In addition, some associations fell short of statistical selleck Gemcitabine significance because of the small sample size. This is the first study to examine test�Cretest reliability of a CPT (to our knowledge) and provides important support for the previous studies using a CPT by suggesting that the observed relationships would be expected to be robust over time. Moreover, these findings converge with another recent study reporting high temporal stability of an APT (Murphy et al., 2009) and other behavioral economic decision-making tasks (Baker et al., 2003; Ohmura et al., 2006; Takahashi et al., 2007). Both the current study and the previous APT study found the highest reliability coefficients for intensity and Omax, although the breakpoint and elasticity coefficients were acceptable.

With regard to elasticity, this is probably because the �� parameter is a derived value and although the exponential model provides a very good fit to the data, it is almost always imperfect (i.e., R2 < 1.0), making the coefficient less likely to be stable over time. However, as the only demand index that empirically quantifies demand across the aggregated choices, it is quite distinct from the other indices and may provide important incremental information. A notable area not addressed in the current study or previous studies has been the relationship between demand indices and biomarkers of nicotine dependence. As purchase tasks depend on self-reported estimates, it is all the more important to examine and establish their relationship to objective indicators of smoking involvement, including CO and other biomarkers such as cotinine.

Although CO was collected in the current study, it is relatively sensitive to recency of exposure, which was not controlled, and the CO assessment took place at varying times during the session. Taken together, these factors invalidated systematic considerations of CO in relation to the demand indices, but this is clearly a priority for future studies. There are also several limitations to the current study that similarly point to priorities for future studies. Chief among these are the use of a relatively small sample and a relatively brief interval between the first and second assessment. With regard to the former, small sample size particularly pertains to the correlations beyond those pertaining to test�Cretest reliability.

Here, substantial caution should be exercised because even large associations fell short of statistical significance at times and the small sample size may have substantially influenced the findings. With regard to the latter, future studies validating the test�Cretest reliability of the CPT over Anacetrapib longer periods of time will be important. In addition, most assessments exhibit some decrement in reliability as time intervals increase, even for putatively long-standing characteristics, such as personality (e.g., Block & Block, 2006).