Skip to content

AP Human Geography · technique

How to write a DBQ

On this page · 5 sections

Start with a correction that will save you hours of misdirected practice: AP Human Geography has no DBQ. No question on this exam hands you a stack of documents and asks you to construct a thesis-driven argument, so if you have been drilling APUSH-style document essays, redirect that preparation immediately.

Section II contains three free-response questions in 75 minutes, and it is worth 50 percent of your composite score. FRQ 1 arrives with no stimulus attached, and FRQ 2 supplies exactly one: a choropleth map, a population pyramid, a scatterplot, a data table, a landscape photograph, a land-use model diagram, or a short textual excerpt. FRQ 3 supplies two, usually selected so that they are intelligible together. Because a later part may draw on both, examine them as a pair before you begin. Nothing guarantees a part that explicitly combines them. Every point on FRQ 3 remains one of the standard seven, earned by satisfying that particular part's task verb. Those stimulus questions are what students in this course generally mean by the DBQ, and the scoring resembles a genuine DBQ in no respect whatsoever.

Each FRQ contains seven parts, labeled A through G, and every part is worth exactly one point, producing 21 identically weighted points. There is no thesis point, no evidence point, no complexity point, and no credit whatsoever for organization. A reader examines your part C, determines whether those one or two sentences satisfy the requirement, and immediately proceeds. A wrong part C therefore costs you nothing on part D.

The parts escalate in difficulty. Early parts say identify or define, middle parts say describe, and late parts say explain or compare. There is no essay to structure and no place for an introduction, only twenty-one small targets approached individually.

The steps

  1. 1

    Cut 75 minutes into three 24-minute blocks plus a closing sweep

    There is no separate reading period. Three questions, 21 points, and every point weighted identically, so divide the clock evenly and bank the remainder. Give each question a 24-minute block, which consumes 72 minutes and reserves the final 3 minutes for a closing sweep across all three responses. Inside a block, spend 90 seconds on the stimulus, roughly 20 minutes writing, and 2 minutes checking. FRQ 1 carries no stimulus, so that block becomes approximately 22 minutes of writing plus the identical 2-minute check. Establish a hard rule now: when a block reaches 24 minutes, you advance immediately, even in the middle of a part. This matters because the points are completely flat, and a fresh question's part A, which usually asks you to define a single term, is a far cheaper point than the part G currently defeating you. Guard FRQ 3 most aggressively, because it carries two stimuli to interpret and therefore needs its entire block rather than whatever minutes happen to survive. Reserve those 24 minutes before you write a single word.

  2. 2

    Strip the stimulus before you read any part

    Ninety seconds, in this order: title, legend or key, axis labels, units, year, source, and scale. Units determine everything, because a map of population change measured in percentages tells a completely different story from one measured in absolute numbers, and confusing the two contaminates every part beneath it. Then locate the extremes, meaning the highest region, the lowest region, and the direction of the trend, and state that entire pattern in one sentence to yourself before you read part A. For a population pyramid, read the width of the base against the width of the top. For a choropleth, establish what the darkest class represents and where those units cluster. For a scatterplot, read direction and outliers, and for a cartogram, identify which countries swelled and which contracted. For a line graph, locate the crossover point. On FRQ 3, examine both stimuli before part A and decide how they relate: the same place at two different dates, the same variable in two different places, or quantitative data paired with a photograph or a textual excerpt.

  3. 3

    Convert each task verb into a required sentence shape

    The verb determines what earns the point, and readers apply it literally. Identify means name it, nothing more, and define means supply the precise meaning, with no example required. Describe means state features or a pattern, meaning what the thing actually looks like. Causation is not required there, and attaching one does not disqualify your answer; it simply earns nothing while consuming clock you need later. Explain means supply a mechanism, meaning how or why, with a genuine causal link inside the sentence. Compare means addressing both cases and stating the relationship outright, using a word like higher, lower, faster, or unlike, because implied comparison earns nothing.

    Watch for scale. Prompts in this course routinely fix one: at the local scale, at the regional scale, at the national scale, at the global scale. Scale of analysis is its own named skill in the course framework, and an answer pitched at the wrong scale misses the point even when the underlying geography is entirely correct. If the prompt specifies national, your evidence is a country-level pattern, law, or policy, never a single neighborhood, and if it specifies local, you name the neighborhood, the county, or the census tract. If a part asks how a pattern differs across scales, state the pattern at each scale inside the same answer, so the reader can see both.

    Then read the remainder of the instruction, because the restrictions are where points get denied. Note the count, which is usually ONE, note exclusions like other than the example in part C, and note requirements like using the map. Before writing anything, mark the verb, the scale, the count, and the restriction on all seven parts. That preliminary pass takes 40 seconds and prevents the most common zero on this section.

  4. 4

    Write one labeled paragraph per part, in order, and nothing else

    No introduction, no thesis, no conclusion, and no restating the prompt, because none of it is scored and all of it consumes minutes you need for part G. Label every answer with its letter and address the parts in order. If your answers run together, or one lands beneath the wrong label, a reader scores your part E as your part D and you forfeit both points. Two to four sentences is the correct size for almost every part, since additional length does not score higher and a lengthy answer that wanders can contradict itself. Do not hedge by offering two answers when the prompt requests one. Readers generally credit the first answer you give, so a wrong first answer can bury a correct second one, which means you should commit to a single response. Never leave a part unanswered: a plausible sentence costs you 40 seconds and sometimes purchases a point, whereas a blank guarantees a zero.

  5. 5

    Build every explain answer as a because-chain ending in a named place

    An explain point lives in the causal link, not in the vocabulary, so write claim, then mechanism, then consequence, then a specific example. Test it by asking whether you could draw an arrow between two clauses. If the sentence contains because, which causes, or leading to, and the material following it genuinely does follow, you have a mechanism. Then anchor that mechanism geographically, because vague geography kills otherwise correct answers, and people move for better opportunities identifies neither a place nor a process. Compare it with a sentence that actually works: factory wages in Shenzhen far exceed farm income in rural Sichuan, so workers migrate toward the coast, which is why China's population passed 50 percent urban in 2011. The underlying idea is identical, but it names the source region, the destination, the driver, and a number. Maintain a mental inventory of specific cases from every unit, so that you always have a place name available.

  6. 6

    Spend the stimulus

    When a part says using the map, using the graph, or according to the table, you must point at something inside it: a value, a labeled region, a year, or a class on the legend. An answer that is geographically accurate but never touches the stimulus fails that part, because the part is testing whether you can interpret the source. Put the number on the page, whether that number is about 6.8 births per woman or the darkest class covering the Sahel. Even when a part does not require it, a stimulus detail is free specificity and costs you nothing. On FRQ 3, a later part often draws on both stimuli, and when it does, state the connection explicitly rather than implying it. Name both sources inside the sentence, so the reader can see the relationship you constructed instead of guessing at it.

  7. 7

    Run a three-minute sweep

    Before the section closes, verify five things quickly. One: all 21 parts have an answer beneath them. Two: each answer matches its verb, so nothing that should explain merely describes. Three: no answer sits under the wrong letter. Four: no example is reused where the prompt prohibited reuse. Five: any part demanding the stimulus actually cites it, and any part naming a scale is answered at that scale. Repair the cheapest gap first, which is almost always a missing answer or a describe-when-explain error, since both require only a single added clause. Adding the word because plus a genuine consequence to a describe-shaped answer converts a zero into a point in roughly 15 seconds. Do not spend the sweep polishing prose, because readers score content against a rubric and no point whatsoever exists for style.

Worked through

Here is an FRQ 2 in the authentic format, with one stimulus.

STIMULUS: A population pyramid for Japan in 2020. The base is narrow, and the widest bars sit at ages 45 to 54 and 70 to 74. The bars above 80 are long, and the female side extends noticeably further than the male side at the top.

PART A. Identify the stage of the demographic transition model represented by the pyramid. Answer: Stage 5, in which deaths exceed births and the population declines naturally. Why it scores: identify wants a name, and the appended clause is insurance rather than decoration. Use whichever version of the demographic transition model your own course teaches, and if that model terminates at four stages, write late Stage 4 and supply the identical reason. Either way, attach the reason to the label and then stop, because the point is already earned.

PART B. Define dependency ratio. Answer: The number of people outside the working ages, meaning under 15 plus 65 and over, measured against the number aged 15 to 64, usually stated per 100 working-age people. Why it scores: define wants the meaning alone, with no example, no consequence, and no policy attached. Japan's total dependency ratio sat near 70 per 100 in 2020.

PART C. Describe the shape of the pyramid shown. Answer: The base is constricted and each young cohort is smaller than the one above it. The widest bars are in the middle-aged and older cohorts, and the top is heavy, with long bars above 80 and a noticeably longer female side. Why it scores: describe wants appearance and pattern. Writing because fertility collapsed after 1975 supplies a cause, and describe parts award nothing whatsoever for causes. Causal reasoning belongs in the explain parts below, though notice that part D wants the consequences of this age structure rather than the reasons its base narrowed. Save your causes.

PART D. Explain ONE economic challenge facing a country with the age structure shown. Answer: A narrow base means the labor force contracts as large older cohorts retire, so the number of taxpayers falls at precisely the moment pension and health spending rises. Japan, where people 65 and over reached 28.6 percent of the population in the 2020 census, funds programs like its Long-Term Care Insurance system, introduced in 2000, from a smaller working population every year. Why it scores: the point is the causal chain, not the phrase aging population. Falling workers plus rising obligations equals fiscal strain, and a named country carrying a number closes the argument.

PART E. Explain ONE pronatalist policy a government could adopt in response. Answer: Subsidized childcare and paid parental leave raise births by reducing the cost of a child, in money and in forgone career earnings, for a household where both adults work. Japan's Angel Plan of 1994 expanded daycare capacity for exactly that reason. Sweden pulls the identical lever considerably harder: 480 days of parental leave per child, 390 of them paid at roughly 80 percent of previous income, with 90 days reserved for each parent so that fathers actually take them. Why it scores: naming a policy by itself is identify-level work and earns nothing here. The mechanism sentence, which explains how the policy alters a household's decision, is what converts a named policy into an explain point.

PART F. Explain how sustained in-migration would change the pyramid. Answer: International migrants are heavily concentrated in ages 20 to 40, so sustained inflows widen the working-age bars first and lower the dependency ratio without any change in births. Their subsequent children then widen the base. Japan had roughly 2.9 million foreign residents in 2020, about 2.3 percent of its population, too small a share to reshape the pyramid. Why it scores: it uses the age selectivity of migration, which is the mechanism this part is fishing for, and it then bounds the claim with realistic scale.

PART G. Explain ONE reason a country in Stage 2 has a pyramid shaped opposite to the one shown. Answer: In Niger, where total fertility was roughly 6.8 births per woman in 2020, imported vaccination, antibiotics, and improved sanitation reduced the death rate, especially infant mortality, decades before birth rates responded. Family-size norms tied to farm labor and old-age security change slowly, so births remained high while deaths fell. Each cohort is therefore larger than the one above it, producing a wide base and concave sides. Why it scores: the point lives in the lag between a falling crude death rate and a still-elevated crude birth rate. Naming Niger, and dating its fertility figure, blocks any accusation of vagueness.

Total: seven parts, seven points, roughly 300 words written, and not one sentence of thesis or conclusion anywhere in it.

Now apply the identical method to different content, in the FRQ 3 shape with two stimuli.

STIMULUS 1: A bid-rent curve. Land rent per acre peaks at the central business district, drops steeply across the first few kilometers, then runs nearly flat through the outer suburbs. STIMULUS 2: A photograph taken at a suburban highway interchange. Single-story big-box stores and a distribution warehouse sit low and wide, each surrounded by surface parking.

PART. Using both stimuli, explain why the land uses shown in the photograph locate where they do. Answer: The curve shows rent per acre peaking downtown, because accessibility to the entire metropolitan area is greatest at that point. Warehouses and big-box retail require enormous floor space and free parking per dollar of sales, so they cannot outbid offices for downtown land. They settle where the curve has flattened and land is cheap by the acre, and the interchange still delivers the truck and customer access their business model requires. Why it scores: it names a feature of each stimulus, the curve's flattening and the photograph's low-rise buildings ringed by expansive parking, and it then links them with a mechanism, competitive bidding for accessible land. Notice what the answer refuses to do. It does not summarize both stimuli side by side and leave the reader to draw the arrow between them.

Unit 6 content, Unit 2 method. Read the stimulus, obey the verb and the scale, construct the because-chain, and name the specifics, because none of that procedure changes with the unit.

Common mistakes

  • Writing it as an essay. An introduction, a thesis, and a conclusion earn zero points and consume the minutes you need for parts F and G, because there is no argument to construct, only seven separate questions to answer.

  • Downgrading the verb. Describing when the prompt says explain is the single most common way to forfeit a point you actually knew. If your answer contains no because, no which causes, and no leading to, it is a description, and it will not score on an explain part.

  • Ignoring a scale the prompt already fixed. If the part specifies at the national scale and you answer with a single neighborhood, or it specifies at the local scale and you answer with an entire continent, the geography can be perfectly correct and still miss the point.

  • Restating the prompt, or naming a term without its mechanism. Answering explain a cause of suburbanization with suburbanization happened because people moved to suburbs is a restatement, and naming a policy, a model, or a process is identify-level work that does not satisfy explain.

  • Using place-free examples. Some countries have high birth rates and people migrate for jobs are both unscoreable, so name Niger, Shenzhen, Detroit, the Sahel, or the Rust Belt instead, and attach a dated number whenever you have one.

  • Never touching the stimulus. On a part that says using the map or according to the graph, an answer that never cites a value, region, year, or legend class fails, even when it is geographically correct. On FRQ 3, when a part does draw on both stimuli, implying the connection instead of stating it leaves that point on the table.

  • Hedging, or running out of clock. Offering two answers where the prompt requests one is dangerous, because readers generally credit the first answer, so a wrong first answer can cancel a correct second one. Leaving the late parts of FRQ 3 blank, because that question received the leftover minutes, is the easiest way to surrender four or five points you could have written.

Checklist

  • Can you state, without looking, that Section II is three FRQs in 75 minutes, that parts A through G are worth one point each, and that FRQ 3 carries two stimuli?
  • Can you read a population pyramid, a choropleth map, a scatterplot, and a cartogram, and then say what each one claims in a single sentence within 90 seconds?
  • Can you say what identify, define, describe, explain, and compare each require, and can you name what makes an answer fail for each of those verbs?
  • When a prompt fixes a scale, whether local, regional, national, or global, can you pitch your evidence at that scale instead of defaulting to whichever example you happen to know best?
  • Can you spot the count, the exclusion, and the using-the-stimulus requirement inside a prompt before you start writing?
  • Can you turn any explain part into a because-chain that ends with a named country, city, or region and, wherever possible, a number?
  • Can you make an explicit comparison, using a word like higher, faster, or unlike, instead of leaving the reader to infer the relationship?
  • On a two-stimulus question, can you state the connection between them in one sentence that names both, rather than describing each of them in turn?
  • Can you write a complete seven-part FRQ in 24 minutes with no introduction, no conclusion, and no part left unanswered?
  • Can you run a three-minute sweep that catches a missing answer, a describe-when-explain error, and an answer sitting under the wrong letter?

More AP Human Geography technique