文章基本信息

标题：Beyond Cowboy Science: Simple Methods for Conducting Credible and Valid Research
本地全文：下载
作者：Diane L. Haase
期刊名称：Tree Planters' Notes
印刷版ISSN：0096-8714
出版年度：2014
卷号：57
期号：2
页码：32
出版社：US Department of Agriculture, Forest Service
摘要：Many nursery and field trials are conducted every year to test new products and techniques. Some of these trials, however, can produce data that are too variable or confounded to accurately assess the question(s) of interest. A “cowboy science” approach can yield results that are statistically invalid and biologically untrue; using such data can lead to erroneous conclusions. By incorporating a few basic principles of study design and data collection, anyone can yield credible data that can be used to answer questions or make decisions. Despite beliefs to the contrary, using a valid experimental design usually requires little or no additional input of time and resources, nor does it require an in depth understanding of statistics. Good study design also ensures that the time and resources invested in research yields meaningful results. This article describes the “Three Rs” of study design—representation, replication, and randomization—along with examples of pitfalls and successes. It also describes how to create a study plan to guide effective research in the nursery or the field. An earlier version of this article was published in the 2013 National Proceedings: Forest and Conservation Nursery Associations (Haase 2014).