Media Studies · Media Studies: Interpreting, Evaluating and Producing Complex Media

Credibility, Misinformation and Synthetic Media: Evaluating Information Online

1.3 1.5About 70 minutes

Learning goals

I can distinguish misinformation, disinformation and malinformation, and fabricated, manipulated and false-context media.
I can use SIFT and lateral reading to evaluate the credibility of online information and images.
I can identify the perspective of a credible source and explain why credible is not the same as neutral.

Words to know

misinformation
False or misleading content shared without intent to harm.

disinformation
False content created or shared deliberately to deceive or cause harm.

malinformation
True information used to cause harm.

synthetic media
Images, audio or video generated or altered by software, including generative AI.

deepfake
AI-made or AI-altered video or audio that makes a real person seem to say or do something they didn't.

SIFT
Stop, Investigate the source, Find better coverage, Trace to the original.

Remember from Grade 10

In Grade 10 you used lateral reading: leaving a page to check what other sources say about it. You also identified perspectives and biases and learned that news outlets frame stories. This lesson applies those skills to a hard problem for Grade 11: evaluating information online, where misinformation spreads quickly and images and video can be synthetic (made or altered by software, including generative AI).

Hook: a moose at the mall

A photo goes viral on a Saturday morning: a full-grown moose standing calmly in a shopping-mall food court beside a pretzel stand. The caption: This morning in Thunder Bay!! Stay safe everyone. It has 40 000 shares by noon. Your cousin sends it to the family chat.

It's funny. It's plausible: moose do sometimes wander into Canadian towns. Is it real? How would you find out in two minutes? (This post is invented for the lesson.)

Three kinds of information disorder

Misinformation: false or misleading content shared without intent to harm. Your cousin sharing the moose photo because she believes it.
Disinformation: false content created or shared deliberately to deceive or cause harm, for money, attention, politics or chaos.
Malinformation: true information used to harm, such as posting someone's private address.

The same post can move between categories as it travels. That is why we judge both the content and the source.

Why false content spreads

It is often emotional (outrage, fear, delight), and emotion makes people share before thinking.
It is often novel and surprising.
It confirms what people already believe (confirmation bias).
Platforms reward engagement, and false content often engages well.
Sharing feels like social connection, even when the content is wrong.

Researchers Claire Wardle and Hossein Derakhshan sorted "information disorder" by two questions: Is it false? and Is it meant to harm? Misinformation is false; disinformation is false and deliberately shared to harm or deceive; malinformation is true but used to harm (for example, leaking someone's private information).

Synthetic and manipulated media

Images, audio and video can be misleading in several ways:

Fabricated: made from nothing, for example by a generative AI tool.
Manipulated: a real image or clip that has been edited (objects added or removed, faces swapped).
Deepfake: AI-generated or AI-altered video or audio that makes a real person appear to say or do something they didn't.
False context: a real, unedited image shared with a false caption, date or place. This is one of the most common kinds, because it needs no special tools.

People used to look for glitches (strange hands, warped text) to spot fake images. Those clues are becoming less reliable as tools improve. Checking the source and tracing the original works much better than staring at pixels. Some companies and news organizations now attach content credentials (information recording how an image was made and edited), but most images online have none.

The SIFT method

Instead of reading a doubtful post more closely, move sideways to check it.

The SIFT method, developed by digital-literacy educator Mike Caulfield: Stop, Investigate the source, Find better coverage, Trace to the original.

Credibility and perspective

Evaluating credibility connects to perspective and bias (expectation 1.5). Ask:

Who made this, and what is their expertise or track record?
Why did they make it? Who benefits if I believe it?
What perspective is represented, and whose is missing?
How do they know? Do they show evidence, sources or original documents?

A source can be credible and still have a perspective: an environmental group, an industry association and a government agency may all report accurate facts but select and frame them differently. Credible is not the same as neutral.

Worked example

Let's SIFT the moose photo

Use SIFT to evaluate the invented viral photo of a moose in a Thunder Bay mall food court.

Stop. Notice the reaction: delight and surprise. Those emotions make people share fast. Don't share yet.
Investigate the source. The account that posted it was created three weeks ago, has no profile details and posts only viral animal photos. Searching the account's name finds nothing about who runs it. Not a credible original source.
Find better coverage. A moose in a city mall would be local news. Searching moose mall Thunder Bay finds no reports from local news outlets, the mall or police. Absence of coverage is a strong warning sign for an event this unusual.
Trace to the original. A reverse image search (uploading the image to a search tool) finds the same photo posted two years earlier by a digital artist, labelled AI art: "Moose goes shopping." The caption was changed and the label removed.
Classify. The image is fabricated (AI-generated) and shared with false context. The account that removed the label likely spread disinformation to gain followers; your cousin spread misinformation by mistake.
Respond kindly. Reply in the family chat with the original link: "Fun picture, but it's AI art from 2 years ago, not Thunder Bay."

The moose photo is not a real event. SIFT shows an unreliable new account, no coverage from local sources, and an original labelled AI art. Tracing beats squinting: no pixel inspection was needed.

The 'who else?' test

For any surprising claim, ask: If this were true, who else would be reporting it? If the answer is lots of reliable outlets and none are, be very cautious.

Just for fun

Funny but true: moose really do wander into Canadian towns sometimes, which is exactly what makes a fake moose photo believable. The best false content is built on something that could happen.

Hands-on activity

Two-minute fact check challenge

You need: a device with a search engine (with a grown-up's OK), a timer and paper.

With a grown-up, choose three claims or images you've seen recently (a surprising headline, a "did you know" fact, a dramatic photo).
Set a timer for two minutes per item. Use SIFT. Write what you found for each letter.
Try a reverse image search on at least one image (most major search engines offer one: you upload the image or paste its address).
Rate each item: Trustworthy, Uncertain or Not trustworthy, with one sentence of evidence.
Reflect: which step gave you the most useful information?

Safety note: don't click unknown links or download files while fact-checking; use search results from known sites.

What to look for: notes for every SIFT step, at least one lateral search, a rating backed by evidence, and an honest Uncertain when the evidence is mixed.

Your turn

Define misinformation, disinformation and malinformation.

What do the letters in SIFT stand for?

What is a deepfake?

Why is false context so common?

Why is checking the source more reliable than looking for glitches in an image?

Give two reasons false content spreads quickly.

More practice: warm-up

What does lateral reading mean?

Is a real photo with a false date fabricated, manipulated, or false context?

What is the first step of SIFT?

What is confirmation bias?

True or false: a credible source is always neutral.

More practice: on your own

Your friend shares a post claiming a new law bans snowmen on front lawns. Apply the who else? test.

An industry group publishes accurate statistics showing its product creates jobs but leaves out environmental costs. Is it credible? Biased? Explain.

Explain how one false post can be both disinformation and misinformation.

A dramatic video claims to show this week's flood in a Canadian city. How would you trace it?

Why should you correct a family member kindly?

Challenge

Evaluate this claim: "If an image has content credentials, it is definitely true."

Explain how the business model of social platforms can help disinformation spread, and what a platform could change.

Write a short paragraph evaluating the credibility and perspective of an invented post: a doctor with 200 000 followers says a vitamin cures colds, linking to her own vitamin shop.

Check your understanding

Quiz

A person shares a false post because they believe it's true. This is:

Which step of SIFT involves a reverse image search?

A real 2019 photo of a storm is shared as today's storm. This is:

What is the most reliable way to check a suspicious viral image?

Credible and neutral are:

Why does emotional content spread faster?

For parents and teachers

What this lesson teaches: evaluating how effectively and honestly online texts communicate information (1.3) and identifying the perspectives and biases of sources (1.5). Students learn the categories of information disorder (Claire Wardle and Hossein Derakhshan's framework), the SIFT method (Mike Caulfield), lateral reading and reverse image search. The moose post is invented.

How to help at home:

When a surprising post arrives in a family chat, try SIFT together. Make it a game, not a scolding.
Model a kind correction with a link to the original.
Supervise searches and avoid clicking unknown links or downloading files.

Common difficulties:

Reading more closely instead of reading laterally. The answer is usually off the page.
Equating credible with neutral. Accurate sources still have perspectives.
Over-trusting visual clues. AI images are getting harder to spot by eye.