Finding an old letter, photograph or newspaper page can feel like finding the answer to a history assignment. Often, it is the beginning of the work. The item may contain valuable evidence, but a persuasive paragraph still has to explain what that evidence supports, what it leaves uncertain and why it matters to the question being asked.
This guide presents a practical way to organise that work. It is designed for a student beginning a small history project, with examples that can be adapted to a teacher's assignment. It does not supply a ready-made interpretation of a real historical event. The classroom example below is entirely fictional, so its details should never be cited as historical facts.
The goal is a set of notes that makes your reasoning visible. A reader should be able to trace a claim back to a particular part of a source and understand how you moved from one to the other. A longer bibliography cannot replace that connection, and a confident writing style cannot repair a claim that the evidence does not establish.
Use an established analysis framework as a starting point
The [Library of Congress teacher's guide](https://www.loc.gov/static/programs/teachers/getting-started-with-primary-sources/documents/Analyzing_Primary_Sources.pdf) encourages observation, reflection, questioning and further investigation. It treats these activities as connected rather than a rigid one-way sequence. Its suggestions include testing ideas against additional sources and revising conclusions as the evidence develops.
The [National Archives document-analysis resources](https://www.archives.gov/education/lessons/worksheets) provide worksheets for different kinds of material and learner levels. They move from identifying a document and examining its parts toward interpreting it and using it as historical evidence. The resources include a tool specifically concerned with perspective.
These are useful teaching frameworks, not automatic answer generators. The workflow in the rest of this article is an original practical adaptation built around a fictional school-project example. Follow your teacher's instructions where the required method, citation style or assessment criteria differ.
Begin with a question small enough to investigate
Suppose a class is studying changes in a fictional town called Alderbridge. A student starts with the question, What was life like in the past? That question is too broad for a short assignment. It does not tell the student which people, period or aspect of life to investigate, and almost any interesting source could seem relevant.
A more workable question might be: what concerns did the organisers of a fictional evening reading room express about its opening hours? This asks about a defined group and a defined issue. It also leaves room for an answer that is limited or uncertain. The project does not have to prove that the reading room transformed the entire town.
Write your working question at the top of your notes. Underneath it, list what kind of evidence might help. An announcement could show advertised hours. Meeting minutes might record organisers' discussion. A visitor's letter could provide one person's response. Those possible sources would do different jobs, even if they referred to the same institution.
Do not decide your final conclusion before seeing what is available. A working question helps you choose where to look; it should not become a demand that every source confirm your first idea. If the surviving evidence answers a narrower question, changing the question can be a sensible improvement.
Save the identity of the item before extracting quotations
Create a source card with the item's title or description, creator if known, date as supplied, holding institution, catalogue identifier and a stable link. Add the date you accessed it. If you are looking at a scan, note the page or image number you actually used rather than only the collection's home page.
Keep the archive's description separate from your own interpretation. A catalogue may give a tentative date or an attributed creator. Preserve that qualification. Replacing probably with definitely in your notes would make your later essay more certain than the record you consulted.
For a transcription, record whether you made it yourself or used one supplied by the institution. If a word is unreadable, mark it as unclear instead of quietly filling the gap with the word that best fits your argument. You may later resolve it by comparing another page or asking for help, but the uncertainty should remain visible until then.
The [National Archives research starting page](https://www.archives.gov/education/research/primary-sources) points students toward its catalogue, online exhibits and research tools. It is a directory for beginning a search, not proof that every linked resource is complete or appropriate for every project. Check the individual item and its description when choosing evidence.
Give observations and interpretations different spaces
Divide your notes into three areas: what the item contains, what you think it may mean, and what you still need to find out. The headings can be simple. Their purpose is to prevent a plausible explanation from becoming indistinguishable from something directly visible on the page.
In the fictional Alderbridge exercise, imagine an invented notice that lists Tuesday and Thursday evening opening hours. The observation is that those two evenings are named. A possible interpretation is that the organisers expected evening attendance. A further question is why those evenings were chosen. The notice alone does not answer that last question.
Now imagine that the notice also invites donations of lamps. You can record the request as part of the invented document's content. You cannot jump straight to the claim that the building had no lighting at all. The organisers might have wanted more lamps, replacement lamps or lamps for another purpose. Those possibilities are questions to investigate, not facts to insert into the essay.
This separation does not prohibit interpretation. Interpretation is part of the work. It simply makes clear which step requires explanation and which additional evidence could strengthen or challenge it. Your notes become easier to discuss because a teacher can see exactly where the argument changes from description to inference.
Ask what the source was trying to do
A source was usually created for some purpose other than helping a future student finish an assignment. In the fictional example, a public notice might be intended to attract visitors or donations. Private minutes might be intended to record decisions. A personal letter might describe one person's experience for a particular recipient.
Write a short purpose statement, but avoid pretending you can read the creator's mind. Base it on the form, wording and available context. If the intended audience is uncertain, say so. An interpretation becomes more useful when it names the evidence behind it rather than relying on a label such as obviously biased.
Perspective does not automatically make an item useless. An enthusiastic announcement may tell you a great deal about how organisers wanted an activity to be seen, while telling you much less about whether visitors agreed. The question is what the source can support, not whether it can serve as an all-purpose account of reality.
For the Alderbridge project, the student might write that the invented notice presents the reading room as welcoming. That is a claim about the notice's presentation. A claim that every resident felt welcome would require evidence about residents' experiences, including whose voices are available and whose are missing.
Compare sources at the level of a specific claim
Adding a second item is most helpful when you know what you are comparing. Put the claim in the first column of a small table. In the next columns, record what each source contributes. Add a final column for remaining uncertainty. This can be done in a notebook; a spreadsheet is not necessary.
For example, the fictional notice advertises two evenings, while invented committee minutes discuss a proposal for three evenings. These statements need not contradict each other. One might describe a proposal and the other the schedule eventually announced. The dates and wording would matter before you could decide how the two fit together.
If two sources really disagree, preserve the disagreement. Ask whether they were created at different times, describe different people or use terms differently. Do not choose the version you prefer merely because it gives your essay a cleaner storyline. A carefully explained conflict can be part of an honest answer.
Also check whether two items are independent. A later summary that repeats an earlier announcement does not necessarily provide a separate confirmation of the event it describes. Your notes should indicate when one account appears to depend on another. The number of links in a bibliography is not the same as the number of independent observations.
Build a paragraph around a limited claim
A useful draft paragraph has a claim, a precise piece of evidence, an explanation and a boundary. The boundary tells the reader what the evidence does not establish. This structure is a writing aid, not a requirement that every paragraph contain exactly four sentences.
Here is a fictional example: the reading-room organisers planned some evening access. The invented public notice names Tuesday and Thursday evening sessions. That detail supports a claim about the advertised schedule. It does not show how regularly the sessions occurred or how many people attended them.
Notice how modest the conclusion is. It is still informative because it answers a defined question. If later evidence supports a wider conclusion, the argument can grow. Starting with a limited claim makes it easier to identify what additional evidence would actually change the answer.
When you include a quotation in a real assignment, copy it accurately and use the citation format your teacher requires. Explain why the quoted detail matters. A quotation placed between two broad assertions does not automatically support either of them. The connection belongs in your own reasoning, not in the reader's guesswork.
Keep a research log that records decisions
At the end of a session, write a few lines about what you searched, what you found and what you will do next. Include unsuccessful searches when they explain a gap. This can prevent you from repeating the same search next time and helps you distinguish a missing result from a source you simply have not examined yet.
A useful log entry might say that you found an announcement but still need evidence of attendance. A less useful entry says only that research is finished. The first entry describes the state of the argument; the second hides the unresolved question.
If a web page becomes unavailable, keep the citation details you already recorded and note the access problem. Do not manufacture a replacement quotation from memory. Depending on the assignment, you may need to choose another accessible item or explain the limitation to your teacher. Preserve local notes and permitted copies without treating them as proof of details you never checked.
Set a stopping rule for the current assignment. You might decide to stop searching once you have enough material to answer the narrow question and explain the main limitation. This is different from claiming you have found every relevant source. A school project can be complete within its stated scope while the broader historical subject remains open.
Review the draft for claims that outrun the evidence
Read each paragraph and underline words such as everyone, always, proved and caused. They are not automatically wrong, but they often carry more weight than a small source collection can support. Ask what evidence would be necessary for each word to be justified in the particular sentence.
Check dates and categories as carefully as quotations. A proposed policy is not the same as a policy in operation. An advertised service is not the same as a service used by every eligible person. An individual's account should not silently become the experience of a whole population.
Then ask a classmate to identify your main claim and the source supporting it. If they cannot find the connection, revise the explanation rather than simply adding another quotation. You can also perform this check yourself after a break, using only the draft and the source cards.
Finally, remove any fictional practice details from a real historical submission. Alderbridge and its reading room exist only as teaching examples in this article. Their purpose is to demonstrate reasoning, not to provide content for a factual essay. Replace them with verified material from your own assigned topic.
Conclusion: show the path from item to argument
Good source notes preserve more than interesting extracts. They keep track of an item's identity, distinguish observation from interpretation and show where the argument still needs support. That makes the eventual essay easier to check and easier to improve.
Start with a manageable question, keep the source's details intact and write claims at the scale the evidence allows. A carefully explained, limited conclusion is more useful than a sweeping story built from uncertain steps. The result is a piece of work whose reasoning you can discuss, not merely a document you can submit.
Sources and research note
Research checked October 27, 2026. Primary teaching resources consulted: the [Library of Congress analysis guide](https://www.loc.gov/static/programs/teachers/getting-started-with-primary-sources/documents/Analyzing_Primary_Sources.pdf), [National Archives document-analysis resources](https://www.archives.gov/education/lessons/worksheets) and its [primary-source research directory](https://www.archives.gov/education/research/primary-sources). The directory displays a 2018 review date; individual resources should be checked separately. The Library's landing page was unavailable during retrieval, but the linked official PDF was read directly. Framework summaries are brief; the workflow and fictional examples are original editorial guidance. No classroom trial or learning-outcome guarantee is claimed. The cover is an original AI-generated conceptual illustration with invented papers, not a reproduction of historical evidence.