Workshop: Intensive, interdisciplinary short courses on Bayesian inference for ecologists

环境生物学的进步需要应用统计方法来从模型和数据中获得理解。模型用数学术语描述关于环境过程的科学假设。统计允许研究人员评估其数据的价值，作为支持或反驳这些模型中表达的假设的证据。这一程序是所有科学探究的基础，但许多处于职业生涯各个阶段的现代研究人员缺乏自信地应用这一程序所需的训练。该奖项将为研究生、博士后、大学教师和机构研究人员提供贝叶斯统计建模方面的强化培训。参与者将积极从科学界代表性不足的群体中招募。这项培训将提高参与者推进科学和解决社会面临的复杂环境问题的能力，并将促进生态学家和统计学家之间的合作。培训工作坊将教授理解贝叶斯推理所需的基本统计理论，训练参与者为贝叶斯模型编写适当的数学表达式，使参与者能够使用贝叶斯方法解决广泛的分析问题，并传播材料以加强其他人的教学。培训将强调理解贝叶斯建模的基本原理，通过讲座和实验室练习，每年提供三个为期两周的短期课程。课程涵盖的主题包括概率论和分布理论，贝叶斯定理及其与最大似然的关系，编写简单和分层贝叶斯模型，马尔可夫链蒙特卡罗算法，以及模型检查。专业的、高级的主题将被选择来匹配课程参与者的特定兴趣。未来将鼓励参与者和统计学家之间的合作，包括六名统计研究生在课程中协助解决问题。培训将提高环境生物学学科的智力资本。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Progress in environmental biology requires application of statistical methods to gain understanding from models and data. Models describe scientific hypotheses about environmental processes in mathematical terms. Statistics allow researchers to assess the value of their data as evidence supporting or refuting the hypotheses expressed in those models. This procedure is fundamental to all scientific inquiry, but many modern researchers at all stages of their careers lack the training needed to apply it confidently. This award will provide intensive training in Bayesian statistical modeling to graduate students, post-docs, university faculty, and agency researchers. Participants will be actively recruited from groups underrepresented in science. This training will enhance participants’ ability to advance science and to solve complex, environmental problems confronting society and will foster collaborations between ecologists and statisticians.The training workshops will teach basic statistical theory needed to understand Bayesian inference, train participants to write proper mathematical expressions for Bayesian models, enable participants to use Bayesian methods to solve a wide range of analysis problems and disseminate materials to enhance teaching by others. The training will emphasize understanding basic principles of Bayesian modeling using lectures and laboratory exercises during three, annually offered short courses spanning two weeks each. Topics to be covered include probability and distribution theory, Bayes theorem and its relationship to maximum likelihood, writing simple and hierarchical Bayesian models, the Markov chain Monte Carlo algorithm, and model checking. Specialized, advanced topics will be chosen to match the specific interests of course participants. Future collaborations between participants and statisticians will be encouraged by including six statistics graduate students to assist with problem solving during courses. The training will enhance the intellectual capital of the disciplines of environmental biology.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.