• Sunday, December 22, 2024
businessday logo

BusinessDay

Information analysis online: Opinion, belief, truth and fact

Analytics and metrics (3)

Aphorisms abound, for the internet age, to help guide us through the digital jungle that’s the World Wide Web. “Correlation is not causation” is one, another is “you are entitled to your opinions but not your own facts.”

An ‘opinion’ can belong to anyone without any consideration to ‘belief,’ ‘truth’ or ‘fact’ and it is always subjective.

‘Belief’ is also subjective and closely related to opinion – it is an acknowledgement of a state of affairs (real or unreal). ‘Truth’ on the other hand is less subjective and can be personal or generally believed by a group of people. ‘Truth’ is less subjective than an opinion and closely related to ‘fact.’ A ‘fact’ is objective and not personal: it is generally believed by most people and supported with evidence.

Distinguishing ‘fact’ from ‘truth’ can be tricky but let me try. A colour-blind person sees a red ball but mistakes it for a purple one because of her inability. For this colour-blind person, it is her ‘truth’ that the ball is purple but the fact is that the ball is red. This is different from “alternative facts” or “alternative world-views.”

We further categorise online content as ‘analytical’ if it is logical, coherent and contains claims supported by evidence. ‘Non-analytical’ online content can be a mess of opinion, superstition, bias, prejudice and sentiments. What about ‘data,’ ‘information’ or ‘knowledge’? These terms are hard to define strictly and are used interchangeably by people.

But suffice it to state that the clearest relationship between the three terms is that ‘information’ is ‘data’ as input gathered and transformed or interpreted, which can produce knowledge as output. The global definition of information (GDI) defines ‘information’ as “data that is well-formed and meaningful.”

In our book, Kirsti Ryall, I state: increasingly, more people online believe one point of view over another instead of checking for facts when an article is encountered online. Simple rule we need: read the whole piece before judging content and seek further information when necessary. In our online content analysis work, we use an analytical content checklist that is descriptive: it does not endorse or condemn articles online. It upholds freedom of speech.

The rating (information quality) we give to content we analyse is objective. Readers are given a more informed choice to agree or disagree with content. We hold that opinions, beliefs, truths, and facts are important to online content analysis if they can lead to knowledge (true or false), decision or action.

Read also: Enabling affordable and sustainable internet access in Africa

History is mutable: ‘facts’ from the past can be discounted or broadened to include further research by people many years later. Our analytical checklist system is not prescriptive: it is not some ruling demanding readers accept or reject an online article rated by us. Our analysis system is descriptive and exploratory.

Nor is it that our checklist’s scoring cannot be reconsidered. It can and it should. New information could reveal how truthful or not stated ‘facts’ are within an article (online content analysed), the scoring we gave previously to the content would be reconsidered. Based on feedback, from the public, our system has been adjusted to reflect this.

Apparent ‘truths’ and opinions change over longer periods of time. What was once viewed as general, normal and consensus can change as we learn more and society progresses. Take the example of the ‘lab leak’ hypothesis of the origin of the COVID-19 virus.

This hypothesis was dismissed and labelled initially as misinformation or as a conspiracy theory. However, with the passing of time this particular hypothesis is now generally regarded as plausible.

The failure to set aside bias, prejudice and sentiments often cripples cognitive ability and distorts decision making.

In the subfield of management science known as decision theory or decision analysis, a ‘good’ decision is differentiated from the ‘right’ decision. It all has to do with information, objectivity and analysis.

A good decision is a decision based on available evidence, data, information and analysis at the time the decision is made. It is generally not possible to know if a good decision will be the right decision until time and events say so. The right decision is simply a decision taken in the past that with hindsight turns out to be correct: the consequences of the decision are revealed as positive, predictive or the decision served the purpose of the decision maker.

Omoregie, principal analyst at Avram Turing, writes from Canada

Join BusinessDay whatsapp Channel, to stay up to date

Open In Whatsapp