g., nicotine dependence, craving, self-efficacy). However, deciding how to measure these constructs is complicated by the fact selleck screening library that several different scales have been developed to assess each one (see Shadel & Shiffman, 2005), and there is almost no guidance available as to which scale one should choose to assess a specific construct. Most reviews of the smoking assessment literature describe strengths and weaknesses of the scales available to assess a construct but typically stop short of recommending a specific scale (e.g., Shadel & Shiffman, 2005). Findings from the few studies that compare the predictive utility of different scales used to assess particular constructs (e.g., nicotine dependence) are mixed, which complicates efforts to choose scales on purely empirical grounds (e.g.

, Courvoisier & Etter, 2010; Etter, 2008). In any case, given that the results of assessment are not helpful in selecting more effective smoking cessation treatments (Kassel & Yates, 2002), any effort expended over deciding which scale one should choose, particularly for this purpose, may well be wasted. Unless the field is going to abandon assessment (which is unrealistic), a fundamentally different approach to assessment is needed. Ideally, this approach would focus on core constructs, utilize a select set of reliable and validated items designed to assess those constructs, make the items widely available, and offer clear guidance as to which items one should select. PROMIS (Patient-Reported Outcomes Measurement Information System, http://www.nihpromis.org/default.

aspx), part of the National Institutes of Health’s Roadmap initiative, emerged as a solution to the problem of measurement choice more generally (Cella et al., 2007). PROMIS has as goals to develop, validate, and standardize item banks to measure key constructs (e.g., pain, anxiety, alcohol use) relevant to a range of chronic medical conditions (e.g., cancer, depression, arthritis). PROMIS uses modern measurement theory (item response theory; see Edelen & Reeve, 2007) and advances in computer technology so that (a) all items in the item banks have parameters describing their measurement properties enabling the calculation of reliability for any subset of items within a given bafank, (b) all items within a bank are calibrated with respect to the same underlying scale allowing scores based on different sets of items within the bank to be compared, (c) the existence of item banks means that items can be added and deleted as the understanding of each bank’s construct matures over time based on scientific findings, and (d) the comparability of scores within a bank allows for the use of tailored tests and combined with computer-based assessment enables minimization of respondent burden.

Measurement advantages Entinostat (e.g.