ability to analyze quantitative data has far outstripped our ability to measure the constructs we aim to study. Researchers now routinely use sophisticated latent variable models, hierarchical linear models, and hazard rate models in their work. Unfortunately, as the articles in this special issue reveal, there are serious, fundamental problems with the measures on which these techniques rely. Although there are a number of problems with the measures we use, the articles in this issue are concerned primarily with the lack of measurement invariance across different racial and ethnic groups. Simply put, there is convincing evidence that quantitative study measures do not always mean the same thing to people in different racial and ethnic groups.1 When answers to quantitative survey items are perceived and understood in different ways, subsequent analysis may amount to little more than comparing apples with oranges. If the goal is to better understand health-related issues in our multicultural society, then it is imperative that we devote significant effort to improving the quality of quantitative measures. The reason for doing so was provided some time ago by Bohrnstedt2 who argued that, Measurement is a sine qua non of any science (p. 69). Without sound measures, high-powered quantitative statistical procedures are conducted in vain. The articles in this issue provide valuable guidelines on how the quality of quantitative survey items may be improved so that they capture and reflect the diverse cultural perspectives found in our nation. A primary emphasis in each article is placed on using qualitative methods for this purpose. The goal of this commentary is to review each article with an eye toward elaborating and extending some of the points that have been made. Specific comments will be provided through review of each article in succession. After this, more general observations are made on the tasks that lie ahead. The article by Johnson3 provides a useful framework for organizing our thinking on how quantitative survey items may be improved with qualitative techniques. He proposes 2 ways of categorizing our efforts in this regard: procedures geared primarily toward establishing interpretive forms of equivalence and procedures aimed at determining procedural forms of equivalence. Interpretive equivalence is concerned with assessing cultural differences in the meaning of survey items and is primarily qualitative in nature, whereas procedural equivalence focuses on the application of sophisticated quantitative procedures to assess crosscultural equivalence such as confirmatory factor analysis and measurement models based on item response theory. Johnson3 correctly points out that both are needed to create the best possible quantitative survey items. There are 2 ways to elaborate and extend the important points made by Johnson.3 First, he advocates the use of multiple methods, ie, the combination of qualitative and quantitative procedures. This is fine, but the same reasoning can be applied to the use of qualitative methods alone. Johnson3 identifies a range of qualitative procedures that can be used to improve measurement quality, including focus groups, input from experts, and ethnographic approaches. However, the quality of measures is likely to be maximized when not one, but