Rate this article from 1 to 10 on its
evidence, reasoning and balance

How would you summarise the quality of the evidence,
reasoning and balance in this article?

Summary of your analysis




No counters, alternatives or objections are considered.

Under 13s are not encouraged to share on social media

Try again Back to main menu
In the martial art of critical thinking, you are...
A black belt in evidence analysis!


Identifying the main claim - 0/2

The main claim is surprisingly hidden in the body of the article. Overall the article is coming down on the side of petrol being cleaner.

Identifying the reasons - 0/3

After a bit of background explanation, the first key reason concerns the higher levels of particulates produced by diesel engines. Then the next bit of evidence comes from an expert who explains why particulates are a health hazard.

Identifying evidence types - 0/4

That engines produce higher levels of particulates is a fact, but not one that is being treated as either a statistical fact or an example. But there's no reason to think the claim is untrue, and it does indeed work to support the claim. So this is a good bit of reasoning.

For the second reason, we have expert opinion. We have no reason to think the source didn't make the claim, the article gives us information that he is a credible authority in the area, there's no evidence that he's biased, and there's no reason to think he's presenting an unusual or contentious claim amongst experts in his field. So this is another good bit of reasoning.

Identifying counters - 0/1

This article has a very great deal of material on the other side of the claim – it is an example of balanced, impartial reporting. There are two areas where alternative positions are expressly introduced (though there are others that are less clear-cut). The first introduces particulate filters; the second accepts that the overall claim doesn't apply to all diesel cars.

Under 13s are not encouraged to share on social media

Try again Back to main menu

Main Claim

The main claim of an article is what everything else is pointing towards: the big idea.

Claims can be surprisingly hard to spot. They're typically near the beginning or, sometimes, near the end – though occasionally can be found almost anywhere.

They often have a strong thematic connection to the headline – but not always.

They're sometimes signalled with special phrases such as "to sum up" or "the main point is" or "our conclusion is". But not always.

They're often shorter and punchier than other sentences.

Look for what the author is trying to convey as the take-home message – what do they want you to remember about this article if you only remember a single thing?

Listen to this example from the Moral Maze where where the witness marks her main claim very clearly with "I still have to conclude..." The software will then guide you to what it thinks is the main claim by marking it with a wavy red line.



The claim is backed up by reasons. There are usually several reasons backing up the main claim of an article.

There is often a part of a reason that ties in very closely with a part of a claim – a claim that, say, pineapple imports from Cost Rica have been increasing, might be supported by evidence talking about pineapples or about Costa Rica.

Sometimes evidence and reasons are introduced with special phrases such as "a key reason is" or "that's because" or "due to". But not always.

Reasons can involve things like statements from people or reports, statistical details, personal experience and more.

Look for what the author of the article is supplying to get you to understand or believe the main claim.

Listen to this example from the Moral Maze where in his introduction, Michael Buerk lays out the issue and then (after 30 seconds) the most important reason. You can also listen to this example, in which a series of reasons are introduced, each marked by "because". The software will then guide you to what it thinks are the reasons by marking them with a wavy red line.



Reasons are connected to claims in many different ways. The Evidence Toolkit allows you to tease apart two different broad categories of reasons: those that are presented as facts and those presented as opinions. Facts should in principle be objectively checkable; opinions are intrinsically subjective. (Though there is more grey area than you might expect...)



Within the category of facts, there are several specific types of reasoning. Statistical reasoning uses evidence from a sample of a population; example-based reasoning uses just a single case to justify a claim. There are many other types of factual reasoning though – here we lump them together as 'Other'. Click to hear how statistical , example and other fact-based reasoning can be used effectively in debates from the Moral Maze.



Within the category of opinions, there are three more specific types depending on whose opinions are being used. Expert opinions carry a lot of weight if the expert is appropriately qualified. Popular opinion can be influential too, although we need to be more careful how we treat it. And personal opinion can be valuable but only gives a narrow, limited evidence base. Click to hear how expert, popular, and personal opinion-based reasoning can be put into practice with illustrations from the Moral Maze.



One part of being impartial and balanced is considering alternative opinions. These are sometimes known as 'objections.' Authors might not agree with objections to their claim, but might nevertheless mention them (either to partly concede, to emphasise that other opinions exist, or to subsequently provide counterarguments of their own).

These objections are sometimes introduced with special phrases such as "on the other hand" or "admittedly" or "to some extent". But not always.

Objections are usually mentioned in the middle or towards the end of an article.

Click here to hear an example of such balance, in which Douglas Murray agrees with his opponent, Giles Fraser, and concedes that the Qur’an, has passages, which are "highly inflammatory" whilst maintaining his point that that’s not a reason to ban the book (or any other book, for that matter).



The goal of The Evidence Toolkit is to encourage thinking about thinking, and in particular to stimulate and scaffold critical thinking of print and online media. To do this, we have carefully selected articles representing different political and thematic orientations, both news and op-ed, including one article that is satirical. For each article, we have designed the software, the lesson plans and the teacher notes to encourage students to ask deep questions of the content and its context. Students are guided towards answers to those questions that have been assembled by a team of experts. Help is available in the form of illustrations of each concept taken from episodes of the BBC Radio 4 radio programme, The Moral Maze, in which combative, provocative and engaging live debate depends upon the quick witted reasoning skills of the participants. (To listen to these clips you will need flashplayer installed so that you can use iPlayer. You will need to stop the clips manually). Finally, after having worked on the pre-prepared articles, students are invited to try out their skills on articles of their choosing, where there is no longer guidance from experts, only hints and tips on the process from the software. At that point the training wheels are off, and the students are free to exercise their new-found skills of critical thinking.


Now you've learnt about critical thinking and how to dissect an article to figure out whether it's accurate, well-supported and impartial, you can apply your expert skills on articles of your choosing.

Some are easier than others. Start by looking for opinion and editorial pieces written by columnists and correspondents – these often have clearer claims. Aim for shorter articles of under 1,000 words to avoid too much detail.

And remember, the Reason Checker is there to help you – but it's only a guide and uses cutting edge artificial intelligence technologies. Sometimes it'll be spot on. But other times it might get the wrong idea. So you need to think about its suggestions carefully!