Focus your AG preparation on three demonstrated skills: knowing one accreditation region's records thoroughly, classifying evidence correctly, and defending a written proof argument aloud. Use the readiness checklist at the end of this guide as your milestone check before applying.
Choosing and Building Depth in Your Accreditation Region
ICAPGen's accreditation process is organized around a geographic region of specialization, so your preparation should center on one region's record landscape, boundaries, and repositories. Confirm the current region definitions and application details directly on the ICAPGen website before committing.
Region depth is not memorizing a list of famous archives. It means knowing which record types exist in your region, when each type begins, what survival gaps and boundary changes matter, and which repositories hold which originals. For example, preparing in a colonial American region means understanding parish and county record survival, land and probate patterns, and how jurisdictional lines shifted across the period you will research.
Compare candidate regions honestly using two criteria: what you already research and what you can physically or digitally access. A region you know moderately well but can reach is usually a better base than a region you have only read about. Build a region notebook that grows weekly; it becomes both your study tool and the backbone of your work samples.
- Region notebook, tab 1: record types with starting dates, gaps, and current custodians
- Tab 2: county or district boundary changes relevant to your research period
- Tab 3: migration and settlement patterns that explain where records were created
- Tab 4: repository inventory with what you have actually used and what remains
Telling Sources, Information, and Evidence Apart in Your Own Work
Evidence analysis separates three layers: the source, the information within it, and the evidence it offers for your question. Practice classifying all three for every record you touch, because a source's format and an informant's knowledge are different judgments.
A common conceptual slip is collapsing these layers into one quality rating, for example calling a death certificate simply reliable or unreliable. The certificate may be an original source, contain primary information about the death itself, and yet carry secondary information about the deceased's birthplace supplied by a relative. Each layer gets its own assessment, and each triggers a different research action.
The third layer, evidence, depends on your question rather than the record. A census entry gives direct evidence of residence, indirect evidence of a marriage when a spouse appears in the household, and no evidence at all for many other questions. Working examples help: take one record and write one sentence per layer, then note what search the classification sends you to next.
| Layer to classify | What you are judging | Action the classification triggers |
|---|---|---|
| Source: original or derivative | The record itself — an original deed book page versus a transcribed index of it | Seek originals where possible; for derivatives, identify who created them and what copying errors they may carry |
| Information: primary or secondary | How the informant knew the fact — the bride herself on a marriage record versus her grandson decades later | Weight each fact by the informant's proximity to the event and find a source with a closer informant |
| Evidence: direct, indirect, or negative | What the item offers for your specific question — a stated fact, an implication, or a documented absence | Correlate indirect items into one argument; report negative searches with their scope as findings, not failures |
Worked Scenario: Resolving Three Conflicting Birth Dates in a Proof Argument
When sources conflict, a proof argument weighs each item using the source, information, and evidence layers, correlates what agrees, and states a conclusion with reasoning. Practice writing these arguments in a few hundred words, not pages.
Scenario: your subject's birth year appears as 1848 in census returns, 1850 in a family Bible, and 1847 on the death certificate. A plausible mistake is treating the death certificate as authoritative because a government office created it, then writing around the conflict. The better decision is to classify first: the certificate is an original source, but the birth date is secondary information, likely supplied by a grieving adult child with no firsthand knowledge.
The Bible entry, though from a derivative transcription you located, may record primary information entered near the event. The census returns are secondary for age but were given by a household member close to the event. A sound argument correlates the Bible and census agreement around 1848 to 1850, explains why the certificate date carries the least weight, and states a reasoned conclusion while flagging what further record — such as a baptismal entry — could strengthen it. This matters because examiners evaluate your reasoning process, not just your answer.
- Step 1: classify each source, informant, and evidence type before comparing dates
- Step 2: explain the weight of each item in terms of informant knowledge and record timing
- Step 3: state a conclusion with reasoning and name the record that could settle any residue
Worked Scenario: Writing a Research Report a Reviewer Can Follow
Written testing rewards reports structured for a reader: a clear objective, what was searched, what was found, what was not found, and logical next steps. Practice the structure on your own projects before you need it under timed conditions.
Scenario: you are asked to identify a woman's parents using a region's probate and land records, and you exhaust two plausible record groups without finding her. A plausible mistake is reporting only the successful finding elsewhere and compressing the fruitless searches into a single footnote. A reviewer cannot then judge whether your research was reasonably thorough or merely brief.
The better decision is to document negative searches as findings: name each record set, the jurisdiction and years covered, and the search terms used, then connect the negative result to your next hypothesis. Structure the report as objective, scope and limitations, findings in the order you reasoned, and recommendations. This matters because the same structure supports both written responses and oral answers, where you must reconstruct your reasoning aloud for a panel without a document to lean on.
Exercise: The Cite-and-Explain Drill for Oral Examination Practice
Oral examination practice works best as a speaking drill: take one citation, classify it through all three evidence layers, and explain your reasoning aloud in under a minute. Score yourself against a rubric, then repeat with harder records weekly.
The drill: choose three sources for one event in your region — ideally one original, one derivative, and one with a conflicting informant. For each, speak aloud without notes: what the source is, who informed it, what evidence it offers for a stated question, and what your next search would be. Record yourself, then transcribe only the first and last sentences. Vague openings and trailing conclusions are the habits this drill is designed to expose.
Expected observations after a week of daily repetitions: your classifications become automatic enough that you can begin speaking within a few seconds of reading the citation, and your next-search statement stops being generic. Use the rubric below to track a target of consistent fours before you schedule formal testing steps, and repeat the drill with records from your weakest region tab.
- Rubric, 1 point: classification given but no reasoning offered
- 2 points: correct layer names with reasoning that does not address the question asked
- 3 points: correct and reasoned, but the next search is vague or unrelated to the gap
- 4 points: all three layers correct, reasoning tied to the question, and a specific next record named
- Self-check milestone: four consistent 4-point ratings across three different record types — a learning measure, not a prediction of exam results
An Adaptable Preparation Sequence Before You Apply
A realistic sequence runs: build the region notebook, write and defend proof arguments, assemble work samples that demonstrate both, then rehearse orally. Adjust the time spent on each phase to your existing experience rather than following a fixed calendar.
Phase one, spend several weeks expanding the region notebook from actual research, not reading alone, so every entry is tied to a record you handled. Phase two, write short proof arguments — a few hundred words each — for real contradictions in your own lines, applying the source, information, and evidence layers from the scenario above. Aim for arguments across different record types rather than repeating one strength.
Phase three, shape your materials into samples that show region depth and reasoning together, since the testing process evaluates both written submissions and oral performance. Phase four, join a mentored study group if one fits your situation; ICAPGen publishes information on its study groups and posts its current testing dates on its website, so check the issuer's site for the administrative calendar and any requirement details before you finalize your plan.
Readiness Checks Before You Submit for Testing
Treat readiness as a set of observable checks: you can map your region's records without notes, classify any citation quickly, resolve a genuine conflict in writing, and defend your reasoning aloud. If a check fails, return to the matching exercise.
Run the checks cold, with no preparation on the day itself. Pick an unfamiliar county in your region and sketch its record types, starting dates, and custodians from memory, then verify against your notebook and note every gap. Gap-free recall is not the milestone; knowing where your knowledge is thin, and being able to find the answer in minutes, is the demonstration reviewers respond to.
Then complete one full cycle in a single sitting: a citation classified through all three layers, a two-hundred-word proof argument on a genuine conflict, and a spoken one-minute defense of that argument. If the spoken version introduces reasoning absent from the written one, your writing and thinking have drifted apart — practice converting one into the other until they match. Repeat the full cycle monthly as you approach application.
- Check 1: sketch an unfamiliar county's record landscape from memory, then verify
- Check 2: classify three citations through source, information, and evidence layers at speaking speed
- Check 3: write a short proof argument resolving a real, unresolved conflict in your research
- Check 4: defend that same argument aloud in one minute with no notes, and compare it to the written version
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.