Every research project begins with a search, and every search happens inside an information ecosystem. If you do not understand the ecosystem, you cannot search it well. You may find the wrong kind of source for your question. You may trust a source that looks credible but was produced without accountability. You may miss excellent evidence simply because you did not know where to look.

Consider two students working on the same research question: What are the long-term health effects of air pollution in urban areas?

The first student types the question into a general web search engine and reads the first three results — a news article summarizing a recent study, a blog post written by an individual with no stated credentials, and a sponsored page from an air purifier company. She treats these as roughly equivalent because they all appeared on the same page of results.

The second student knows that this question involves scientific and medical evidence, so he begins with a subject-specific academic database. He finds peer-reviewed studies published in public health journals. He also checks government health agency reports for official statistics. When he reads a news article summarizing a study, he knows to find the original study and read the actual data.

Both students searched. Only one understood the information ecosystem well enough to search it responsibly.

The word ecosystem comes from biology. In ecology, an ecosystem is a community of living organisms interacting with each other and with their physical environment — plants, animals, soil, water, sunlight, and the relationships among them. Remove one element and the entire system shifts.

An information ecosystem works by similar principles. It is the whole system of information sources, producers, distributors, platforms, institutions, technologies, and users that together create, store, circulate, and consume information in a society.

Every source you find during research is a product of this ecosystem. A peer-reviewed journal article was shaped by academic peer review norms, university research infrastructure, publisher economics, and the conventions of a specific discipline. A tweet was shaped by the constraints of a commercial social media platform, its recommendation algorithm, and the social dynamics of the user's network. A government statistical report was shaped by agency methodology, political oversight, data collection resources, and reporting requirements.

None of these sources simply "contains information." Each one was produced by a system, and that system left marks on the source — in what it includes, what it excludes, how it is framed, and who it was intended to reach.

Ecosystem thinking helps researchers ask better questions about sources. Instead of asking only Is this true?, a researcher who understands information ecosystems also asks who produced this and for what purpose, what system the source exists within, what incentives shaped its creation, what standards governed its production, who the intended audience was, what the source type does well and where its structural limits lie, and what other parts of the ecosystem might give a different or more complete picture.

These are not suspicious questions. They are not signs of distrust or cynicism. They are the questions that any responsible researcher asks, regardless of whether a source turns out to be reliable or not. Understanding the ecosystem is how you assess each source fairly and accurately.

For most of recorded history, the information ecosystem was relatively small and slow. Books took years to write and expensive processes to reproduce. Access to libraries was limited. News traveled by physical mail. If you wanted information on a specialized topic, you needed a specialist.

The modern information ecosystem is orders of magnitude larger, faster, and more complex. A single internet search may return millions of results within a fraction of a second. A single social media post can reach millions of people within hours. AI tools can generate thousands of words of text in seconds, drawing on vast amounts of training data. Anyone with internet access and a free account can publish information globally without any editorial review.

This abundance creates opportunity: more information than ever before is accessible to more people. It also creates serious challenges: more misinformation, more noise, more sophisticated manipulation, and more difficulty distinguishing reliable from unreliable sources. The skills taught in this textbook — and especially the ecosystem awareness taught in this chapter — exist precisely because navigating this complexity requires deliberate skill, not just access.

Books are among the oldest and most trusted sources in the information ecosystem. For centuries, they were the primary vehicle for communicating sustained, organized, and deeply developed knowledge. Understanding what books are and how they work as research sources is still essential, even in a digital age when a great deal of information arrives in shorter formats.

A book, in the research sense, is a substantial work of written content that develops a subject at length. Books may be monographs — scholarly books written by one author or a small group presenting original research, analysis, or argument on a specific topic; edited collections, in which a scholar or team invites multiple experts to contribute chapters on related topics and then edits and compiles the result; textbooks, written primarily to support instruction and typically surveying a field rather than presenting original research; handbooks and companions, comprehensive reference volumes covering the full scope of a discipline; and trade books, written for general audiences rather than academic specialists.

Books go through processes that vary significantly by type. Academic monographs typically go through peer review — expert readers who evaluate the scholarly quality of the manuscript before an academic press accepts it for publication. Trade books go through editorial review by publishing house editors, but not through the same kind of expert scholarly review. Textbooks are reviewed for pedagogical accuracy and appropriate level, and often go through multiple editions as knowledge in a field updates.

Reference works are books designed not to be read cover to cover but to be consulted for specific information.

Books are especially valuable for research that requires depth — a book-length treatment can develop complex arguments, present extensive evidence, and explore a subject from multiple angles in ways no article can match. They are valuable for context, since foundational arguments, theoretical frameworks, and the intellectual history of a debate are often laid out most fully in books. And reference works are invaluable early in the research process for building the background knowledge needed before moving to specialized sources.

The MGU Library makes books available in physical and digital formats. The library catalog allows you to search by title, author, subject, keyword, and call number. Subject headings in the catalog link you to related books in the same area. The encyclopedia section of the MGU Library is particularly valuable for reference works useful in the early stages of research.

Scholarly articles are the primary vehicles for communicating original research across most academic disciplines. Understanding how they are produced, where they are found, and what they can and cannot do is one of the most important research skills a university student can develop.

A scholarly article (also called an academic article, peer-reviewed article, or research article) is a relatively short, focused piece of writing — typically between a few thousand and twenty or thirty thousand words — that reports on original research, advances a scholarly argument, or reviews the existing literature on a topic. Scholarly articles are published in academic journals: periodicals published on a regular schedule, often monthly or quarterly, that specialize in a particular discipline or sub-field.

The defining characteristic of scholarly articles, and the feature that distinguishes them most sharply from other types of writing, is peer review. When a researcher submits an article to a peer-reviewed journal, the editor sends it to two or more expert reviewers — other researchers in the same field — who evaluate the work without knowing the author's identity. In many journals, the review is double-blind: neither the author nor the reviewers know each other's identities.

Peer review is not perfect. Reviewers have their own biases. Important work is sometimes rejected; flawed work sometimes passes. High-profile journals have retracted articles that turned out to be fraudulent or methodologically unsound. But peer review is the most rigorous systematic quality-control mechanism that exists in academic knowledge production, and a peer-reviewed article carries more evidentiary authority than a source that has undergone no comparable review.

Scholarly articles are almost never freely accessible through a general web search. They are indexed and often stored in academic databases: specialized search systems that index thousands of journals in particular disciplines or across disciplines generally.

Access to most academic databases requires either a subscription, paid by a library on behalf of its users, or a fee per article. The MGU Library provides access to databases appropriate for the academic programs it supports. When you search through the MGU Library's database access points, you are searching curated, scholarly collections — very different from what a general web search returns.

Not all information comes from academic institutions. A large and vitally important part of the information ecosystem is produced by journalists, editors, commentators, and media organizations working outside the academy. Understanding this zone of the ecosystem — its strengths, its variations, and its limitations — is essential for any researcher.

News reporting is the attempt to inform a general audience about recent events, conditions, or issues. Professional journalism at its best is reported by journalists with training in finding, verifying, and interpreting information; based on interviews, documents, official records, and on-the-ground observation; subject to editorial oversight that checks accuracy before publication; transparent about sources to the extent possible while protecting confidential sources; and corrected when errors are discovered.

Quality news reporting is an essential part of the information ecosystem. It documents events as they happen, creates a record of public life, holds institutions accountable, and covers topics — from local government decisions to international crises — that academic researchers cannot address in real time.

News reports are particularly valuable in research for documenting that events occurred and when, capturing public reactions and debates at the time of an event, providing details about fast-moving situations before academic analysis catches up, and tracking policy changes, legal decisions, and institutional actions.

However, news reporting also has limitations. Journalists work under deadline pressure. Coverage is often incomplete, based on early information that may later be revised. Some news organizations have editorial biases. The business model of many news outlets — advertiser-supported or subscription-based — shapes what gets covered and how. Sensationalism sells; nuance often does not.

Magazines occupy a middle ground in the information ecosystem. They range from general-interest publications, which cover current events, culture, science, and politics for broad audiences, to specialized publications targeting specific professions or interest communities.

Feature journalism takes more time and depth than daily news reporting. A feature article in a quality magazine may take weeks or months to research and write, involving extensive interviews, document review, and analysis. Well-researched feature journalism can be an excellent source for understanding the human dimensions of a story, the broader context of an issue, or the perspectives of people affected by a situation.

Trade magazines and professional publications — which circulate within specific professions or industries — often report information not covered by general media, including industry data, regulatory developments, and professional debates.

Popular media sources — news articles, magazine pieces, blog posts — have characteristics that researchers must keep in mind. They are typically written for general audiences, which means they simplify complex topics. They are rarely subject to the kind of expert review that academic sources receive. They may reflect the editorial perspective of the organization that publishes them. They are often written quickly, increasing the risk of errors. And advertising models can create pressure to generate clicks rather than ensure accuracy.

The open web — everything publicly accessible through web search engines — is one of the most complex and variable zones of the information ecosystem. It contains some of the most valuable information available, and also some of the least reliable.

The open web includes government websites and official agency pages; university and educational institution websites; research organization and think tank websites; nonprofit organization websites; news and media organization websites; corporate and business websites; personal and individual websites; blog platforms and independent publishing sites; community wikis and collaborative information sites; open access academic repositories and preprint servers; health information portals; and legal information databases.

The range is extraordinary. A government statistical agency website offers rigorously collected official data. A personal blog offers one individual's perspective, verified by no one. A university research center website may offer expert analysis. A corporate website offers promotional material written to serve the company's interests. A Wikipedia article offers a collaboratively edited summary that may be accurate, outdated, or somewhere in between.

The web address, or URL, of a website provides some limited information about its type.

An important development in the information ecosystem is the open access movement: the effort to make scholarly research freely available online rather than locked behind subscription paywalls. Open access takes several forms: gold open access, in which articles are published immediately as freely available on the publisher's website, sometimes funded by author fees; green open access, in which authors deposit pre-publication or accepted versions of articles in institutional repositories or subject repositories such as PubMed Central; and preprint servers, platforms where researchers post papers before peer review, such as arXiv for physics and mathematics or medRxiv for health research.

Because the open web is so variable, evaluating websites requires deliberate attention to several questions: Who runs this website, and what is their purpose? Who wrote this specific page or document, and what are their credentials? What evidence does the page cite, and can those sources be traced? When was the page last updated? Does the information agree with what other reliable sources say? Does the site disclose its funding, affiliation, or agenda?

Chapters 25 and 26 address source evaluation in detail. For now, the key point is that open web sources cannot be evaluated by appearance alone. A professional-looking website may contain misinformation. A plainly formatted page from a government archive may contain exactly the data you need.

Social media platforms now occupy an enormous portion of the modern information ecosystem. Billions of people use platforms such as Facebook, Instagram, X, YouTube, TikTok, Reddit, LinkedIn, and dozens of others to produce, share, and consume information. Understanding the social media zone of the ecosystem is essential — both because it is where much of the information people encounter now lives, and because it presents some of the most significant challenges to responsible research.

Social media platforms are primarily built to facilitate social interaction and the sharing of user-generated content. The information that appears on these platforms includes personal posts, opinions, and commentary; news links shared and commented on by users; videos from individuals, organizations, news outlets, and entertainment creators; discussion threads and community forums; professional updates and industry commentary; real-time reporting from the scene of events; humor and satire; advocacy and activist organizing; misinformation, disinformation, and propaganda; and advertising and sponsored content.

The information that appears in a social media feed is not neutral or comprehensive. It is shaped by recommendation algorithms: automated systems that predict what content each user is likely to engage with and prioritize that content accordingly.

These algorithms are designed to maximize engagement — not to maximize accuracy, diversity, or the quality of information. Content that produces strong emotional reactions, such as outrage, excitement, fear, or humor, often receives more algorithmic promotion than calm, measured, accurate reporting. This creates systematic pressure in the social media information environment toward sensationalism and emotional content, even when that content is misleading.

Despite its challenges, social media and user-generated content can be legitimate research sources in specific contexts. Social media posts from participants in events, protests, or public discussions can serve as valuable primary source documentation for historians, sociologists, and communication researchers studying those events. Analyzing patterns in social media content with appropriate methodology can reveal something about public attitudes, the spread of narratives, or the organization of social movements. And during breaking events, social media often carries eyewitness reporting before traditional news organizations have coverage — though this material requires careful verification.

When using social media content as a research source, the same evaluation questions apply: Who created this? What is the purpose? What evidence is offered? How can it be verified?

Wikipedia deserves specific attention because it occupies an unusual position in the information ecosystem. It is a collaborative, online encyclopedia written and edited by a volunteer community of contributors, governed by policies about verifiability, neutral point of view, and reliable sourcing.

Wikipedia is not a reliable academic source for direct citation in most research contexts. Its articles can be edited by anyone, may contain errors, and may lag behind current research. Disciplines, departments, and instructors almost universally ask students not to cite Wikipedia as a source.

Information does not only come in the form of written text. A full understanding of the modern information ecosystem requires recognizing that data sets, photographs, maps, audio recordings, video footage, and other nontext sources are legitimate and often essential forms of evidence.

A data set is a structured collection of numerical or categorical information, typically organized in rows and columns, that can be analyzed to identify patterns, trends, or relationships. Data sets are produced by government statistical agencies, research institutions and universities, international organizations such as the World Bank, the World Health Organization, and the United Nations, private research firms and companies, nongovernmental organizations, and individual researchers sharing their data publicly.

Data is increasingly available for research purposes. Government open data initiatives have made vast amounts of official statistics publicly accessible. Research data repositories allow scholars to share the data underlying their published findings.

Evaluating data requires attention to the same questions raised for other sources, plus additional technical considerations: How was the data collected? What is the sample size and sampling method? What are the definitions used for key variables? Who funded the data collection, and could that affect what was measured or how? Chapter 29 covers data and statistical evidence in detail.

Visual images — photographs, illustrations, diagrams, charts, graphs, and infographics — are a significant category of sources and evidence. Images can function as documentary evidence that records events, conditions, or people at a specific moment; as scientific visualization that represents data or complex relationships; as historical records that provide direct visual evidence of the past; or, when edited, misattributed, or taken out of context, as propaganda and manipulation.

The research value of visual sources depends heavily on context. A photograph of a protest taken in 1968 is a primary historical source — when we know where and when it was taken, by whom, and in what circumstances. The same photograph shared on social media with a misleading caption claiming it depicts something else entirely becomes a piece of disinformation. Chapter 30 covers visual evidence evaluation, including reverse image searching and deepfake detection, in detail.

Maps are a specialized form of visual information that represents geographic, spatial, or thematic data. Maps produced by government agencies, academic institutions, and reputable research organizations can be excellent sources of geographic, demographic, environmental, and historical information. They can also be tools of manipulation: the choices of what to include, what to exclude, how to draw boundaries, and what colors and scales to use can present the same geographic reality in very different and potentially misleading ways. Evaluating maps involves the same attention to producer, purpose, and methodology as any other source.

Audio recordings — interviews, speeches, broadcasts, oral histories, music — are primary sources in many research contexts. Video footage — news broadcasts, documentary films, recorded events, personal recordings — similarly functions as direct documentary evidence.

Legitimate audio and video sources accessible to researchers include broadcast news archives, oral history collections in libraries and archives, congressional and parliamentary records, recorded interviews published by reputable organizations, and documentary films, which themselves should be evaluated for perspective and accuracy.

Archives are collections of primary source documents — letters, diaries, official records, photographs, maps, audio and video recordings, organizational records, and other materials — preserved for historical and research purposes. Academic libraries often maintain special collections and archives alongside their general holdings.

Archives hold evidence that cannot be found anywhere else: the original manuscript of a novel, the internal correspondence of a government agency, a collection of personal letters that documents daily life in a specific era, official records of a historical event. For researchers working in history, literary studies, political science, biography, and related fields, archives are often essential.

Many archives have begun digitizing their holdings and making them accessible online. But a great deal of archival material remains only in physical form, accessible only by visiting the archive in person.

The emergence of large language model AI tools — software that can generate human-like text, answer questions, summarize documents, and assist with writing — has added a fundamentally new element to the information ecosystem. As of 2025, tools such as chatbots, AI writing assistants, and AI-powered research tools have become widely available and widely used by students, researchers, professionals, and the general public.

Understanding what these tools are, how they work, and what they can and cannot do responsibly is now an essential part of information literacy.

AI language models are software systems trained on enormous amounts of text data — books, articles, websites, and other written sources — to learn statistical patterns in language. When prompted with a question or instruction, these models generate responses by predicting what words would be most appropriate to follow, based on those learned patterns.

This means that AI language models do not "know" things in the way humans know things. They do not consult databases or look up current information unless specifically integrated with search tools. They generate text that is statistically likely to be useful, coherent, and relevant — but this generation process does not guarantee accuracy. An AI tool can produce confident, well-structured, grammatically correct text that contains significant factual errors.

One of the most significant problems with AI-generated information in research contexts is a phenomenon called hallucination: the tendency of AI language models to generate plausible-sounding but fabricated information, including entirely fictional citations.

Despite these significant limitations, AI tools can be genuinely useful at certain stages of the research process when used thoughtfully and with appropriate verification. They can assist with brainstorming and exploration, helping you think through the dimensions of a topic or generate a preliminary list of questions. They can offer plain-language explanations that build background understanding before you read primary and secondary sources. They can provide drafting assistance, helping you think through organizational structures or giving feedback on draft writing — though the ideas and evidence must still be yours, and disclosure requirements apply. And they can help with identifying search terms, brainstorming keywords and related terms for database searching, though you must verify these against controlled vocabulary in actual databases.

The question of when and how to disclose AI tool use in academic work is an evolving and important area. Different institutions, departments, and individual instructors have different policies on AI assistance. Some assignments prohibit AI assistance entirely. Some permit it for specific stages of the process, such as brainstorming or editing, but not others, such as generating the argument or the evidence. Some require explicit disclosure of any AI use.

The rule is: always follow the specific disclosure and use policies of your institution, your department, and your instructor for the assignment you are completing. When in doubt, ask your instructor before using AI tools, not after. Chapter 45 addresses responsible AI tool use in research in detail.

AI tools have introduced a new kind of information into the ecosystem — information that is generated by software rather than researched, observed, or analyzed by human experts. This means that when information from AI tools circulates without attribution, for example when text generated by an AI tool is presented as original research, it adds a layer of unverifiable, unaccountable content to the information environment.

This is not an argument against AI tools. It is an argument for understanding what they are. An AI language model is not a researcher, a journalist, a scientist, or an expert. It is a sophisticated pattern-matching system that generates text. The ecosystem benefit of such tools depends entirely on whether the humans using them understand that distinction and act accordingly.

The following table summarizes the key characteristics of the major information ecosystem zones covered in this chapter.

One of the key lessons of this chapter is that different zones of the information ecosystem serve different research purposes. No single zone is always better than all others. A peer-reviewed article is the right source for many research questions but the wrong source for finding the text of a federal law. A government website is the right source for official demographic data but the wrong source for a peer-reviewed analysis of what that data means.

The skill researchers need is not just finding sources, but matching source types to research needs. The following examples illustrate this matching process.

This ecosystem map will help you build a research strategy that uses the right parts of the information environment for your specific question.

Every source you use in research must be cited, but different source types require different information to be recorded. When you encounter a source, make sure to record all elements you will need for a citation.

A key benefit of using the MGU Library rather than relying solely on general web search is that the library has already applied curation and quality control. The databases accessible through the library index peer-reviewed journals. The reference works on the library's shelves and in its digital collections have been selected by professional librarians based on quality and relevance. The MGU Library does not guarantee that every source you find through it is accurate — evaluation is still your responsibility — but it does mean you are starting in a part of the information ecosystem that has been filtered for scholarly reliability.