Abstract

This wiki page, a work in progress, describes best practices for documenting research with the goal of enabling others to understand that work and to reproduce experiments / results, and assess the data resulting from that research.

Disclaimers

This document is intended to be a collaborative effort, drawing on the expertise of many contributors. Because the goal is accuracy and adherence to sound scientific methods, not all perspectives may survive the peer-review process.

Overview

In general, scientific research should be well documented. The process does not dictate the document timeline but it does require accuracy and sufficient detail for others to clearly understand and duplicate that research (assuming they have the skills and resources available.) The final document should include 1) an Abstract, 2) a discussion of measurements, materials, procedures, observations and techniques, 3) details about the data, accuracy, resolution, etc., 4) interpretation and analysis of the observations and data which includes ownership of it's limitations and 5) discussion of conclusions which have been proven by the observations.

When any or all of these basic fundamentals are missing or lacking in detail or scope, the author is effectively asking the reader to take what they say on the basis of assumption, not the scientific method based on facts and observable measurements. Resolving such omissions can be resolved either by clarification and added detail / data or by retraction until additional work can be performed which supports the prior assumption.

Abstracts

Abstracts should generally be just a short paragraph, at the top of the document, and should quickly outline 1) the focus of the work, 2) the reason for the work, 3) the nature of the experiments and observations and 4) a broad hint at the conclusions derived from that work.

Methods, Procedures and Protocols

Methods, procedures and protocols describe the processes used by the author to perform the experiments which directly relate to the research being described. A method is the more general description of performing a task; eg. assembling kites from paper, dowels, glue and string. A procedure is more specific to the sequence of events and the expected result; eg. the cut, fold and glue sequence for paper, wood and string to make a box kits. A protocol is a strict definition of a procedure which, if followed precisely, will produce the same results (within given tolerances) every time; eg. the exact same kite design, shape, size and aerodynamic properties every time.

All three can be used, even within the same research, but it is best to be clear on which is being described by choosing the terminology carefully; for example, never call a method a protocol.

Materials

Within this context, the term 'Materials' must reference anything and everything from chemicals, to plastics, to optics, to lasers and LEDs to electronics and even data. In short, materials is just the large umbrella of the components which are often combined in the process of creating or observing something new. While chemicals, plastics and LEDs seem obvious, data may appear out of place. However, data often represents the result of combining physical materials under some other circumstance, to be re-applied to the new research; eg. calibration data of a reference light source.

The importance of discussing "Materials" is in understanding the impact of how the materials influence the results of the research which is being described. All materials can, in one way or another, be described by specifications. Without specifications, the research results will be questionable at best.

Specifications

Specifications (specs) describe the nature of a "material" which, in turn, helps describe how that material will interact with other materials. If you don't know the specs of your chemical sample and you don't know the specs of your sample container, the container might dissolve and leave you with a mess. If you don't know the specs of your UV LED and then try to power it from your laptop USB port, you might damage your laptop.

However, besides just the potential for damage, not knowing and not reporting the material specs (at least where applicable) places your work and your conclusions in question. For example, if your results show a 10% rise in river water temperature but you don't know, and report, that your thermal detector has a +/- 10% drift with battery voltage, your conclusion that the river is warming could be completely wrong. In addition, without properly reporting and handling specs, your work may negatively impact the work of others who rely on your results.

Measurements

The topic of measurements is integral to the topic of specifications. Measurements are simply the recording of what a detection device is reporting -- raw data in many instances. However, raw data alone is not sufficient to achieve a meaningful measurement of some parameter. A meaningful measurement is only achieved by including the measurement devices specifications and by performing the measurement either at a time or over a time period which can be validated as representing the parameter you are attempting to observe. Such a measurement also requires that the device performing the measurement is not influencing the parameter you are trying to measure.

Data

Analysis

Conclusions