Variance Modeling of Smoking-related EMA Data

描述（由申请人提供）：正如项目公告# PAR-08-213“行为与社会科学中的方法与测量”中所述，需要“开发适当的分析技术，用于行为与社会科学研究的新数据和新方法”。该提案旨在解决对日记或生态瞬时评估（EMA）方法产生的数据的需求。在吸烟和癌症研究中使用EMA方法已成为一种新的和重要的方法。EMA研究的数据本质上是多层次的，例如（1级）观察嵌套在（2级）天和（1级）受试者内。因此，线性混合模型（lmm，又称多层或分层线性模型）越来越多地用于分析EMA数据。在EMA研究中，每个受试者有多达30或40个观察结果并不罕见，这比传统的纵向数据lmm允许的建模机会更大。特别是，一个非常有前途的扩展方法是将方差建模为协变量的函数，除了它们对总体平均水平的影响之外。例如，如果吸烟者的情绪是结果，那么人们可以考虑协变量对他们情绪水平的影响（例如，他们平均有多高兴/悲伤），以及他们情绪的变化（例如，他们的情绪有多不稳定/不稳定）。或者，当一个人吸烟时，可以从均值（情绪是否改善？）和方差（情绪是否稳定？）的角度来检查情绪的变化，以及哪些变量可能与吸烟相关的情绪水平和变化的变化有关。因此，通过允许主体内方差是协变量的函数，我们可以更直接地检验吸烟有助于调节情绪的假设。因此，本研究的目标是：(1)开发可访问的软件，用于EMA数据的均值和方差的一般三层建模；(2)利用我们的项目资助“青少年吸烟模式的社会和情感背景”（NCI资助#PO1 CA98262）的数据，研究吸烟对青少年情绪调节的作用，该项目建立了一个吸烟和尼古丁依赖高风险青少年队列。这项研究有可能在分析EMA数据和理解情绪变化与吸烟依赖之间的关系方面做出显着的方法和实质性贡献。这些方法可以很容易地推广到各种与癌症相关的研究领域，包括疼痛和症状的评估，以及饮食和运动。公共卫生相关性：在吸烟和癌症研究中使用生态瞬时评估（EMA）方法已经成为一种新的和重要的方法，允许检查吸烟相关现象随着时间的推移而发生。正如项目公告# PAR-08-213，行为和社会科学的方法论和测量所指出的那样，需要“开发适当的分析技术，用于行为和社会科学研究的新数据和新方法。”本提案旨在通过开发EMA数据的统计方法和软件来解决这一需求，EMA数据包括在几天内和受试者内嵌套的观察结果，允许对平均水平（变量是否始终较高或较低）和变异水平（变量是否更不稳定或不稳定）的影响。随着对变异的新关注，拟议的研究将检查以前在吸烟研究中无法解决的问题。

DESCRIPTION (provided by applicant): As noted in Program Announcement # PAR-08-213, Methodology and Measurement in the Behavioral and Social Sciences, there is a need for "developing appropriate analytic techniques for use with new kinds of data and new approaches to behavioral and social science research." This proposal is aimed at addressing this need for data generated from diary or Ecological Momentary Assessment (EMA) methods. Use of EMA methods in smoking and cancer research has become a new and vital approach. Data from EMA studies are inherently multilevel in nature with, for example, (level-1) observations nested within (level-2) days and (level-1) subjects. Thus, linear mixed models (LMMs, aka multilevel or hierarchical linear models) are increasingly used for analysis of EMA data. In EMA studies, it is not unusual for there to be up to thirty or forty observations per subject, and this allows greater modeling opportunities than what conventional LMMs for longitudinal data allow. In particular, one very promising extended approach is the modeling of variances as a function of covariates, in addition to their effect on overall mean levels. For example, if a smoker's mood is the outcome, then one can consider the effect of covariates on their mood level (e.g., how happy/sad are they on average), as well as on their variation in mood (e.g., how labile/erratic is their mood). Or, one can examine mood changes when a person smokes in terms of the mean (does mood improve?) and variance (does mood stabilize?), and what variables might be related to those smoking-related changes of mood level and variation. Thus, by allowing within-subject variance to be a function of covariates, we can more directly examine the hypothesis that smoking helps to regulate mood. Thus, the aims of the proposed study are to (1) develop accessible software for general 3-level modeling of means and variances of EMA data; and (2) examine the role of smoking on mood regulation in adolescents using data from our program project grant, "Social and Emotional Contexts of Adolescent Smoking Patterns" (NCI grant #PO1 CA98262), which established a cohort of adolescents at high risk for the development of smoking and nicotine dependence. This study has the potential to make notable methodological and substantive contributions for analysis of EMA data and understanding the relationship between mood variation and smoking dependency. These methods can easily generalize to a variety of cancer -relevant research areas, including the assessment of pain and symptoms, as well as diet and exercise. PUBLIC HEALTH RELEVANCE: Use of Ecological Momentary Assessment (EMA) methods in smoking and cancer research has become a new and vital approach, allowing for the examination of smoking-related phenomena as they happen over time. As noted in Program Announcement # PAR-08-213, Methodology and Measurement in the Behavioral and Social Sciences, there is a need for "developing appropriate analytic techniques for use with new kinds of data and new approaches to behavioral and social science research." This proposal is aimed at addressing this need by developing statistical methods and software for EMA data consisting of observations nested within days and subjects, allowing for effects on both average levels (is the variable consistently higher or lower) and levels of variation (is the variable more labile or erratic). With the new focus on variation, the proposed research will examine previously un-addressable questions in smoking research.