What Analogies Measure
Analogies are one of the most reliable indicators of abstract reasoning ability and have been central to high-stakes assessment for over a century. They measure three core cognitive capacities that predict academic success and professional performance: relational reasoning (the ability to identify relationships between concepts), conceptual mapping (understanding how those relationships transfer to new domains), and vocabulary depth (recognizing that mastery of relationships requires rich semantic knowledge).
The research evidence is consistent across multiple testing bodies. SAT analogies appeared in the Scholastic Aptitude Test from its inception in 1926 through the 2004 redesign, when the test committee determined that analogies were redundant with sentence completion and reading comprehension for measuring verbal reasoning. However, the Miller Analogies Test (MAT), which measures reasoning for graduate-level admission, remains heavily analogy-based, MAT reasoning subtests contain 120 analogies and are still administered thousands of times annually to psychology, law, and business school candidates. The GRE (Graduate Record Exam) historically featured analogies until 2002, when they were replaced with critical reading. MCAT analogies appear in the verbal reasoning section and serve as a key predictor of medical school performance. Across all these instruments, analogies correlate strongly with measures of general intelligence (g) and with later academic and professional achievement, independent of domain-specific knowledge.
The Standard Analogy Format
Analogies follow a consistent structural format that appears across all major testing bodies: A is to B as C is to what? This is typically written as A:B :: C:?, and is read aloud as "A is to B as C is to blank." Understanding how to parse this structure is the foundation of analogy problem-solving.
The key to reading the format correctly is recognizing that the phrase "is to" serves as a relational shorthand. It encodes the direction of the relationship from the first term to the second. If the analogy is "Surgeon:Scalpel :: Carpenter:?", the "is to" between Surgeon and Scalpel means "uses as a primary tool." That same directional relationship must hold between Carpenter and the correct answer. The four terms arrange into a strict logical pattern: two pairs, where the relationship within the first pair must parallel the relationship within the second pair. This structure is so fundamental that even ambiguous relationships (ones that could be expressed multiple ways) follow the same A→B, C→? vector. Correctly identifying that vector is the entire problem.
Common Relation Types
Analogies rely on a finite set of relationships that appear repeatedly across test versions. Recognizing these patterns accelerates problem-solving and helps you distinguish between answer choices that may seem superficially plausible but encode different relationships.
- Synonym or Near-Synonym: Words with very similar meanings. Example: Courageous:Brave :: Frightened:Scared. The relationship is semantic equivalence.
- Antonym or Opposite: Words with opposite meanings. Example: Expand:Contract :: Hot:Cold. The relationship is semantic opposition along a spectrum.
- Part-to-Whole: A component relates to the larger entity containing it. Example: Petal:Flower :: Verse:Poem. The first term is a constituent part of the second.
- Cause and Effect: One concept produces or directly causes the other. Example: Poison:Death :: Education:Literacy. The first creates or enables the second.
- Tool or Function (Object to Use): A person or agent uses the first term to accomplish the second, or a tool performs the second function. Example: Artist:Brush :: Surgeon:Scalpel. The first term uses the second to practice their work.
- Category and Instance: The first term is a broader class; the second is a specific member of that class. Example: Color:Red :: Emotion:Anger. The second is a concrete example of the first's category.
- Action and Object: An action is performed on or with a specific object. Example: Read:Book :: Eat:Food. The first term is the action; the second is the typical object receiving the action.
- Sequence or Progression: Terms follow a logical or temporal order. Example: Spring:Summer :: Infancy:Childhood. The first precedes the second in a standard sequence.
- Degree or Intensity: The terms represent the same concept at different intensities. Example: Annoy:Infuriate :: Dislike:Detest. The second is a stronger form of the first.
These nine relation types account for the vast majority of analogies on standardized tests. Becoming fluent in recognizing them allows you to solve problems faster and with greater certainty.
How to Approach Analogy Questions
The most effective strategy for solving analogies follows a three-step process that separates high performers from those who guess. This method is taught by test preparation organizations across MAT, GRE, and MCAT programs.
Step 1: Identify the Precise Relationship Between A and B. Before looking at any answer choices, form a clear statement of how A relates to B. Don't settle for vague descriptions like "they go together" or "they're related." Be specific: Is one a synonym for the other? Is A a part of B? Does A cause B? Does A use B? The more precisely you articulate the relationship, the easier it is to spot the correct answer and avoid traps. Write your relationship statement down or say it aloud if possible. Example: If the analogy is "Prologue:Book," your relationship statement should be "A prologue is the opening section that precedes the main content of B."
Step 2: Find the Parallel Relationship in the Answer Choices. Now move to the C term and check each answer option for one that creates the same relationship with C that B has with A. Read each option as a complete analogy: "C is to [option] in the same way that A is to B." Does the relationship hold? Is the directionality correct? Many test-takers fail this step by finding an answer that is related to C in some way, without checking that the relationship type matches the A:B relationship. Use your relationship statement from Step 1 as a filter. If the A:B relationship is "tool used by," then the C:? relationship must also be "tool used by," not just "related to C in some other way."
Step 3: Check for Trap Answers With Similar But Different Relations. Test makers deliberately include wrong answers that encode relationships similar enough to confuse quick thinkers. An answer might be related to the test terms in a different way, perhaps it's a category rather than a part, or a cause rather than a tool. For example, if the analogy is "Cardiologist:Heart," the relationship is "medical specialist who studies/treats." An answer like "Anatomy" (something you study the heart in) is related but encodes a different relationship (subject of study, not practitioner). Eliminate answers that are related to C but through a different relationship type than the one you identified in Step 1. This final check often separates a good guess from a confident, correct answer.
Common Analogy Traps
Test makers use three main strategies to create plausible wrong answers that catch careless test-takers. Recognizing these traps will help you avoid them.
- The Same-Domain Trap: The answer is from the same semantic field or category as the correct answer but encodes a different relationship. Example: For "Surgeon:Scalpel :: Baker:?", the answer "Oven" might be tempting because both Oven and Rolling Pin (the correct answer) are tools in a bakery or kitchen. However, Oven is where you do the primary action (bake), whereas Rolling Pin is the tool you use while baking. The relationship is different. Both answers are in the same domain, but only one encodes the correct relationship.
- The Reverse-Relation Trap: The answer encodes the relationship backwards, B:A instead of A:B. Example: For "Petals:Flower," a trap answer might be "Forest" (because Flower relates to Forest as a specific instance to a broader category, the opposite direction). The relationship must flow in the same direction as the original analogy. This trap is especially effective because the wrong answer is still related to the correct answer; test-takers often fail to notice the directional mismatch.
- The Too-Broad or Too-Narrow Trap: The answer is related to C but at the wrong level of specificity. Example: For "Ulna:Bone :: Carrot:?", a trap answer might be "Vegetable" (too broad, carrot is a vegetable, but the relationship is more specific) when the correct answer requires the same specificity as "Bone" (like "Root Vegetable"). Similarly, too-narrow answers trap test-takers who overgeneralize. The specificity of the A:B relationship (how general or specific the category is) must match in C:?.
These traps succeed because they exploit the natural tendency to find related concepts and assume they fit. By keeping your relationship statement from Step 1 in mind and checking both the relationship type and the directionality, you avoid these patterns.
Training Analogy Reasoning
Analogy reasoning is not a fixed ability, it improves substantially with deliberate practice. Research on expertise in reasoning tasks shows that improvement follows a dose-response curve: more practice correlates with better performance, with diminishing returns after 100-150 problems of focused training.
The most effective training approach combines vocabulary expansion with relational thinking. Analogies require both: you can't recognize relationships between concepts you don't understand, and memorizing vocabulary without learning to see relationships leaves you vulnerable on novel test items. Start by studying the Miller Analogies Test (administered by Houghton Mifflin Harcourt), which provides a dedicated practice bank of 120 analogies per full test. Work through MAT practice tests under timed conditions, aim for 1.5 minutes per problem initially, then reduce to 1 minute as you improve. The MAT is more challenging than historical SAT or GRE analogies and provides better preparation for abstract reasoning.
Supplement timed practice with daily analogy puzzles, spend 5-10 minutes each morning solving analogy problems without time pressure, focusing on the accuracy of your relationship identification. Many sites like Analogies Online or the Miller Analogies Test Study Guides offer free banks of problems. As you solve, keep a log of the relationship types you encounter; tracking which relationships give you the most trouble helps you focus your study. The final and most transferable training technique is deliberate relation-categorization: take any set of 10-20 random words and generate as many accurate analogies as you can between pairs of words, ensuring each encodes a different relationship type. This trains your brain to recognize relationships actively rather than just matching pre-formed analogies, and that skill transfers directly to novel test items.
To measure your progress, track your accuracy and speed on practice tests over 4-6 weeks. Most test-takers improve 15-25% on timing and 10-15% on accuracy with consistent daily practice. The gains are largest in the first 2 weeks (establishing the three-step process) and continue steadily through week 6, after which diminishing returns set in unless you switch to more advanced problem sets.
Explore more about structured verbal reasoning assessment strategies at our verbal reasoning tests overview.