Tell Source Types Apart

In Part 1 you decided what you were looking for and planned how to go find it. Now the results are coming back, and a new problem appears: the page in front of you might be a careful study, a quick news summary, someone’s opinion, or an advertisement in disguise, and at a glance they can all look equally solid. Before you can decide whether to trust a source, you have to know what kind of thing you are actually looking at. This chapter gives you a simple, reliable way to name it.

Sort Any Source on Two questions, Not One

Imagine your search for “does magnesium help sleep” returns two pages. One is a published study that ran a sleep experiment. The other is an online shop’s product page for magnesium pills. Both say magnesium helps you sleep, but they are clearly not the same kind of source, and you would not trust them in the same way.

A source type is just a label for what a source is and what it was made for. The mistake almost every beginner makes is trying to sort sources into a single list: “is this primary, secondary, opinion, or an ad?” That feels tidy, but it jams two completely different questions into one, and it falls apart fast. Where does a vendor’s page that quotes a real study go? Is it “an ad” or “secondary”? The honest answer is both, because you are actually asking two separate things.

So we sort every source on two axes at once.

The first axis is provenance: where the information originally comes from.

• A primary source is the original, first-hand record itself: the study that ran the experiment, the raw dataset, the official statistics, an eyewitness account, the original document. It is the thing closest to the facts, before anyone else repackaged it.

• A secondary source reports on, summarizes, or comments on a primary source: a news article about a study, a blog explaining the research, a review that pulls several studies together. It points back upstream to something more original than itself.

The second axis is intent: what the source is mainly trying to do.

• Informational: to inform or report (a study, a statistics page, a straight news report).

• Opinion: to argue a viewpoint or persuade you (an op-ed, a columnist, a “why I think…” post).

• Promotional: to sell or promote something (a product page, an ad, a sponsored article).

The two axes are independent, and that is the whole point. A source has one answer on each axis, and any combination is possible.

Because the two axes protect you from different mistakes. Provenance tells you how close the source is to the original facts, so you do not treat a quick summary as if it were the evidence itself. Intent tells you what the source wants from you, so you do not mistake a sales pitch for a neutral report. One label alone always leaves a blind spot; together they give you a fair first read on almost anything.

Spot the Type in the Wild

Labels are easy to define and harder to apply, so let’s run real-looking sources through both axes. Notice that the topic keeps changing on purpose, because classifying a source is a skill you use everywhere, not just in one field.

• A peer-reviewed article reporting a sleep experiment the authors ran. This is the original record of the research itself, written to report findings. Provenance: primary. Intent: informational.

• A newspaper article headlined “New study says magnesium improves sleep.” It did not run the study; it is telling you about one. Provenance: secondary (it points back to the study). Intent: informational.

• A supplement shop’s product page that says “clinically proven to improve sleep.” It reports on research it did not produce, and its real job is to sell you pills. Provenance: secondary. Intent: promotional.

• An opinion column arguing your city should fund more cycle lanes. It may mention statistics, but it exists to push a viewpoint. Provenance: secondary. Intent: opinion.

• A government road-safety portal publishing the city’s own injury figures. This is the original dataset, released to inform. Provenance: primary. Intent: informational.

Look at the last two next to the first one. A study and a government statistics page are both primary, even though one is “research” and the other is “official numbers”, because each is the original first-hand record. Primary does not mean “scientific”; it means closest to the source. And look at the shop page: it is secondary and promotional at the same time, which a single-list approach could never capture.

Read a Source at a Glance

With practice, a few signals will let you place most sources in seconds. Treat these as hints that point you in the right direction, not as guarantees, and always confirm by asking what the source actually is.

Your Turn!

Time to classify some sources yourself.

The situation. You are researching whether a popular electric toothbrush actually cleans better than a manual one, and your search returns five results. Before trusting any of them, you want to know what each one is.

A clinical study, published in a dental journal, that ran a brushing experiment and reports the results.A news health-section article titled “Electric brushes may beat manual, study suggests.”The toothbrush maker’s own product page stating “proven 2x better plaque removal.”A dentist’s personal blog post arguing that technique matters far more than the brush.A national dental association’s published guidance summarizing what current research recommends.

Your task. For each of the five, give it one label from each axis: provenance (primary or secondary) and intent (informational, opinion, or promotional). Add a few words saying which signal gave it away.

Deliver: a five-row list, each with its two labels and a short justification. Keep it: this is exactly the “what is this source?” call you will record for every source on your project’s Source Map.

Then check your answer against the model solution found at the end of this chapter.

Let’s Recap!

• A source type is a label for what a source is and what it was made for, and it always takes two labels, not one.

• Provenance asks where the information came from: primary sources are the original first-hand record, secondary sources report on or summarize a primary source.

• Intent asks what the source is trying to do: inform, argue an opinion, or promote and sell.

• The two axes are independent, so a single source can be, for example, secondary and promotional at the same time.

• Quick signals place most sources fast, but always confirm by asking what the source actually is before you trust it.

You can now name what a source is on both axes. In the next chapter, you’ll follow the provenance axis upstream and learn to tell the original source of a claim from whoever just happened to report it first.

Model Solution

There is no trick here, just a fair read on both axes. Here is a strong answer.

• 1. The clinical study.Primary, informational. It is the original experiment, reporting its own results. The giveaway: there is nothing more original to trace it back to.

• 2. The news article.Secondary, informational. “Study suggests” tells you it is reporting on research it did not run, with no product or side to push.

• 3. The product page.Secondary, promotional. It leans on research it did not produce, and its purpose is to sell the brush. “Proven 2x better” plus a buy button is the tell.

• 4. The dentist’s blog post.Secondary, opinion. It argues a viewpoint (“technique matters more”); the “I think” framing and a single named author’s stance mark it as opinion, even if it is well informed.

• 5. The dental association guidance.Secondary, informational. It summarizes and recommends based on existing research, so it points upstream rather than presenting its own first-hand study.

If your labels named both axes and you could point to the signal behind each, you have got the move. The topic will keep changing; the two questions never do.