AI & Study Science Glossary (40 Terms)

Popularized in LSAT prep: re-solving flagged practice test questions untimed BEFORE looking at the answer key to separate real comprehension limits from time-pressure errors.

Working alongside another person (or an interactive AI sprint check-in partner) to anchor focus and reduce executive-dysfunction wandering, especially helpful for students with ADHD.

A metric used by AI text detectors measuring variation in sentence length and structure; human writing naturally alternates short punchy sentences with longer complex ones, whereas raw AI output has uniform length.

Having an AI model break a multi-step math, chemistry, or logic problem into explicit intermediate reasoning steps so you can verify the logic at each line.

The maximum amount of text, PDFs, or conversation history an AI model can hold in active memory at one time before forgetting earlier details.

The short-term dip in practice performance caused by mixing problem types or adding exam constraints, which paradoxically leads to much higher test-day scores.

A 2×2 prioritization grid sorting tasks by Urgency and Importance so high-impact exam prep doesn't get crowded out by low-value busywork.

A spreadsheet or notebook tracking every practice question you miss by (1) Topic, (2) Why you fell for the trap, and (3) The rule that prevents repeating that mistake.

Analyzing a returned midterm to diagnose whether lost points came from a Content Gap, a Careless Algebra/Reading Slip, or a Time-Management Bottleneck.

When an AI detector like Turnitin or GPTZero mistakenly labels 100% human-written student work as AI-generated—best defeated by keeping Google Docs or Word version history.

The habit of checking every citation, DOI, formula, or historical claim generated by an AI against primary textbook sources or citation-grounded tools like Perplexity and NotebookLM.

Writing a specific "If-Then" study plan ("If it is 7:00 PM after dinner, then I will open my laptop at my desk and do 15 minutes of Calculus") to automate habit execution.

The observation that work expands to fill the time allotted for it; giving yourself a tight 30-minute deadline forces efficiency where an open-ended Sunday afternoon leads to distraction.

In AI text detection, a measure of how unpredictable word choices are to a language model; raw LLM text chooses high-probability next words (low perplexity), while human writing is more varied.

Attempting to answer practice questions on a chapter BEFORE reading it—even when you guess wrong, it primes your brain to spot the key concepts as soon as you read.

Pairing a high-effort behavior (30 minutes of Anki flashcards) immediately before a high-reward habit (watching an episode or grabbing coffee) to reinforce daily study consistency.

Technology used by tools like NotebookLM where the AI searches ONLY your uploaded class PDFs and lecture notes to answer questions, drastically reducing hallucinations.

Uploading your professor's or exam board's official grading rubric alongside your draft so an AI coach can audit whether you hit every required scoring criterion.

System instructions built into Vertech Academy tutors that prevent the AI from dumping finished essays or homework solutions, keeping the student in the driver's seat.

Assigning every task to a specific calendar slot (e.g., "4:00–4:45 PM: Drill 15 AP Chem equilibrium problems") instead of working from a vague, endless to-do list.