators of any arbitrary point on psi only when M=Psi, and otherwise they yield bias whose magnitude can be considerable as the target level moves away from the midpoint of the range of Psi. The standard error of the estimator also increases as the target level approaches extreme values whether or not M=Psi. Contrary to widespread belief, neither the performance level at which bias is null nor that at which standard error is minimal can be predicted by the sweat factor. A closed-form expression nevertheless gives a reasonable fit to data describing the dependence of standard error on number of trials and target level, which allows determination of the number of trials that must be administered to obtain estimates with prescribed precision.