Research Data Management

The National Library of Medicine define 'data collection' as: 

the process of gathering data, typically in the context of a research project or for ongoing surveillance and tracking. This involves measuring information on variables of interest in a pre-established, systematic way that enables researchers to address research questions, test hypotheses, and evaluate outcomes.

In this section you should briefly describe the source of data that you are using in your study. Examples of data collection include: 

• Data captured using data collection instruments such as surveys, wearable devices, standardised scales.
• Data generated by the processing of biological materials, such as blood and tissue samples, that were collected as part of routine care.
• Text, audio or video data captured by observational or interview methodologies.
• Images captured by microscopes, and other types of imaging hardware.

You should give clear details of how new data will be collected or produced. You can describe the methods and software that you will use to generate, capture or extract data. 

Sample answer 1: 

A digital questionnaire will be distributed using MS Forms to a sample of 600 people who are engaging with the service, as baseline and again 3 months after they complete treatment. Data will also be captured from a sub-sample of 200 people via a wearable device for the duration of their treatment (3 months). 

Sample answer 2:

The study will involve conducing 20-25 semi-structured interviews with young adults (aged 16-24) who are participating in [name] programme. Interviews will be conduced face to face and recorded for transcription using digital audio recording where interviewees permit. In case they do not, interviews will be undertaken in pairs to enable detailed note-taking. Interview recordings/notes will be transcribed by the research assistant for analysis.  

Now that you've described where the data you are using for your study will be sourced from, the next step is to describe the attributes of that data. A good approach is to describe the type, format and volume of each type of data that you will be working with. 

You may find it clearer to present this information as a table in your DMP and there is an example of how this might look  on the following tab. 

Data type: You should clearly define the type of data that you are going to generate or work with. Research data can take many forms, and are often discipline-specific, but at a basic level, research data can be described as any information that has been collected, observed, generated or created to validate original research findings. 

Common examples of research data include measurements, experimental results, fieldwork, observations, interview recordings and images. Although usually digital, research data also includes non-digital formats such as hand-written laboratory notes. 

What type of data are you using in your study?

• Text - such as field or laboratory notes, survey responses, transcription of interviews
• Numeric or tabular (data in rows and columns) - such as tables, counts, measurements
• Audiovisual (audio / video recordings) - such as sound recordings or videos of qualitative interviews, observations 
• Images from microscopes, scans, ultrasound, x-ray or other equipment that generate images 
• Models, computer code
• Discipline-specific - such as CIF in chemistry
• Instrument-specific - such as equipment outputs (source: DMPTool)

Data format: For each type of data you are using, list the file format in which the data will be collected and stored during the research process. The file format is usually reflected by the filename extension (such as .doc or .pdf). Not all data formats are electronic, and if you must hold some data in paper copy then your format will be listed as paper or hard copy, but of course this will limit the accessibility of the data. 

While open formats are recommended to ensure long-term usability of data, it is acceptable for you to store data in proprietary format during the research process, if you require the functionality of that format. In the DMP, you should explain why you have chosen certain formats, and this can be based on staff expertise, because a certain software tool us widely used within your community for this type of data, or because it is the format accepted by the repository where you will store the data in the long term. 

Below are some examples of common file formats for data:

Spreadsheets stored as:

• comma-separated values file (.csv)
• Excel file (.xls, .xlsx)
• OpenDocument Spreadsheet file (.ods)

Text files stored as: 

• plain text file (.txt)
• PDF/A file (.pdf)
• Word file (.docx)
• XML file (.xml)
• HTML file (.html)

Image files stored as: 

• TIFF file (.tif)
• JPEG file (.jpg)
• Portable Network Graphics file (.png)

Video / audio stored as: 

• Material Exchange Format file (.mxf)
• FLAC file (.flac)
• MPEG-4 file (.mp4)

Reports or images stored as: 

• PDF/A file
• PDF/X file

At the end of your study you should consider converting the data files to a more accessibility-friendly format that can be opened and used without the need of a specific software license. If you store your data in a preservation-recommended format, it will save you time at the end of the study in converting your files for long-term access. This topic is discussed in more detail in the 'Preserving Research Data' section of this guide. 

Formats for preserving text data

• Plain text (.txt) format is generally the preferred format for both accessibility and preservation
• Formatted text (such as .docx or PDF files, or well-structured HTML rendered in a browser) are usually accepted by data repositories for preservation.

Formats for preserving tabular data

• Comma- or Tab-Separated Value (CSV/TSV) formats are generally the preferred for by data repositories.
• Any proprietary format that is a de facto standard for a profession or supported by multiple tools (e.g. Excel .xls or .xlsx, Shapefile) are usually accepted by data repositories for preservation.

Formats for preserving image data: 

• JPEG Image Encoding family (.jpeg, .jpg)
• TIFF (.tiff, .tif)
• Portable Network Graphics (.png)
• Scalable Vectors Graphics (.svg)

Formats for preserving audio data: 

• Material Exchange Format (.mxf)
• FLAC (.flac)

Format for preserving video data: 

• Material Exchange Format (.mxf)

4TU.ResearchData have provided a 1-page guide on preferred file formats for long term access to files which is a useful reference point. For more detailed and updated guidance, visit the the Library of Congress Recommended Formats Statement. Every year the LOC release a statement of recommendations for digital file formats, guiding the community on the text, tabular and audiovisual file formats that are preferred and which are accepted for long term access to the contents. There are further resources at the end of this page. 

Data volume: For each type of data you are working with, you should give a clear estimate of the volume of that data in the DMP. This will help you to plan how much storage capacity you’re going to need while the project is active, and how much storage capacity you'll need to preserve the data at the end of the study. 

Some research (e.g. research using wearable devices) will produce lots of data over an extended period, and will require a lot of storage space. Other research will produce a relatively small amount of data, in terms of bytes, and the cloud storage provided by RCSI will be more than sufficient to store this volume of data. 

In general, data volume should be expressed in bytes such as:

• gigabytes / GBs
• megabytes / MBs
• terabytes / TBs

If you cannot express the volume in terms of bytes, alternative ways to express data volume include: 

• Audio recordings: Number of expected interviews x average length of interview
• Tabular files: Number of columns (variables) x rows (respondents/cases) 
• Survey data: Number of survey questions or variables x expected sample size
• Image data: Number of images x average size of image this software produces