You are given a claim, a hypothesis, or an argument, and you choose the evidence that best supports or undermines it. In textual evidence questions the choices are statements or quotations. In quantitative evidence questions the passage includes a table or graph and the choices describe data from it, so you have to check both that the description matches the data and that the data actually bears on the claim. Some questions ask you to complete a sentence so the example logically illustrates the point.
Command of evidence usually accounts for roughly 3 to 5 of the 54 reading and writing questions, inside the roughly 26 percent Information and Ideas share.
Where students lose points
Choosing a choice that accurately describes the graph but says nothing about the specific claim being tested.
Supporting a claim about one group with evidence drawn from a different group in the same table.
Picking evidence that restates the claim instead of giving an independent reason to believe it.
How to attack these questions
1Write the claim in your own words and mark exactly which variable or group it is about.
2Check each choice twice: first that it is factually consistent with the passage or figure, then that it is relevant to the claim.
3If the question asks what would weaken the claim, look for evidence pointing in the opposite direction, not merely unrelated evidence.
Example questions
Written for this skill, with the answer explained. 347 more are in the app.
Try one
Command of evidence
Real question
Dark dust settles on the surface of mountain glaciers every summer. The layer mixes rock flour, soot, and algae that grow in the meltwater. Glaciologist Sanna Väisälä proposed that this dust speeds melting because dark surfaces absorb more sunlight than clean ice does. To test the idea, her team swept a set of marked squares on one glacier clean each morning and left neighboring squares untouched, reading stakes drilled into every square to measure the season's ice loss.
Which result, if true, would most strongly support Väisälä's proposal?
The answer is C
Väisälä's proposal is that the dark layer speeds melting, and her design sets swept squares beside untouched ones, so the decisive result is the swept squares losing less ice over the season. One finding compares years instead of squares: The glacier as a whole lost more ice in the study year than in the previous year. A warmer summer would produce the same number, so nothing separates dust from weather. Another concerns where the dust came from: The dust on the untouched squares came mostly from farmland to the south. The source of the dust does not bear on whether a dark surface absorbs more sunlight. The last reports no difference at the coldest hour: Swept and untouched squares were the same temperature at dawn each day. Equal temperatures before sunrise is what the proposal expects and shows no absorption effect either way.
Medium
Using account books from a manor in fourteenth-century England, historian Edward Oduya compared the size of each year's wheat harvest with the price of wheat at a nearby market over five years. Accounts of this kind were drawn up each autumn by the reeve who ran the estate for its lord, and they record the grain threshed from each field beside the acreage sown, letting a yield per acre be reconstructed centuries later. Prices at the time depended heavily on local supply, since grain was costly to move over long distances. Oduya concluded that poor harvests drove wheat prices up: ______
Which choice most effectively uses data from the table to complete the statement?
A) the smallest harvest, in 1334, coincided with the highest price, 13 pence per bushel. ✓
B) the largest harvest coincided with the highest price.
C) the price of wheat was lowest in 1334.
D) the harvest was larger in 1333 than in any other year.
Why: Oduya's conclusion links harvest size to price, so the completion has to pair the two, which is what the smallest harvest, in 1334, coincided with the highest price, 13 pence per bushel. does by putting the smallest harvest, in 1334, together with the highest price of 13 pence per bushel. Answering the largest harvest coincided with the highest price. pairs them the other way, with the largest harvest and the highest price, which would argue against his conclusion. Answering the price of wheat was lowest in 1334. misreports the table, where 1334 carried the highest price rather than the lowest. Answering the harvest was larger in 1333 than in any other year. names the largest harvest but gives no price, so it supplies only one side of the relationship the conclusion rests on.
Medium
In a well-known study, people offered a small immediate payment chose it over a larger payment a month away. A researcher concluded that people undervalue future rewards as such. Behavioral scientist Noor Farhadi has repeated the study several times without obtaining the original pattern, and she notes that the original participants were told the later payment would arrive by mail, while nothing was said about how the immediate one would be handed over. Farhadi's repetitions used the same sums and the same one-month delay as the original.
Which finding, if true, would most strongly weaken the researcher's conclusion?
A) Participants who chose the immediate payment were younger on average than those who did not choose it.
B) Participants offered a two-month delay chose the immediate payment even more often than before.
C) Participants chose the later payment when told it would be handed over in person on the same terms. ✓
D) Farhadi's own participants were paid a fixed fee for taking part regardless of their choice.
Why: The conclusion is that delay itself is what participants avoided. Choice C removes the mail and keeps the delay, and the preference reverses, so what they were avoiding was the delivery method. Choice B is the trap for a reader who reads any stronger effect as trouble for a claim, but a bigger gap producing more impatience is what undervaluing the future predicts. Participants who chose the immediate payment were younger on average than those who did not choose it. is who made the choice, not why the delay was avoided. Participants offered a two-month delay chose the immediate payment even more often than before. is a steeper version of the original pattern, which supports the conclusion. Farhadi's own participants were paid a fixed fee for taking part regardless of their choice. is a detail of her procedure, not a finding about the future's value.
Harder
Deep peat bogs preserve pollen in datable layers. Paleobotanist Ngozi Eze proposed that a sharp drop in tree pollen in one Irish bog around 1300 CE records clearance by farmers rather than dieback from a colder climate. The bog lies in a lowland basin, and Eze counted grains at every centimeter of a five-meter core, several hundred grains to a level. She fixed the chronology with radiocarbon dates and with two ash layers laid down by Icelandic eruptions recorded in 1104 and 1362.
Which finding, if true, would most strongly support Eze's proposal?
A) Tree pollen recovers in layers dated about two centuries later.
B) The same layers that lose tree pollen gain cereal pollen and charcoal, and tree species that tolerate cold fall as sharply as those that do not. ✓
C) The bog has accumulated peat continuously for eight thousand years.
D) Temperature records reconstructed from other sites show cooling across Europe in the fourteenth century.
Why: Eze proposes clearance by farmers rather than dieback from cold, so the supporting finding must add a human signal and remove the climate one, which cereal pollen and charcoal in the same layers, plus cold-tolerant trees falling as sharply as the rest, does. One finding is about later layers: Tree pollen recovers in layers dated about two centuries later. Recovery two centuries on follows abandoned fields or a warming climate equally. Another is about the site: The bog has accumulated peat continuously for eight thousand years. Continuous peat accumulation supports the record's reliability, not the cause of the drop. The last supports the explanation she rejects: Temperature records reconstructed from other sites show cooling across Europe in the fourteenth century. Cooling across fourteenth-century Europe is evidence for the dieback account.
Questions students ask
How do I answer quantitative evidence questions with a graph?
Read the axis labels and units before you read the choices. Then check each choice against the figure for accuracy, and separately check whether it addresses the claim in the passage. A choice can be a correct reading of the data and still be the wrong answer.
What makes evidence strong for a claim?
Strong evidence is about the same subject, the same group, and the same variable as the claim, and it points in the direction the claim predicts. Evidence that is true but about a different group or a different measure does not support the claim.
What does it mean to weaken a hypothesis?
It means providing a result the hypothesis did not predict, or one that fits a competing explanation better. The right answer usually shows the expected pattern failing to appear, not just a missing piece of information.
Drill this skill until it is automatic
Questions on command of evidence alone, with explanations on every one, and the app tracks which ones you keep missing.