Introduction to Data Tree

The continued, available for use, ongoing usability of a digital resource, retaining all qualities of authenticity, accuracy and functionality deemed to be essential for the purposes the digital material was created and/or acquired for. Users who have access can retrieve, manipulate, copy, and store copies on a wide range of hard drives and external devices.

- CASRAI Dictionary

Information collected primarily for administrative, and not research purposes. It includes profiles and curriculum vitae of researchers, the scope and impact of research projects, funding, citations, and research outcomes. This type of data is collected by government departments and other organisations for the purposes of registration, transaction and record keeping, usually during the delivery of a service. These data are also recognized as having research value.

- CASRAI Dictionary

Data that are expressed in a summary form (e.g., summary statistics)

- CASRAI Dictionary

The bringing together of elements. Types of aggregations differ by the nature of the processes by which elements are brought together and the reason understood for aggregating or contained as a unit. Aggregations differ in the nature of relations between the member parts.

- CASRAI Dictionary

A computable set of steps to achieve a desired result.

- CASRAI Dictionary

• H0,that the two means are not equal, i.e. there is a difference between the two types of year..

A research data lifecycle stage that involves the techniques that produce synthesised knowledge from organised information. A process of inspecting, cleaning, transforming, and modelling data with the goal of highlighting useful information suggesting conclusions, and supporting decision making. Data analysis has multiple facets and approaches, encompassing diverse techniques under a variety of names, in different business, science, and social science domains.

- CASRAI Dictionary

A facility providing IT services, such as servers, massive storage, and network connectivity.

Many data centres act as a data repository for certain types of research data.

Data cleaning is a continuous process that requires corrective actions throughout the data lifecycle. Data cleaning is the process of detecting and correcting corrupt or inaccurate records from a dataset. Data cleaning involves identifying, replacing, modifying, or deleting incomplete, incorrect, inaccurate, inconsistent, irrelevant, and improperly formatted, data. Typically, the process involves updating, correcting, standardising, and de-duplicating records to create a single view of the data, even if they are stored in multiple disparate systems. 

- CASRAI Dictionary

The most important thing to realise about data cleaning is that it is not just a one-time activity. Cleaning can (and should!) occur at every stage of the research data lifecycle.

A managed process, throughout the data lifecycle, by which data/data collections are cleansed, documented, standardised, formatted and inter-related. This includes versioning data, or forming a new collection from several data sources, annotating with metadata, adding codes to raw data (e.g., classifying a galaxy image with a galaxy type such as "spiral"). Higher levels of curation involve maintaining links with annotation and with other published materials. Thus a dataset may include a citation link to publication whose analysis was based on the data. The goal of curation is to manage and promote the use of data from its point of creation to ensure it is fit for contemporary purpose and available for discovery and re-use. For dynamic datasets this may mean continuous enrichment or updating to keep it fit for purpose. Special forms of curation may be available in data repositories. The data curation process itself must be documented as part of curation. Thus curation and provenance are highly related.

 - CASRAI Dictionary

A data dictionary, at its simplest, is a list and description of every variable within a dataset, including information such as the units of measurement and what the variable represents.

For more complex datasets, including multi-level or larger database structures, the data dictionary also includes descriptions of the relationships between tables, and for categorical data with a pre-defined set of possible options (e.g. an "enum" datatype in SQL, data coming from a "select" question in a survey, or other defined data such as days of the week), the data dictionary should also include the list of all possible values.

This term defines the practise of analysing large volumes of data, seeking possible relationships without any pre-defined hypothesis or goal. 

This practise is sometimes employed when working with 'big data' and often leads to premature conclusions. When working with big data, it is very easy to find statistically significant results, but it is much harder to spot if those results are actually meaningful, especially without a specific hypothesis. 

Data dredging is sometimes described as "seeking more information from a dataset than it actually contains" (CASRAI).

Data exploration, often called exploratory analysis, uses descriptive statistical methods to learn about and understand the characteristics of a dataset. 

This includes exploring measures of central tendency (e.g. mean, median), measures of spread (standard deviation, range, variance). It also might include exploring the structure of the data, for example splitting the dataset by a categorical variable, or creating visualisations to view the data in different ways. 

This stage of analysis is often where a lot of data cleaning happens, as you can often spot missing data or outliers during this process. 

The layout of a file in terms of how the data within the file are organised. A program that uses the data in a file must be able to recognise and possibly access data within the file. A particular file format is often indicated as part of a file's name by a filename extension (suffix). Conventionally, the extension is separated by a period from the name and contains three or four letters that identify the format. 

There are as many different file formats as there are different programs to process the files. Examples include: Word documents (.doc), Web text pages (.htm or .html), Web page images (.gif and .jpg), Adobe Postscript files (.ps), Adobe Acrobat files (.pdf), Executable programs (.exe), Multimedia files (.mp3 and others). 

Preferred formats are those designated by a data repository for which the digital content is maintained. If a data file is not in a preferred format, a data curator will often convert the file into a preferred format, thus ensuring that the digital content remains readable and usable. Usually, preferred formats are the de facto standard employed by a particular community. 

- CASRAI Dictionary

1. In the context of data and network security: The assurance that information can only be accessed or modified by those authorized to do so. 

2. In the context of data quality: The assurance the data are clean, traceable, and fit for purpose.

- CASRAI Dictionary

A formal statement describing how research data will be managed and documented throughout a research project and the terms regarding the subsequent deposit of the data with a data repository for long-term management and preservation.

 - CASRAI Dictionary

The process of analysing multivariate datasets using pattern recognition or other knowledge discovery techniques to identify potentially unknown and potentially meaningful data content, relationships, classification, or trends. Data mining parameters include: Association (looking for patterns where one event is connected to another event); Sequence or path analysis (looking for patterns where one event leads to another later event); Classification (looking for new patterns); Clustering (finding and visually documenting groups of facts not previously known); Forecasting, or predictive analytics (discovering patterns in data that can lead to reasonable predictions about the future.

- CASRAI Dictionary

The reliability and application efficiency of data. It is a perception or an assessment of dataset's fitness to serve its purpose in a given context. Aspects of data quality include: Accuracy, Completeness, Update status, Relevance, Consistency across data sources, Reliability, Appropriate presentation, Accessibility. Within an organisation, acceptable data quality is crucial to operational and transactional processes and to the reliability of analytics, business intelligence, and reporting. Data quality is affected by the way data are entered, stored and managed. Maintaining data quality requires going through the data periodically and scrubbing it. Typically this involves updating, standardising, and de-duplicating records to create a single view of the data, even if it is stored in multiple disparate systems. 

- CASRAI Dictionary

Data quality assurance (DQA) is the process of verifying the reliability and overall quality of data. 

 - CASRAI Dictionary

It is a process that should ideally be planned in advance, and integrated into your entire project workflow, from creation / sourcing of data, through processing and analysis and to storage and sharing of data.

Data quality checklists can be used to help identify potential issues with a dataset before you begin your exploratory analysis work. 

A data quality checklist is a list of possible issues with a dataset. This list can be created before you start exploring your data to help streamline your data cleaning. If there are physical or logical boundaries that your data should conform to, such as humidity not being above 100%, or age always being 0 or greater, these can form part of your checklist. As such, there is often a strong relationship between a data quality checklist and the data dictionary for your dataset.

1. A storage location for a collection of data that is too valuable to discard, but is only accessed occasionally.

2. An archival service providing the long-term permanent care and accessibility for digital objects with research value. The standard for such repositories is the Open Archival Information System reference model. 

- CASRAI Dictionary

Any organised collection of data in a computational format, defined by a theme or category that reflects what is being measured/observed/monitored. 

- CASRAI Dictionary

A standard way to express a numeric calendar date that eliminates ambiguity, acceptable formats being defined by ISO 8601. ISO 8601 is applicable whenever representation of dates in the Gregorian calendar, times in the 24-hour timekeeping system, time intervals and recurring time intervals or of the formats of these representations are included in information interchange. It includes calendar dates expressed in terms of calendar year, calendar month and calendar day of the month; ordinal dates expressed in terms of calendar year and calendar day of the year; week dates expressed in terms of calendar year, calendar week number and calendar day of the week; local time based upon the 24-hour timekeeping system; Coordinated Universal Time of day; local time and the difference from Coordinated Universal Time; combination of date and time of day; time intervals; recurring time intervals. 

- CASRAI Dictionary

Research data resulting from processing or combining 'raw' data, often reproducible but expensive.

Examples: compiled databases, text mining, aggregate census data.

A name (not a location) for an entity on digital networks. It provides a system for persistent and actionable identification and interoperable exchange of managed information on digital networks. A DOI is a type of Persistent Identifier (PID) issued by the International DOI Foundation. This permanent identifier is associated with a digital object that permits it to be referenced reliably even if its location and metadata undergo change over time.

- CASRAI Dictionary

An initiative to create a digital "library card catalog" for the Web. Dublin Core is made up of 15 metadata elements that offer expanded cataloging information and improved document indexing for search engine programs. The 15 metadata elements used by Dublin Core are:

• title (the name given the resource), 
  
• creator (the person or organisation responsible for the content), 
  
• subject (the topic covered), 
  
• description (a textual outline of the content), 
  
• publisher (those responsible for making the resource available), 
  
• contributor (those who added to the content), 
  
• date (when the resource was made available), 
  
• type (a category for the content), 
  
• format (how the resource is presented),
  
• identifier (numerical identifier for the content such as a URL),
  
• source (where the content originally derived from),
  
• language (in what language the content is written), 
  
• relation (how the content relates to other resources, for instance, if it is a chapter in a book), 
  
• coverage (where the resource is physically located), 
  
• rights (a link to a copyright notice).
  

Computationally intensive, large-scale, networked and collaborative forms of research and scholarship across all disciplines, including all of the natural and physical sciences, related applied and technological disciplines, biomedicine, social science and the digital humanities.

- CASRAI Dictionary

Individual items or records (both digital and analogue) usually obtained by measurement, observation or modelling of the natural world and the impact of humans upon it, including all necessary calibration and quality control. This includes data generated through complex systems, such as information retrieval algorithms, data assimilation techniques and the application of numerical models. However, it does not include the models themselves. 

 - NERC Data Policy

Examples of research data:

• Model output from running a numerical climate model
• Time series logged by environmental instrumentation
• Conductivity-Temperature-Deptch casts from oceanographic cruises
• Groundwater chemistry and stable isotope measurements
• Butterfly abundance observations.

Error is the difference between the measured value and the ‘true value’ (NPL, 1999).  Errors can come from the measuring device itself, including bias, changes due to wear, instrument drift, electrical noise and device resolution.  Other errors can be introduced by difficulties in performing the measurement and by operator skill.  To avoid sampling error, sufficiently dense measurements in space and time should take place to make sure that full variability is captured e.g. diurnal cycles, variations across a site.

Errors can be random or systematic (NPL, 1999).  With random errors, each measurement gives a different result, so the more measurements (of the same thing) the better the estimate and the more certain the measurement becomes.  Systematic errors arise from a bias, e.g., a stretched tape measure, and more measurements do not produce a better estimate of the ‘true value’.

Research data from experimental results, often reproducible, but can be expensive.

Examples: data from lab equipment, metagenomic sequences recovered from soil samples, results of a field experiment.

Data that follows the FAIR principles must be:

• Findable
  
• Accessible
  
• Interoperable
  
• Reusable

FAIR is a set of guiding principles for data management and stewardship designed by stakeholders representing interests in academia, industry, funding agencies and scholarly publishers. The FAIR principles define a set of core enabling conditions which, if fulfilled for a given set of data, would ensure that they remain accessible and re-usable over the long term. 

A key element of these principles is the focus on the use of structured information and persistent identifiers to enable machine discoverability and use of the data.

The full set of principles are published in the article "The FAIR Guiding Principles for scientific data management and stewardship". - DOI: https://doi.org/10.1038/sdata.2016.18

A legal concept that allows the reproduction of copyrighted material for certain purposes without obtaining permission and without paying a fee or royalty. Purposes permitting the application of fair use generally include review, news reporting, teaching, or scholarly research. When in doubt, the quickest and simplest thing may to request permission of the copyright owner.

- CASRAI Dictionary

A two-dimensional representation of data in which values are represented by colors. Heat maps communicate relationships between data values that would be would be much more difficult to understand if presented numerically in a spreadsheet.

- CASRAI Dictionary

The aggregation of data to make coherent observations about the world, meaningful data, or data arranged or interpreted in a way to provide meaning.

- CASRAI Dictionary

It is often considered the job of the scientist, researcher or statistician to derive information from raw data.

The capability to communicate, execute programs, or transfer data among various functional units in a useful and meaningful manner that requires the user to have little or no knowledge of the unique characteristics of those units. Foundational, syntactic, and semantic interoperability are the three necessary aspects of interoperability.

- CASRAI Dictionary

Background or contextual data about a dataset. Literally "data about data". Metadata is required to enable someone to properly understand and interpret a main dataset. 

Examples: 

• The research questions that the data was collected to address
  
• Any relevant environmental conditions affecting the main variables
• The instruments used to collect or generate the data, including specification and calibration details.
• The methodology used to collect or generate the data,
• Definitions of variables, including units - also called a data dictionary. 

Research data captured in real time, usually unique and irreplaceable. 

Examples: Weather records, species census surveys.

A set of terminology to describe important concepts, often specific to a particular domain or discipline. It's a way of describing a vocabulary that can be shared among practitioners in a field, to allow for easier communication and a standardised way of defining and labelling, for example when writing metadata.

Structured data that are accessible, machine-readable, usable, intelligible, and freely shared. Open data can be freely used, re-used, built on, and redistributed by anyone - subject only, at most, to the requirement to attribute and share-alike.

-  CASRAI Dictionary

The practice of science in such a way that others can collaborate and contribute, where research data, lab notes and other research processes are freely available, under terms that enable reuse, redistribution and reproduction of the research and its underlying data and methods. 

- FOSTER Open Science Consortium

Open Science encompasses a broad set of practices, including: 

• Open Access publication of research results
  
• Open Source licensing of software code
  
• Open Dissemination of research data.
  

Open source software is software whose source code has been made freely available for re-use and modification under and Open Source license. 

There are different types of open source license, but to be truly open source they all conform to the guidelines laid out at the Open Source Initiative. 

Data in the form of physical samples.

Examples: Soil samples, ice cores.

Data that has been created or collected first hand to answer the specific research question.

This is data regarding 'qualities', which do not take the form of numbers. 

Examples: interview transcripts, ethnographic materials, photos

Data that is a measure of quantity. It comes in the form of numbers. 

Examples: Measurements, sample analyses, observations and model outputs.

Raw data are data that have not been processed for meaningful use. A raw dataset is exactly what is collected, before any data cleaning, processing or analysis has been completed. 

It is often useful to store raw data as well as the cleaned, processed data, as it can help your work to be more easily reproduced. If another researcher has your raw data and the steps you used to process and analyse, they can recreate your results. This has to be balanced with the cost of storing raw data, and the likelihood of the raw data being useful compared to data that has undergone an initial process of data cleaning.

A static or organic conglomeration or collection of smaller (peer reviewed) datasets, most probably published and curated, e.g. UK Tide Gauge Network, IUCN Red List of Endangered Species

The evidence that underpins the answer to the research question

 - UK Concordat on Open Research Data (2016)

Recorded factual material commonly retained by and accepted in the scientific community as necessary to validate research findings.

 - EPSRC Policy Framework on Research Data

Data that are used as primary sources to support technical or scientific enquiry, research, scholarship, or artistic activity, and that are used as evidence in the research process and/or are commonly accepted in the research community as necessary to validate research findings and results. All other digital and non-digital content have the potential of becoming research data. Research data may be experimental data, observational data, operational data, third party data, public sector data, monitoring data, processed data, or repurposed data.

- CASRAI Dictionary

A model to conceptualise the different stages through which data pass during the research process, and the data management activities that relate to those stages.

The model used throughout Data Tree has six stages, corresponding to different activities during the life of a research project. Other institutions or paradigms have slight variations on these stages, but the broad concepts are applicable no matter how you choose to categorise your research activities. 

Our model is based on the UK Data Service model from 2011, and has the following stages: 

• Re-using Data: Often considered both the start and the end of the cycle. Your research might start by gathering secondary data, and your own research outputs might be later used by yourself or others in different sectors.
• Creating Data: Data collection or generation activities.
• Processing Data: The tasks of turning raw data into analysis-ready data. This includes quality control checks, data cleaning and documentation.
• Analysing Data: Includes data visualisations and statistical analysis; tasks that involve the process of getting information out of your data.
• Preserving Data: The tasks of putting your data into a location for long-term storage and access, such as a data repository.
• Making Data Accessible: This includes not just ensuring your data can be accessed, but also making others aware of your data. This might include publishing in a data journal and adding appropriate licences to your preserved data.

Existing data which is being reused for a purpose other than the one for which it was collected.

Research data generated from test models where the model and metadata may be more important than the output data from the model.

Examples: Climate or ocean circulation models.

Individuals, groups or organisations that have an interest or share in an undertaking or relationship and its outcome - they may be affected by it, impact or influence it, and in some way be accountable for it.

- CASRAI Dictionary

Control over time of data, computer code, software, and documents that allows for the ability to revert to a previous revision, which is critical for data traceability, tracking edits, and correcting mistakes. Version control generates a (changed) copy of a data object that is uniquely labeled with a version number. The intent is to track changes to a data object, by making versioned copies. Note that a version is different from a backup copy, which is typically a copy made at a specific point in time, or a replica. 

- CASRAI Dictionary

Version control is very popular in programming, and many coders use Git or Subversion to track changes to their scripts  and other text-based files. Other, simpler version control systems include things like MS Word's "track changes" feature, and the feature that many cloud storage facilities such as Dropbox and Google Drive have that allows users to revert to previous versions of documents stored in their systems for limited periods.