Data & Analysis
Research involves collecting information and understanding what it means. On this page, you will find examples of data collection and ways to analyze that information, keeping in mind this process can vary depending on the discipline you are conducting research in.
Data
In a research context, data is any kind of information you may gather and/or analyze while working towards your research goal. The first step to working with data is getting the data. Data collection is essential for the research process, as it allows researchers to gather novel information, accept or reject a hypothesis, and draw conclusions from their work.
10 Potential Data Collection Methods:
Primary and Secondary Data
Data used by research and other studies can be classified as either primary or secondary. Primary data includes any data generated by the researcher during their experiments. Secondary data is anything retrieved from a published database, such as healthcare facilities or government institutions. Common websites used for research include the CRC Handbook of Chemistry and Physics for chemistry, Statistics Canada for statistics across Canada or try visiting the archives at an art gallery or museum for historical data, and so many others.
The definitions of primary and secondary data are different depending on the context. Noted above are the common definitions in mathematics and science. In literature, primary data refers to written material from the same time period as your research, and secondary data refers to written material from a different era. When pursuing your own research, make sure to find the definitions of primary and secondary data that are relevant to your field.
One highly crucial aspect of collecting secondary data is ensuring the validity of your sources. As you may know, there are tons of incorrect data floating around the internet. For this reason, it is vital to verify that your sources give accurate information. This can be done by retrieving data from references such as published databases, government websites, or peer-reviewed journals. A peer-reviewed journal publishes articles written, reviewed and edited by experts in a field to ensure accuracy. Many universities give students access to hundreds of databases and peer-reviewed journals spanning multiple disciplines.
Various Kinds of Primary and Secondary Data:
Qualitative and Quantitative Data
Data collected can be either qualitative or quantitative. Qualitative data include observations that cannot be directly measured, such as colour, smell, memory, reactions. Quantitative data, on the other hand, can be directly measured and is usually expressed numerically such as percentage, volume, number of people, etc. Depending on the type of research you are pursuing, you may find yourself using one or both kinds of data. Determining which kind of data you need for research will help you decide what method of data collection you wish to use, be it experimentation, examination of existing documents, observation, or any other method.
As most present research data is stored on computers, it is also important to learn how to organize and manage your data. This is called a Data Management Plan (DMP) and may look slightly different for everyone. We all have our own unique preferences, but you should make sure to organize your files in such a way that they are easily found. It is also very useful to save multiple copies of your data just in case it is lost or deleted by accident.
Depending on the type of research you are pursuing, there are numerous ways you could collect data. For example, a scientist can gather data directly through experimental work, or a historian could gather data by examining documents from the time period they are studying. You can also collect data from other people in the form of questionnaires, interviews, or social media monitoring. For information on more methods of data collection, visit Global Patron.
For data that is collected without a computer, such as in a laboratory or in the field, it is essential to record it and any observations in a journal or notebook. A well-made DMP will help you first decide what kinds of data you need to collect, how you will collect it, how you will save/store your data, and how you can present it to others.
For more information:
General Information Required When Creating a Data Management Plan:
Analysis
The word analysis, originating from Latin, means “resolution of anything complex into simple elements”. Taking your data and putting it in terms that are understandable to interested readers is a necessity. This is where analysis comes in. Since it involves interpretation of your data, it is one of the most important steps during your research journey. Keep this in mind during your analysis process: garbage in, garbage out. Errors during analysis can halt the next step of your research. There are two main parts to analyzing data. The results and the discussion.
Results
Under your results section, you will either display your raw data or the outcome of your analyses or both. Ensure that if you display only your analyses, you attach a supplementary information document containing the data and indicate clearly in your results section where this information can be found. Since there are two types of data, there are two types of analyses: quantitative and qualitative data analyses.
Quantitative Data Analyses
Quantitative data analyses usually involve the use of statistics (like mean, mode, and so many others). The statistical analyses a researcher uses are dependent on what is being studied therefore, having a great understanding of statistics is very important.
- Precise: Your data should be within a specific acceptable range. If there are outliers (not within the expected range), the researcher should explain this in the discussion.
- Accurate: Important in cases where a researcher has data from past works to compare their results to. If new findings are different, one should be able to explain why in the discussion.
- Reproducible: If your method is performed multiple times by you or another researcher, similar results should be obtained. If this is not the case, one should either troubleshoot their method or be able to explain why in the discussion.
There are more expectations that nominal data is required to reach, however, every research project is different. Therefore, during preparation, the researcher must determine exactly what they expect their data to look like and which analysis best suits their project.
Qualitative Data Analyses
Qualitative data analyses are different from quantitative data analyses because there are no numbers involved. A common example would be in psychology experiments where participants’ reactions are collected as data. In these cases, a researcher is looking for themes and patterns in their data.
Summary of Common Steps in Qualitative Data Analysis:
The themes and patterns in qualitative data are often obtained from previous research projects on similar topics.
No matter what kind of data you are using, an analysis should be performed alongside data collection. This allows the researcher time to adjust their method early enough if need be.
Discussion
Usually, before a researcher embarks on a project, they have a hypothesis: expectations about what their data will be or what pattern they will see. These are usually inferred from completed projects in the past covering the same topics you are exploring. If your data confirms your hypothesis, then you should explain how it fits into the existing research. Always keep in mind that the main goal of academic research is to add to the existing pool of information. If your findings do not meet your hypothesis, then you should be able to look through the existing literature and explain why the observed outcome is different.
While a researcher may be able to easily place their project into existing research, this is not the case for all projects. Those exploring novel connections between topics have to chart unexplored links. Such projects require a very in-depth understanding of the topics being linked and the inferences made have to be backed by existing research. The researcher’s thought patterns have to be vividly described in the discussion section and the links must be workable.
In all research papers, the writer must always describe future work. If during the analysis, a certain factor was unclear, then an educated guess and how to confirm this guess must be proposed. It is in this section of the discussion that methods to address unexplainable results are included. Although this is a great way to show predicted continuity of work, it is best to keep it precise.
As we write discussions we must do our best not to fall victim to the curse of knowledge, which is assuming everyone knows what you know. Poorly done discussions, usually lathered with difficult terminology, can result in research papers that may be difficult to understand and therefore, difficult to disseminate. A few explanations here and there about common terminology could be the difference between a good paper and a great read. It is usually a good idea to keep in mind the interested reader that has no background in your topic.
References
Pell Institute. (2019). Analyze Qualitative Data «Pell Institute. Evaluation Toolkit. http://toolkit.pellinstitute.org/evaluation-guide/analyze/analyze-qualitative-data/