Cross-Text Connections: SAT Practice Questions & Study Guide
Comparing two short passages on the same topic to analyze how the authors' perspectives, arguments, or methods relate to each other.
Understanding Cross-Text Connections on the SAT
Cross-Text Connections questions present two short passages—Passage 1 and Passage 2—written by different authors on the same or related topic. The question then asks you to analyze the relationship between the two texts: how one author might respond to the other's argument, whether the authors agree or disagree on a specific point, or what claim is supported by one text and challenged by the other. These questions are analytically demanding because they require holding two distinct authorial stances in mind simultaneously.
The most common question types in this category are: 'How would the author of Passage 2 most likely respond to the claim made in Passage 1?', 'Both authors would agree that…', or 'Unlike the author of Passage 1, the author of Passage 2…'. Each of these requires you to first characterize each author's position with precision. Vague characterizations—'Author 1 thinks technology is complicated'—are rarely precise enough. You need targeted summaries: 'Author 1 argues that algorithmic recommendation systems systematically amplify misinformation and recommends regulatory intervention.'
A critical skill is distinguishing between claims an author explicitly makes, claims an author implies but does not state, and claims an author neither makes nor implies. SAT Cross-Text questions frequently offer distractors that attribute to an author a position that is plausible but not actually in the text. The correct answer requires inference that is tightly tethered to the passage—not logical extrapolation, but supported inference based on what the author actually wrote.
A practical approach: after reading both passages, annotate each author's main claim and attitude in the margin (or mentally). Note areas of overlap and areas of divergence. When you read the question, identify which part of each passage is most relevant to the specific claim in the question stem, and return to that part before selecting an answer. Students who try to answer Cross-Text questions from memory rather than re-reading the relevant sections are far more likely to make errors.
Key Rules & Formulas
Memorize these rules — they come up directly in SAT questions.
Summarize each author's main claim and attitude before looking at the question.
Passage 1: Author argues social media increases political polarization. Passage 2: Author argues social media's effects on polarization are overstated and context-dependent.
Identify the specific claim in the question stem and find where each passage addresses it.
If the question asks how Author 2 would respond to Author 1's claim about filter bubbles, locate Author 2's discussion of filter bubbles—do not rely on Author 2's overall thesis.
An author can only 'agree' or 'disagree' with a claim if their passage directly addresses it—do not assume agreement from silence.
If Passage 2 never discusses algorithmic amplification, the author of Passage 2 cannot be said to agree or disagree with Author 1's claim about it.
Distinguish between the author's explicit claims and the reader's inferences—Cross-Text answers must be grounded in what the text says or clearly implies.
An author who describes the positive effects of urban green space is not necessarily advocating for a specific policy—do not attribute a policy position the author never stated.
When two authors agree on a point, both passages must contain evidence for that agreement—one passage alone cannot establish agreement.
Both Authors agree that X is only correct if you can point to a sentence in each passage that supports X.
Cross-Text Connections Practice Questions
Select an answer and click Check Answer to reveal the full explanation. Questions go from easiest to hardest.
The following texts are adapted from two essays on remote work. Text 1 Remote work, once a rare privilege granted to a small professional class, has become a defining feature of contemporary employment in many industries. Advocates emphasize that remote workers report higher job satisfaction, reduced commuting stress, and, in numerous studies, equal or greater productivity than office-based counterparts. For knowledge workers especially, the flexibility of remote work may represent a durable improvement in working conditions. Text 2 While the productivity data on remote work is often cited optimistically, such data must be interpreted carefully. Most productivity studies measure output rather than collaboration quality, and remote work has been associated with measurable declines in spontaneous knowledge-sharing, mentorship, and the informal social bonds that sustain organizational culture over time. Productivity metrics alone offer an incomplete picture of remote work's full organizational impact. Based on the texts, how would the author of Text 2 most likely respond to the claim in Text 1 that remote workers show "equal or greater productivity"?
The following texts are adapted from two articles on antibiotic resistance. Text 1 The overuse of antibiotics in agriculture—particularly the practice of administering subtherapeutic doses to promote livestock growth—has contributed significantly to the global antibiotic resistance crisis. Resistant bacteria from farm environments can transfer resistance genes to human pathogens through various pathways, including water runoff and direct human contact with animals. Regulatory limits on agricultural antibiotic use are urgently needed. Text 2 While concerns about agricultural antibiotic use are legitimate, attributing the antibiotic resistance crisis primarily to agriculture oversimplifies a complex problem. Hospital-acquired infections involving resistant bacteria are driven substantially by prescription practices in clinical settings, where antibiotics are frequently overprescribed for viral conditions against which they have no effect. Any serious policy response to antibiotic resistance must address clinical overprescription alongside agricultural use. Based on the texts, both authors would most likely agree that
The following texts are adapted from two essays on rewilding. Text 1 Rewilding—the large-scale restoration of natural processes, including the reintroduction of apex predators—has gained traction as a conservation strategy in Europe and North America. Proponents argue that reintroducing wolves and other predators can restore ecological balance, reduce overgrazing by deer and elk populations, and even alter the physical landscape through cascading effects known as trophic cascades. The reintroduction of wolves to Yellowstone National Park is frequently cited as a successful demonstration of these principles. Text 2 The enthusiasm for rewilding, and for trophic cascade theory in particular, has outpaced the supporting evidence. The Yellowstone wolf reintroduction is a compelling narrative, but the scientific literature presents a more complicated picture. Studies attempting to replicate trophic cascade effects in other ecosystems have produced mixed results, and ecologists have cautioned that the Yellowstone case may reflect conditions unique to that ecosystem rather than a generalizable model. Rewilding advocates would serve their cause better by acknowledging this uncertainty. Unlike the author of Text 1, the author of Text 2
The following texts are adapted from two essays on artificial intelligence in medicine. Text 1 Machine learning algorithms trained on medical imaging data have demonstrated diagnostic accuracy that rivals, and in some cases exceeds, that of experienced radiologists. These systems can process thousands of images rapidly and consistently, free from the fatigue and attentional variability that affect human performance. Integrating AI into radiology workflows has the potential to increase diagnostic throughput and reduce the incidence of missed diagnoses. Text 2 The clinical deployment of AI diagnostic tools faces a challenge that benchmark accuracy statistics obscure: performance degradation across populations. Algorithms trained predominantly on imaging data from large academic medical centers may perform significantly worse when deployed in community hospitals serving demographically different patient populations. Before broad clinical deployment, AI diagnostic tools require rigorous external validation across diverse patient populations and imaging environments. The author of Text 2 would most likely characterize the evidence presented in Text 1 as
The following texts are adapted from two scholarly essays on the history of science. Text 1 Thomas Kuhn's The Structure of Scientific Revolutions introduced the concept of the paradigm shift, arguing that science does not progress through steady, cumulative accumulation of knowledge but through periodic discontinuous revolutions in which one theoretical framework is replaced by another. Kuhn's account was transformative: it challenged the dominant view of science as a purely rational enterprise and introduced a social and historical dimension to scientific change. Whatever its limitations, Kuhn's framework remains indispensable for thinking about how science actually develops. Text 2 The influence of Kuhn's paradigm-shift model, while undeniable, has been accompanied by a distortion of his argument in popular and interdisciplinary usage. Kuhn was writing specifically about the physical sciences and about the rare, dramatic episodes that mark genuine theoretical transitions. When the term "paradigm shift" is applied to incremental methodological changes, shifting research trends, or even marketing campaigns, it loses its analytical specificity entirely. The misappropriation of Kuhn's vocabulary has arguably done as much damage to serious philosophy of science as his original work did good. Based on the texts, the author of Text 2 would most likely agree with which claim from Text 1?
The following texts are adapted from two essays on the ethics of gene editing. Text 1 The development of CRISPR-Cas9 gene editing technology raises profound ethical questions, particularly regarding germline editing—modifications to embryos that would be heritable by future generations. Proponents argue that germline editing could eliminate heritable diseases such as Huntington's and cystic fibrosis, reducing suffering on a generational scale. Given the potential benefits, research into therapeutic germline editing deserves continued support, provided it proceeds within robust regulatory frameworks. Text 2 Arguments for therapeutic germline editing consistently underestimate the distinction between treating a living patient and altering a future person who cannot consent to the modification. The analogy to treating existing diseases is fundamentally misleading: a therapy removes a condition a patient wishes to be rid of, but germline editing preemptively determines a characteristic of a person not yet born without that person's input. The inability to obtain meaningful consent from those most affected is a categorical barrier, not merely a procedural obstacle. The author of Text 2 would most likely respond to Text 1's claim that germline editing "could reduce suffering on a generational scale" by arguing that
The following texts are adapted from two essays on digital privacy. Text 1 The collection of personal data by technology companies has reached a scale that earlier generations of privacy theorists could not have anticipated. Every click, search, and purchase is logged, aggregated, and sold to advertisers, insurers, and employers. Individuals who wish to participate in modern digital life have no practical means of opting out of this surveillance economy. Comprehensive data protection legislation, modeled on the European Union's General Data Protection Regulation, is the only meaningful remedy available. Text 2 Legislative approaches to data privacy, while necessary, are insufficient on their own. The GDPR, despite its stringent requirements, has not significantly altered the business models of major technology companies; compliance has largely taken the form of lengthy consent notices that users dismiss without reading. Effective data privacy protection requires not only legislation but also technical infrastructure—privacy-preserving computing techniques, open-source alternatives to surveillance-based platforms, and digital literacy education that empowers individuals to make meaningful choices. Based on the texts, both authors would most likely agree that
The following texts are adapted from two essays on the epistemology of expert testimony. Text 1 Democratic societies face a growing problem of expert-public mistrust. When scientific consensus on issues such as vaccine safety or climate change is dismissed by significant portions of the public, the consequences can be serious and widespread. Epistemic deference—the disposition to trust experts in their fields of demonstrated competence—is not blind faith but a rational response to the division of cognitive labor in modern knowledge-producing institutions. Dismissing expert consensus without commensurate expertise in the relevant domain is not a mark of critical thinking; it is epistemic arrogance. Text 2 The case for epistemic deference is not as simple as its advocates suggest. Expertise is domain-specific: a climate scientist is an authority on atmospheric dynamics, not necessarily on climate policy, yet public discourse frequently conflates these. Furthermore, the history of science includes well-documented cases in which emerging expert consensus was wrong and laypersons or dissenting scientists were eventually vindicated. Calibrated deference—trusting experts on technical matters within their specific domain while maintaining critical scrutiny of their policy prescriptions—is more epistemically defensible than wholesale deference. Based on the texts, the author of Text 1 and the author of Text 2 most directly disagree about
Want more Cross-Text Connections practice?
Access 1,000+ additional questions filtered by difficulty and score band in the full 1600.lol question bank — free, no signup needed.
Common Mistakes to Avoid
These are the most frequent errors students make on Cross-Text Connections questions. Knowing them in advance prevents costly point losses.
- !Attributing to an author a position that seems logically consistent with their argument but is not actually stated in their passage.
- !Confusing which author made which claim—students frequently flip the two authors' positions under time pressure.
- !Selecting an answer that describes what both authors discuss (their shared topic) rather than what they agree or disagree about (their positions).
- !Using information from one passage to answer a question specifically about the other passage's author, ignoring what the second passage actually says.
- !Choosing an answer that is too strong—for instance, selecting 'Author 2 would strongly reject' when the evidence only supports 'Author 2 would question' or 'would find insufficient.'
SAT Strategy Tips: Cross-Text Connections
Write a one-sentence summary of each author's main point before reading the questions—this investment of 20 seconds prevents the confusion that causes most errors on Cross-Text questions.
When a question asks 'how would Author 2 respond to Author 1's claim,' identify the specific sentence in Author 1 that contains the referenced claim, then find the most relevant part of Author 2's passage to evaluate a response.
Use the process of elimination aggressively: one or two answer choices will usually misattribute a position, and one will use language that is far stronger than what the passage supports—eliminate these first.
Practice identifying agreement versus disagreement quickly: look for evaluative language (criticizes, endorses, dismisses, supports) and qualifiers (however, nevertheless, in contrast, similarly) in both passages as signals of each author's stance.
Other Craft and Structure Subtopics
Master Cross-Text Connections on the SAT
These 8 questions are just the start. Unlock the full 1600.lol question bank for 12,000+ official-style SAT questions with the Desmos calculator, instant feedback, and progress tracking.
Join 50,000+ students preparing for the 2025–2026 Digital SAT on 1600.lol