Time

The lesson takes around 20 minutes to complete.

About the lesson ‘What are data, and how are they different from information and knowledge’

Before we can understand concepts such as data, data collection, data processing, data analysis etc., we first need to understand more generally what data are – and what they are not. Data, information and knowledge are popular concepts that are often used interchangeably and in many different contexts. However, in order to understand what they each mean, it’s important to consider how these terms differ from each other.

In this lesson, you’ll be given an introduction to the concepts of data, information and knowledge, and you’ll learn how they relate to each other. You’ll then be introduced to different types of data. You’ll be shown the differences between primary and secondary data, qualitative and quantitative data, and structured and unstructured data.

Learning outcomes

When you have completed the lesson, we expect that you:

can recall the difference between the concepts of data, information and knowledge.
can describe and identify quantitative and qualitative data.
can describe and identify primary and secondary data.
can describe the difference between structured and unstructured data.

Source

The lesson was produced by the University of Copenhagen as part of its learning resources for digital literacy, 2023.

Course directors:

Christian B. Knudsen, specialist consultant, Copenhagen University Library (KUB)
Daniel H. Pryn, information specialist, Copenhagen University Library (KUB)

__________________________________________________________

Data, information and knowledge

Our point of departure is a hierarchical model widely known as the “knowledge pyramid”. The pyramid exists in many forms, sometimes adding concepts such as “insight” and “wisdom” to the model, but here we will focus on a three-way split into “data”, “information” and “knowledge”.

The illustration was adapted from: Rowley, J. (2007). The wisdom hierarchy: representations of the DIKW hierarchy. Journal of Information Science, 33 (2), pp. 163-180.

Data can be described as raw input for a specific observation, which acts as the minimum unit for the analysis that you are conducting. Data are therefore uninterpreted, and are the raw material from which information and knowledge are created.

Information is interpreted data. This means data processed in a specific way, for example, a description, organisation or contextualisation of data for a particular purpose. In other words, information can be described as organised data that have significance or meaning.

As a result, what counts as data in one analysis can be information in another. For example, if you are counting the collected works of Holberg, each work can be interpreted as data. But if you are analysing the collected works of Holberg, the content of a particular work will be interpreted as information. Knowledge is the most complex concept of the three. Knowledge means experience, insight and understanding acquired through learning or reflection. Knowledge is an integral part of our understanding of the world and is based on information that is processed and used in a meaningful context. Knowledge can be described as a fundamental understanding of information and as the ability to deconstruct a particular piece of information and answer the question: why does this information describe what it does? A deconstruction of this kind will naturally lead to a description of the relationship between multiple underlying pieces of information, so knowledge also describes the ability to combine two or more pieces of information and end up with more than the sum of their parts.

The relationship between data, information and knowledge can be illustrated as follows:

Graf – Data-Information-Viden.png

The figure was adapted from: Bellinger, G., Castro, D. & Mills, A. (2004). Data, Information, Knowledge, and Wisdom.

Another way to illustrate the relationship between data, information and knowledge is presented below:

We touched upon the concepts of “insight” and “wisdom” earlier, and they can be added (tongue-in-cheek) to the above illustration:

Illustration by David Somerville (http://www.smrvl.com/blog/) based on the original by Hugh McLeod (https://www.gapingvoid.com/team-members/hugh-macleod/)
– unicorn/conspiracy added by anonymous

__________________________________________________________

Examples illustrating the difference between data, information and knowledge

In order to elaborate on the difference between data, information and knowledge, it’s helpful to think of an example. Let’s say that you want to study what affects plant growth in your garden. You start by collecting data on factors such as the soil pH, the amount of sunshine, the amount of water given to the plants etc. These data are raw observations, measurements or numbers that are not meaningful in themselves.

The next step is to process and organise data in a meaningful way. For example, you could create a table setting out each factor and measuring it on a scale (e.g. pH value: 6, sunshine: 8 hours per day, irrigation: 500 ml per day etc.). This can help you to create information about which factors affect plant growth in your garden.

Finally, you can use the information to create knowledge about how the different factors interact, and to identify patterns in the way sunshine, irrigation and pH value relate to each other. You might discover that plants that are exposed to high levels of sunlight and small amounts of water grow better if they are in soil with a low pH value. From there, you will be able to formulate hypotheses about how the different factors interact and influence each other. In other words, you begin to create new knowledge about how the plants grow, which is more than just the sum of the information provided. You can then use this knowledge to make decisions about how to cultivate your plants, and which factors you should take into account to optimise growth.

When is data data? And when is information information?

Data are the raw input for a particular observation, and data play an elemental role in our understanding of the world. However, this does not mean that the data are elemental per se, just that the data are elemental in the context of cognition. Data can therefore be understood as the necessary minimum unit before a specific observation becomes a cognition, so data for one person can be information about data for someone else. An example of an observation is that it is raining, and if we leave it there, “it is raining” will be the minimum unit for this observation, meaning that the observation of rain constitutes data. However, if this observation is made by a meteorologist, the observation of rain can be seen as information indicating that a low pressure system has formed (based on the knowledge the meteorologist possesses). In this context, the formation of the low pressure system is data and rain is information. Even this example is not straightforward, however, because in practice it is difficult to conceive of a “pure” observation with no further cognition taking place after the raw observation. For example, an observation of rain will naturally lead to a cognition that you will get wet if you go outside. It is even more difficult to describe such an observation in “pure” terms, as any description will be based on language, meaning that interpretation and cognition happen automatically. Now that the observation has become an interpretation, it is no longer data we are talking about, but rather information.

Data must always be understood as data for a specific observational context, and what constitutes data in one observation is not necessarily data in another observation. We have circled back to the original definition of data as elemental for a specific observation. This is not because actual data are indivisible, but because the data lose their character as data (becoming information) if they are divided in this way.

Here’s another example:

A book, in and of itself, can also be considered to be data, but becomes information when it is read. Depending on who is reading it, different (or even contradictory) information will be derived.

If the book is written in Chinese, most Danes will be unable to read it. This means that no understanding can happen (among these non-Chinese speakers), meaning that the book remains data within the relevant context.

What we mean by the relevant context here is the context of language-based interpretation. As described above, though, it’s difficult to imagine a situation in which no information is derived from a particular set of data, and in this example, the mere cognition that this is a book is in itself information about a number of characteristics, such as the fact that it has at least one author etc. However, this information does not fall within the scope of the context of the example described.

__________________________________________________________

All students work with data!

When people talk about data, it’s very easy to think about quantitative data. In other words, things that can be measured and weighed and put into an Excel spreadsheet. By now, though, you’re hopefully aware that this is far from the whole story. It’s true that some data are observations organised in an Excel spreadsheet or similar tool, but data are used in all university studies and take many other forms depending on the academic discipline and the topic.

Are you a student of literature and cultural studies? If so, the collected works of Georg Brandes may be the data you are working with. As you read and take notes, you are deriving information from the raw data – the works themselves. Later, you create knowledge by synthesising the information you acquired while you were reading. Of course, you can also extract information using purely digital methods, in which you instruct the computer to “read” the works, but the works are still your primary data.

Are you studying law? Judgments and preparatory works are important legal sources. As you work through the comments on the Public Administration Act, you’re interacting directly with data which, after a more detailed analysis, explain when a document is an external expert opinion and when it is internal. Digitised analyses of judgments can strengthen the work of defence lawyers or automate the routine aspects of document preparation. This field is often called “legal tech”.

As a student in the natural and health sciences, you’ll encounter lots of data that are instantly recognisable as data. The concentration of potassium in a patient’s blood, pH during a titration or the number of seals in particular waters are obvious examples of data. But things that you might not immediately think of as data can also be data. X-rays can be considered to be data, as can photographs from wildlife cameras or medical histories. Here too, you can often let your computer do the work – but whether it’s you or the computer that notices a badger in the photo, data are being processed.

__________________________________________________________

Terms for different types of data

When we talk about data, the following categories are often used to denote different types of data:

Primary and secondary data

Primary data are essentially data collected directly from the source, e.g. through observations, questionnaires, interviews or experiments. Primary data are often specific to a given study and not available from other sources.

Secondary data are data collected in a context other than the current study. In many cases, secondary data have already been processed and organised in a meaningful way and are available from books, articles, reports, statistics, public databases etc.

However, there may be situations in which the distinction is not quite as simple. For example, in a literature study, it may be the literature itself that is the research object, so the literature can be used as primary data in this context. The literature study could be looking into how works of fiction reflect social structures or ideas, how they affect the reader’s perception of the world, or how they express the author’s views and opinions. In this situation, the works are primary data, and the results of your literature study may be used as secondary data by others.

There are advantages and disadvantages of using primary and secondary data in research. Primary data are often more relevant to the study you are carrying out, but collecting them can be time-consuming. Secondary data may be easier and faster to access, but they may be less relevant to your particular study as they were not collected for the same specific purpose. (Note that this is a generalisation and that there will be cases where the opposite is true. It always depends on the specific data you need to collect and the way they are collected.)

Quantitative and qualitative data

Qualitative data are data that describe the characteristics of a thing or a situation. Examples include emotions, attitudes, opinions, experiences or descriptions of things. Qualitative data are often not quantifiable in concrete numbers, but can be described and analysed by means of text, images or sound, for example.

Quantitative data are data that describe quantitative characteristics of something, i.e. sizes, volumes or amounts. They can be measured by counting or weighing, for example. Examples of quantitative data include quantity, length, weight, temperature, time or other measurable properties.

However, there may be an overlap. A text can express a particular attitude, the character of which is assessed subjectively, and as such it is qualitative. With a large number of texts, however, the number of people expressing an opinion and the number of people expressing a different opinion can be counted, resulting in quantitative data on the distribution of qualitative observations. Similarly, there are areas of research that involve quantifying such things as emotions and attitudes. In other words, the distinction between qualitative and quantitative data is not always quite so clear.

Qualitative data on the one hand and quantitative data on the other are used in research to achieve different things, depending on what is being studied. Qualitative data can often provide insight into meaning, motivation and subjective experiences, whereas quantitative data can provide a more generalisable understanding of phenomena and can be used to test hypotheses and theories. However, it is important to emphasise that this is only a general description of the difference between the use of qualitative and quantitative studies. For example, in some studies, we might use quantitative methods to examine subjective factors, and qualitative methods are occasionally also used to investigate and understand general phenomena. This means you must carefully consider how the methods are used within your own subject area.

Structured and unstructured data

Structured data are data that are organised in a way that facilitates searching, sorting and filtering. This means that there is a fixed structure determining how data are arranged and organised. An example of structured data is a database in which each record has fixed fields, and each record can be sorted and filtered by these fields.

Unstructured data are data that do not have a fixed structure. This means the data are not organised or arranged in any particular way, for example, in a text document or as an audio recording of a conversation between two people. It’s more difficult to search, sort and filter unstructured data because there’s no fixed structure you can use.

The structuring of unstructured data usually involves organising and processing the data into meaningful information. Various tools and software can be used to do this, depending on the reason why the data are being collected, and therefore also depending on an appraisal of what is meaningful and valuable.

For example, questionnaires may contain unstructured data in the form of free text, necessitating a process to structure the data into meaningful information. This can be done by sorting data into categories, by counting responses or by processing data in other ways in order to create structured data consisting of tables, diagrams or other visualisations. The transition from unstructured to structured data is often fluid, as a process is required to structure and process data into meaningful information.

__________________________________________________________

Example of structured data: Plant growth

Revisiting our study of plant growth in relation to hours of sunshine, irrigation and the soil pH, the data might initially look like this:

mlwater/hours sunshine/ph/growth cm	20/6	21/6	22/6	23/6
tomatoes	500/6/7/10	400/8/6.5/12	300/4/7/14	450/7/6,8/16
cucumbers	500;6;7;8	400;8;6,5;10	300;4;7;12	350;5;7.2;14

Data cleaning and organisation normally requires several steps, and a good rule of thumb is that 80% of the time is spent on data cleaning, while 20% is spent on data analysis. The data below have been brought into a more structured form that is easier to search and work with:

Date	Vegetable	Irrigation (ml)	Sunshine (hours)	pH of soil	Growth (cm)
2023-06-20	Tomatoes	500	6	7.0	10
2023-06-20	Cucumbers	500	6	7.0	8
2023-06-21	Tomatoes	400	8	6.5	12
2023-06-21	Cucumbers	400	8	6.5	10
2023-06-22	Tomatoes	300	4	7.0	14
2023-06-22	Cucumbers	300	4	7.0	12
2023-06-23	Tomatoes	450	7	6.8	16
2023-06-23	Cucumbers	350	5	7.2	14

__________________________________________________________

Example of how to create structure in an interview recording

An example of a step in data structuring could be the processing of audio or video recordings of interviews. A process is required to structure such data into meaningful information. The transformation may involve transcribing and cleaning text data, then sorting into categories, counting responses or processing data in other ways. This allows you to organise your data for a clearer overview, while making them easier to process and use to draw conclusions.

Here’s an example of how you can structure data from interview recordings:

If you experience access denied, try reloading the page.
For English subtitles, please look for the CC icon in the lower right corner of the video and press English.

__________________________________________________________

Check what you’ve learned about data, information and knowledge

We’ve created a short quiz so you can check what you’ve learned about the terms introduced in this lesson.

Go to the quiz.

__________________________________________________________

If you want to know more

Places where you can find out more about data in an academic context:

Lesson ‘Using digital data to produce knowledge’ in the ‘Digital reflection’ module
Lesson ‘Take care of your data – data classification and risks’ in the ‘Data management’ module

__________________________________________________________