Chapter 12: Finding and Evaluating Sources

Search Strategically

Evaluate Rhetorically

Give Sources Jobs

All chapter content remains available without JavaScript. The keyword builder, route chooser, source triage feedback, and browser-saved source log require JavaScript, but the instructions and examples remain visible.

Research can begin with a deceptively simple moment: you open a search box, type your entire essay question, and receive thousands or millions of results. Somewhere in that avalanche may be exactly what you need. There may also be advertisements, repeated summaries of the same claim, outdated material, sources written for a different audience, and pages that look polished while saying very little.

The challenge is not merely finding information. The challenge is making good decisions about where to search, what to trust, what to keep, and what each source will actually do in your paper.

A research paper does not become strong because it contains a large pile of sources. It becomes strong when the writer chooses sources deliberately, understands them accurately, and gives each one a clear job.

Key Insight: A source belongs in the paper because it has a job, not because it was difficult to find.

By the end of this chapter, you should be able to:

turn a broad topic into a focused research question and useful search terms;
choose among library databases, Google Scholar, and the open web;
revise a search when the first results are weak, overwhelming, or repetitive;
use SIFT and lateral reading to investigate online information;
distinguish a source’s credibility from its usefulness;
distinguish peer-reviewed academic, other scholarly, popular, trade or professional, government, advocacy, primary, and reference sources;
match each source type to an appropriate research job;
keep a source log that records citation information and a plan for using the source;
save the information needed for accurate MLA citations while you research.

Topic, Question, and Keywords

A broad topic names the territory. A research question gives you a direction through that territory. Keywords create the routes you can try.

Suppose your topic is social media and college students. That topic is real, but it is too large to guide a focused search. Social media could involve attention, sleep, friendship, political information, advertising, identity, privacy, body image, career networking, or dozens of other issues.

A stronger research question might be:

Focused research question: How does late-night use of short-form video platforms affect the sleep routines and academic habits of first-year college students?

That question identifies a behavior, a platform type, possible effects, and a population. Instead of pasting the entire question into every search box, break it into concepts and build a keyword map.

Main concept	Related terms and alternatives
short-form video	TikTok, Instagram Reels, YouTube Shorts, algorithmic video feed
late-night use	bedtime use, nighttime screen use, evening media use, screen time before sleep
sleep	sleep quality, sleep duration, bedtime delay, insomnia, circadian rhythm
academic habits	study habits, attention, class attendance, academic performance, procrastination
first-year college students	freshmen, first-year undergraduates, college students, emerging adults

A first search might combine only two or three concepts:

TikTok AND sleep quality AND college students
short-form video AND bedtime delay
nighttime screen use AND academic performance
social media AND sleep AND first-year students

Different words open different doors. A promising article may introduce a term such as bedtime procrastination, which then becomes a useful search term. Research is partly the process of learning the language other people use to discuss your question.

Keyword Builder

Enter two or three main concepts and some alternatives. The tool will create starting search strings, not an argument or a final answer.

Main concept 1 Related terms, separated by commas

Main concept 2 Related terms, separated by commas

Main concept 3, optional Related terms, separated by commas

Key Takeaway

Your first search is not a final exam. It is a scouting trip.

Strong research often moves among three search lanes: library databases, Google Scholar, and the open web. These lanes overlap, but they are not interchangeable. Each is built for different kinds of discovery.

Lane 1: Library Databases

Library databases collect and organize material that the college provides through subscriptions. Depending on the database, that material may include scholarly journal articles, newspapers, magazines, reference entries, reports, ebooks, statistics, primary documents, and professional publications.

The CT State Library provides multidisciplinary and subject-specific databases, research guides, citation support, and help from librarians. Database availability can change, so enter through the current CT State Library portal rather than relying on an old vendor bookmark.

Databases are especially useful for peer-reviewed scholarship, newspaper or magazine coverage behind paywalls, reliable background information, subject headings, filters, and stable citation records.

Lane 2: Google Scholar

Google Scholar searches academic material across publishers, repositories, court opinions, books, theses, conference papers, and other scholarly locations. It is useful for entering an academic conversation and following trails around a promising source.

It can help you search an exact title or author, limit by year, follow Cited by, use Related articles, check All versions, and locate accessible PDF or HTML copies.

Scholar normally sorts by relevance rather than date. A high citation count shows that other writers referred to a work, but it does not prove that the source is correct, current, suitable, or cited positively. Citation counts are clues, not crowns.

Lane 3: The Open Web

The open web may provide government data, current journalism, organizational reports, policies, laws, speeches, advertisements, social posts, videos, datasets, maps, images, and material too recent to have reached a scholarly journal.

The web is not automatically less credible than a database. A federal agency’s original dataset may be more authoritative for a statistic than an article summarizing it. A company advertisement may be the correct primary source when you are analyzing the company’s rhetoric.

The open web also demands more investigation because publication standards vary. A professional design, official-sounding name, or .org address does not prove reliability.

The Lanes Work Together

A library article may provide vocabulary for a better web search. A government report may name an expert whose scholarship you locate through Google Scholar. A news article may summarize a study, leading you to trace the claim back to the original research.

Choose a search lane based on what you need, not on a rule that one lane is always superior.

Example: A student researching hostile architecture might begin in a multidisciplinary database with hostile architecture AND public space. A useful article may introduce related terms such as defensive design, exclusionary design, or anti-homeless design. Those terms can improve searches in all three lanes.

What you need	A strong place to begin
scholarly research on a focused question	library database or Google Scholar
current reporting on a recent event	newspaper database or credible open-web journalism
a law, policy, public record, or government statistic	official government source
background and vocabulary for an unfamiliar topic	reference database, research guide, or strong overview
an advertisement, speech, social post, or company statement to analyze	original primary source on the open web
professional practice within a field	professional or trade publication

Three-Lane Route Chooser

What do you need right now?

A weak search process repeats the same vague query and hopes the next page of results will somehow improve. A stronger process treats each search as information about what to try next.

A repeatable search cycle:

Search with two or three meaningful terms. Begin broad enough to discover vocabulary but focused enough to avoid a result ocean.
Scan before clicking. Read titles, dates, publication names, authors, abstracts, and result descriptions.
Collect useful vocabulary. Notice recurring terms, names, theories, locations, laws, organizations, and subject headings.
Revise the search. Add a narrower term, replace a weak term, try a synonym, remove a word, or change lanes.
Record what worked. Save strong search strings in your source log.

Useful Search Moves

Quotation marks search for an exact phrase: "bedtime procrastination".
AND requires both concepts in many databases: college students AND sleep quality.
OR searches for alternatives: TikTok OR "Instagram Reels" OR "YouTube Shorts".
Date filters help when currency matters.
Subject terms act like official database tags and often reveal the vocabulary of a field.
References lead backward, Cited by leads forward, and Related articles moves sideways.

When Results Are Too Broad

Add a population, place, time period, platform, discipline, or specific effect. Replace social media effects with something such as short-form video AND sleep quality AND college students.

When Results Are Too Narrow

Remove one concept, use a broader term, try a synonym, or search the references of the one useful source you found. Replace first-semester community college students TikTok bedtime procrastination Connecticut with college students social media bedtime procrastination.

When Sources Repeat One Another

Search for limitations, criticism, a different population, an alternative explanation, a historical comparison, a primary source, a systematic review, a counterargument, or the original dataset or study.

Key Takeaway

Research should deepen the conversation, not build a hall of mirrors.

SIFT is a practical method associated with digital-literacy educator Mike Caulfield. The letters name four moves: Stop, Investigate the source, Find better coverage, and Trace claims, quotations, and media to the original context.

S: Stop

Before reading every word or sharing the page, pause. Ask whether you recognize the source, what claim is being made, what your research purpose is, and whether the claim makes you eager to believe or reject it. Stopping does not mean becoming suspicious of everything. It means refusing to let the page set the pace of your judgment.

I: Investigate the Source

Find out who is behind the information. Examine the author or organization, relevant expertise, publication context, ownership, editorial process, purpose, and institutional role. Do not rely only on the source’s own About page. An organization is unlikely to describe itself as a beautifully designed fog machine.

F: Find Better Coverage

Search for other credible reporting, research, or analysis of the same claim. Look for stronger, clearer, or more complete coverage; sources with closer access to the evidence; and limitations missing from the first version you encountered.

T: Trace to the Original Context

Trace a statistic to the dataset or report, a quotation to the full interview or speech, a scientific claim to the study, an image to its earlier use and caption, and a policy claim to the actual law or rule. The original source is not automatically perfect, but it shows what was actually measured, said, shown, or decided.

SIFT Walkthrough: “Handwritten Notes Are Always Better”

Claim: “A scientific study proves that students who take handwritten notes always learn more than students who use laptops.”

Stop: The word always is unusually absolute. The post does not name the study, students, subject, or kind of learning.

Investigate: The account sells notebooks and study planners. That does not automatically make the claim false, but it reveals a commercial purpose and no obvious research expertise.

Find better coverage: Search for independent reporting, scholarly reviews, and later studies. Look for differences in task, population, distraction, accessibility, typing style, and review habits.

Trace: Follow the post to the blog, the blog to the news summary, and the news summary to the original study. Read the abstract, methods, and limitations. Evidence from a particular task under particular conditions is not an eternal law for every learner and every class.

The SIFT process does not require you to “debunk” the claim. It helps you replace an inflated claim with a more accurate understanding.

Key Insight: Tracing a claim often changes the question from “Is this true?” to “Exactly what does the evidence support?”

Close reading stays inside a text. You examine the writer’s words, evidence, organization, tone, and rhetorical choices. Lateral reading temporarily leaves the page. You open other tabs to investigate the author, publication, organization, evidence, and reputation from the outside.

Professional fact-checkers often resist spending too much time on an unfamiliar page before learning who is behind it. A beautifully organized article can consume twenty minutes before you discover that the “research institute” publishing it is a marketing project run by one person.

A short lateral-reading routine:

Open a new tab.
Search the author’s name, organization, and publication.
Read independent descriptions, not only the source’s biography or About page.
Look for relevant expertise, ownership, funding, editorial standards, corrections, and past work.
Search the central claim separately.
Return to the original page with better context.

What About Bias?

“Bias” is often treated as a trapdoor word: once someone spots a perspective, the source falls through the floor. That is too simple. Every source is created by someone, for some purpose, from some position.

Ask whether the purpose is transparent, the evidence fits the claim, sources and methods are visible, important context is omitted, fact and opinion are distinguished, and the source is appropriate for the job you need it to perform.

An advocacy organization’s page may be the best primary source for explaining that organization’s goals and language. It may be weaker for proving that all of its factual claims are correct. The source type and the research job must fit.

Read the Result Page, Too

Lateral reading begins before you click. Compare domains, titles, dates, and descriptions. This practice is sometimes called click restraint: pause long enough to choose a promising route instead of automatically entering the first result.

Writers often collapse two different questions into one: Is this source credible? and Is this source useful for my paper? They are related, but they are not identical.

Credibility

Credibility asks whether the source is trustworthy enough for the claim you intend to make.

author or organizational expertise;
publication context;
quality and transparency of evidence;
date and currency;
accuracy and traceability;
purpose and possible conflicts of interest;
whether other reliable sources support, complicate, or challenge the claim.

Usefulness

Usefulness asks whether the source provides something your argument needs.

a definition or historical background;
a statistic or concrete example;
a primary text for analysis;
a method or framework;
expert explanation;
a counterargument or limitation;
a different population or context;
evidence that complicates your first assumption.

	Useful for this paper	Not useful for this paper
Credible	Keep and identify its job.	Set it aside, even though it is a good source.
Not yet credible enough	Investigate further or trace to stronger evidence.	Do not spend more time unless your purpose changes.

A credible scholarly article about sleep patterns among retired adults may be excellent research but not useful for a paper focused on first-year college students. A current student newspaper story may not carry the authority of a peer-reviewed study, but it may provide a valuable campus example or lead to a policy document and interviews.

The one-sentence job test: I will use this source to ________________________________.

Strong answers name a task: define bedtime procrastination, provide data on college-student sleep, explain how algorithmic feeds encourage continued viewing, offer a counterargument, analyze a platform’s own language, or provide a local example. Weak answers include “get a quote,” “prove my point,” “because it is scholarly,” or “because it was the first result.”

A source without a job becomes decorative luggage. It makes the paper heavier without helping it travel.

A Rhetorical Reminder: Ethos, Logos, and Pathos

The rhetorical appeals you encountered earlier can help you examine a source, though they should not become a three-box scoring system.

Ethos

Ethos asks how the source establishes credibility. Consider expertise, experience, institutional role, fairness, transparency, and publication context. Relevant credentials, careful citation, acknowledgment of limitations, and a record of accurate reporting can strengthen ethos. A polished design, impressive title, or confident tone may create the appearance of ethos without providing much reason for trust.

Logos

Logos asks how the source supports its claims. Look for evidence, examples, data, methods, definitions, reasoning, and connections between claims. A source can contain statistics and still use them poorly. Numbers are evidence only when readers can understand where they came from and what they mean.

Pathos

Pathos asks how emotion shapes the audience’s response. Emotion is not automatically dishonest. A firsthand story may help readers understand the human stakes of an issue. The question is whether emotional language or imagery deepens the evidence or tries to replace it.

The Appeals Work Together

A public-health report may build ethos through expert authorship, logos through transparent data, and pathos through stories showing why the findings matter. Your task is not to count appeals. Explain how the source earns trust, supports its claims, and directs attention.

Key Insight: Rhetorical effectiveness and factual reliability overlap, but they are not identical. A source can sound highly persuasive and still require investigation.

Source categories are not a ranking from “bad” to “good.” They tell you something about a source’s audience, purpose, writing style, review process, and likely uses. The same topic may appear in a peer-reviewed journal, a newspaper, a professional magazine, a government report, and an advocacy website. Those sources participate in different conversations and do different jobs.

Academic and Scholarly Sources

Academic or scholarly sources are produced within colleges, universities, research institutions, or scholarly fields. They include peer-reviewed journal articles, academic books, book chapters, literature reviews, conference papers, dissertations, and some research reports.

They are usually written by scholars or researchers for readers who already have some knowledge of the field. Common features include specialized vocabulary, an abstract or formal introduction, engagement with previous research, a methods section when original research is involved, extensive citations, careful qualification, and publication through an academic journal, university press, scholarly society, or research institution.

Peer-Reviewed Sources

A peer-reviewed source is a particular kind of academic source. Before publication, specialists in the same or a closely related field evaluate the work. They may examine the research question, methods, evidence, reasoning, use of previous scholarship, and significance of the findings. The author may be asked to revise before acceptance.

Peer review is a form of scholarly quality control. It does not guarantee that a source is flawless, unbiased, current, or useful for your paper. Peer-reviewed studies can use limited samples, make mistakes, disagree, or later be challenged. Peer review tells you something important about how the source entered the academic conversation, not that the conversation has ended.

Scholarly and peer-reviewed are not perfect synonyms. An academic book from a university press may be scholarly without passing through the same journal process. A preprint may present serious research before formal review. A conference paper or dissertation may also be scholarly but require additional checking.

Key Insight: “Peer-reviewed” describes a review process. It does not mean “automatically correct,” and it does not tell you whether the source fits your particular question.

Popular Sources

Popular sources are written for a broad public audience. They include newspapers, general-interest magazines, news websites, podcasts, documentaries, and widely read online publications.

Popular sources often use accessible language, explain unfamiliar concepts, focus on current events or public consequences, use shorter articles and fewer formal citations, rely on journalists and editors, and use images, headlines, storytelling, and other features designed for public readers.

Popular does not mean unreliable. Strong journalism can provide timely reporting, local information, interviews, investigation, public response, and clear explanations of complex research. Distinguish reporting from editorials, opinion columns, sponsored content, press releases, and entertainment. Trace important research claims back to the original study whenever possible.

Trade or Professional Sources

Trade or professional sources are written for people who work in a particular occupation, industry, or professional field. Examples include publications for nurses, teachers, attorneys, engineers, architects, marketers, librarians, criminal-justice professionals, or hospitality managers.

These sources often assume knowledge of the profession, use field-specific vocabulary, discuss current practices and regulations, emphasize application, include case studies or industry data, and carry advertising connected to the profession. They can be excellent for understanding how people in a field describe their work and apply new policies or technologies. They may be less useful for proving a broad causal claim unless their evidence and methods are clear.

Feature	Peer-reviewed academic	Other scholarly or academic	Popular	Trade or professional
Primary audience	researchers and specialists	scholars, students, and specialists	general public	practitioners in a field
Typical author	researcher or scholar	scholar, academic expert, or research institution	journalist, editor, expert contributor, or public writer	practitioner, industry expert, editor, or professional organization
Review process	evaluated by disciplinary peers before publication	varies by format and publisher	editorial review, fact-checking, or platform-specific review	professional or editorial review; sometimes sponsored or industry supported
Evidence style	methods, data, citations, literature review	citations, scholarly analysis, archival or theoretical evidence	interviews, links, public documents, observation, summarized research	case studies, workplace data, regulations, professional experience, industry reports
Language	specialized and formal	often specialized and formal	accessible to broad readers	field-specific and practical
Strongest uses	original research, methods, disciplinary findings	theory, history, detailed analysis, scholarly context	current events, interviews, local examples, accessible explanation	current practice, professional trends, applied examples, industry language
Questions to ask	Is the study current, well designed, and relevant to my population and claim?	What scholarly process or publisher supports it, and how does it fit the field?	Is this reporting, opinion, sponsored content, or summary? Can I trace major claims?	Who supports the publication? Is the evidence independent, transparent, and appropriate?

How the Rhetorical Appeals Appear Across Source Types

A peer-reviewed article may build ethos through disciplinary expertise and review. Its logos may appear through methods, evidence, and qualified conclusions. Pathos is usually restrained, though the choice of problem can signal urgency or values.
A popular news feature may build ethos through a respected publication, named reporters, interviews, corrections, and links. It may use narrative and images for pathos, while its logos comes from documentation and explanation.
A professional publication may build ethos through practical experience. Its logos may rely on case studies, regulations, performance data, or workplace examples. It may appeal to professional values such as efficiency, safety, service, or innovation.

Do not assume that formal language equals strong ethos or that emotional language equals weak evidence. Examine what the source actually does.

Other Source Categories

Government and Institutional

Use laws, official policies, census or public-health data, institutional reports, and original public documents. Official sources show what an institution records or says; independent sources may still be needed to evaluate performance or interpretation.

Advocacy

Use advocacy sources to understand a group’s position, values, campaign language, proposed solutions, and claims requiring independent checking. Name the purpose and evaluate the evidence.

Primary

Primary sources are direct evidence from the person, event, organization, culture, or period being studied: speeches, interviews, advertisements, posts, policies, data, photographs, songs, and artifacts. Direct access does not guarantee an objective or complete account.

Reference and Background

Reference works, overviews, research guides, and encyclopedias help you learn vocabulary, names, dates, theories, and debates. They may guide research even when they do not appear in the final paper.

Key Takeaway

Source type shapes the kind of ethos, evidence, language, and audience a source is likely to use. Identify the type, then decide whether it fits the job your argument needs completed.

A source log prevents the familiar research disaster in which you remember seeing the perfect statistic but cannot remember where it lived, who published it, or whether the quotation in your notes was copied exactly.

A useful source log records both what the source is and what you plan to do with it.

Source-log fields:

author or organization;
title;
publication or container;
date;
stable URL, DOI, database permalink, or locator;
specific source type, such as peer-reviewed journal article, scholarly book chapter, popular news report, editorial, trade article, government dataset, advocacy brief, or primary source;
search lane and search terms;
one- or two-sentence summary;
important evidence and its page, paragraph, section, or timestamp;
credibility notes;
How I’ll use this source;
questions, limitations, or claims to check;
draft MLA Works Cited entry.

Completed Example

Author or organization: Michael A. Caulfield
Title: “Four Moves”
Publication or container: Web Literacy for Student Fact-Checkers
Date: 2017
Locator: Pressbooks chapter
Source type: Open educational resource; digital-literacy instruction
Search lane: Open web
Search terms: Mike Caulfield four moves lateral reading

Summary: Caulfield presents practical online fact-checking moves: look for previous work, trace information upstream, read laterally, and circle back when a route fails. The page notes that the approach was later streamlined into SIFT.

Credibility notes: Caulfield is the author of an openly licensed textbook on web literacy. The chapter directly explains the framework associated with his work, while Civic Online Reasoning materials independently emphasize similar lateral-reading practices.

How I’ll use this source: Define lateral reading and explain why source evaluation sometimes requires leaving the original page.

Question or limit: The 2017 chapter uses the earlier “four moves” wording, so the later SIFT terminology should also be represented accurately.

Draft MLA entry: Caulfield, Michael A. “Four Moves.” Web Literacy for Student Fact-Checkers, 2017, Pressbooks, https://pressbooks.pub/webliteracy/chapter/four-strategies/. Accessed 16 July 2026.

The field How I’ll use this source is the engine of the log. Without it, the source log is a storage closet. With it, the log becomes an early outline.

Keep Your Notes Honest

Use labels that separate direct quotation, paraphrase, summary, your reaction or analysis, and a question to investigate. Quotation marks in research notes are not decorative. They tell your future self which words belong to someone else.

Citation problems usually begin long before the Works Cited page. They begin when a writer saves a quotation without a page number, copies a temporary database-session URL, forgets the author, or cannot tell whether the browser-tab title belongs to the article or the larger website.

Save citation details while the source is open. Depending on the source, record the author or organization, source title, container title, contributors, version, volume and issue, publisher, date, page range, DOI or stable URL, and access date when appropriate.

MLA builds Works Cited entries from core elements. A container is the larger whole that holds the source. An article may be contained in a journal; a page may be contained in a website.

Source: Michael A. Caulfield’s chapter “Four Moves” from Web Literacy for Student Fact-Checkers.

Author: Michael A. Caulfield
Title of source: “Four Moves”
Title of container: Web Literacy for Student Fact-Checkers
Publication date: 2017
Platform or publisher: Pressbooks
URL: https://pressbooks.pub/webliteracy/chapter/four-strategies/
Access date: 16 July 2026

Draft MLA entry:
Caulfield, Michael A. “Four Moves.” Web Literacy for Student Fact-Checkers, 2017, Pressbooks, https://pressbooks.pub/webliteracy/chapter/four-strategies/. Accessed 16 July 2026.

If your sentence names the author, you may not need a parenthetical author reference: Caulfield describes lateral reading as investigating what other sources say about an unfamiliar publication.

If your sentence does not name the author, use the shortest information directing readers to the Works Cited entry: Lateral reading asks researchers to investigate a publication through outside sources rather than relying entirely on the publication’s self-description (Caulfield).

When a source has stable page numbers and you quote or paraphrase a specific passage, include the page number. For an unpaginated webpage, do not invent one.

Citation Generators Are Drafting Tools

Database and Google Scholar citation buttons can save time, but treat the output as a draft. Check capitalization, italics, quotation marks, authors, dates, duplicate publisher information, URLs or DOIs, source and container titles, and whether the citation matches the version you actually used.

Key Takeaway

A generator can arrange information. It cannot always understand the source correctly.

Source evaluation is rarely a simple contest between “good” and “bad.” A better triage system asks whether a source is credible enough, useful enough, and appropriate for the paper’s purpose.

Keep: credible enough and clearly useful;
Maybe, investigate further: potentially useful, but the source or claim needs more checking;
Not for this paper: irrelevant, redundant, too weak for the intended claim, or better replaced.

Practice research question: Should college instructors require handwritten notes, allow digital notes, or let students choose based on the task and their learning needs?

The source cards below are fictionalized for practice. Consider credibility, usefulness, source type, ethos, evidence or logos, emotional framing, and the specific job each source could perform.

Source A: Peer-Reviewed Laboratory Study

A 2014 peer-reviewed psychology study compares students who took longhand notes with students who typed notes during controlled lectures. It reports differences in verbatim copying and performance on some conceptual questions. The sample consists mainly of students at selective four-year institutions.