【英語論文の書き方】第55回　実験計画について論文翻訳・英文校正　研究者専門の翻訳会社　ワールド翻訳サービス

【英語論文の書き方】第55回　実験計画について

2018年4月26日 10時00分

第54回では次の３つのテーマを取り上げました。
(1) Using hyphens in compound adjectives
複合形容詞のハイフンについて

(2) Spaces between sentences and in mathematical expressions
文と文の間に入れるスペース、数字と数学記号や単位の間に入れる
スペースについて

(3) Checking the “Instructions to authors” of the relevant journal before you start preparing your paper
「投稿規定」について

第55回（今回）のテーマは実験計画です。
ジャーナルに投稿する際の大前提として押さえるべきポイントをカバーしました。

具体的には
・実験条件の管理
・サンプルサイズ
・三角測量
・分析したデータの確認
・まとめ
について取り上げています。

見落としている方も多いポイントとのことなので、ぜひお時間を作ってご覧ください！

Designing effective research By Geoffrey Hart

One of the biggest problems I encounter in the journal manuscripts that I edit relates to the experimental design. Often, authors obtain results that are not statistically significant, even though it seems likely that the differences that they observed between treatments were large and significant. When we study biological systems, there is always a risk that high variability will make it difficult to detect significant differences. But there are proven techniques that can reduce the risk. In this article, I’ll remind you of some of the most important ones.

Control experimental conditions

One problem with field research is that so many factors can affect the outcomes, particularly for living organisms. This is why many researchers prefer laboratory research, because they control the variables. If it’s necessary to limit variability in the field, the first step is to review the literature to identify the factors that can affect the outcomes.
Understanding these factors lets you choose sites that have similar values of these factors to minimize their impact. It also lets you perform stratified sampling, in which each group of similar properties forms a “stratum”, with reduced variation within the group.

Sample sizes

Field research can be hard; when I did plant physiology research many years ago,　the measurements took long enough that by the time I reached the final plant, it was time to remeasure the first plant. Time and budget constraints encourage many researchers to use a very small sample size; though this makes better use of limited time and money, it often results in a lack of significant differences. Assuming there is no inherent problem that leads to high variation (e.g., a skewed probability distribution that results from some physical process), increasing the sample size will usually help.
One way to increase the likelihood of significance is to examine sample sizes in　the literature for studies similar to yours. Studies where the results appeared different, but the difference was not significant, suggest that you should use a larger sample size.
If you prefer a more mathematical approach, you can predict the required sample size using the reported variances (s2). Wikipedia provides a good overview of this subject
(https://en.wikipedia.org/wiki/Sample_size_determination).
You can also start by determining the “statistical power” (the probability that you　will correctly reject the null hypothesis) of your design.

Triangulation

In surveying “triangulation” means examining the position of a point by examining it from two or more angles and using basic geometry to calculate the position—something that would not be easy from only one position. In research, triangulation means confirming a finding from multiple directions.
For example, to measure gene expression, you can either quantify the protein products of the gene or the quantity of its RNA transcripts. In ecology, you can measure the volume of carbon sequestered by vegetation using either species-specific allometric equations or by random destructive sampling. In both cases, the two results will be correlated, but different. If one test fails to produce a significant result, the other test may provide the desired result.
Even if it doesn’t, identifying the same trend in both tests makes it more likely that a difference exists.

Confirming you can analyze your data

Sometimes an experiment seems to be well designed, but it becomes difficult to perform statistical significance tests for the data it produces. A simple example might be designing an experiment based on the assumption that the data will be normally distributed, when in fact the data follows a highly skewed or non-normal distribution or even a multimodal distribution such as binomial. The literature will often reveal this, so you can ask a statistician for advice on how to handle such data in your experimental design. You can sometimes still find a way to analyze the data, but it’s better to start with appropriate data in the first place. With a little knowledge of the likely variance structure of the phenomenon or organism that you’re studying, and the likely probability distribution of the results, they can help you find an experimental design that maximizes the ability of your study to detect significant results.

Summary

Most students learn these concepts early in their university career, but a surprising number forget them when it comes time to design their own research. I hope that this article will help you avoid this mistake.

無料メルマガ登録

メールアドレス
お名前

これからも約2週間に一度のペースで、英語で論文を書く方向けに役立つコンテンツをお届けしていきますので、お見逃しのないよう、上記のフォームよりご登録ください。

もちろん無料です。

バックナンバー

【英語論文の書き方】第55回 実験計画について 論文翻訳・英文校正 研究者専門の翻訳会社 ワールド翻訳サービス