Practice Test Geeks(CPI) Certified Principal Investigator Practice Test
← All CPI Flashcard Decks

Data Management & Analysis Flashcards

7 cards from real CPI practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.

Read the first 7 Data Management & Analysis flashcards as text

  1. Which data collection strategy best minimizes recall bias in a study asking participants about dietary habits over the past year?

    Answer: Prospective food diary maintained daily

    Prospective daily diaries capture information as events occur, eliminating the need to recall past behaviors.

  2. A study database flags a creatinine value of 45 mg/dL as a data query. The site confirms this was a transcription error and should be 4.5 mg/dL. What must accompany the data correction?

    Answer: An audit trail entry documenting who changed the value, when, and why

    GCP requires that all data corrections are traceable via audit trail showing the original value, new value, responsible person, date, and reason.

  3. What is the key difference between a cross-sectional study and a longitudinal study in data collection?

    Answer: Cross-sectional studies collect data at one point in time while longitudinal studies follow participants over time

    Cross-sectional studies capture a snapshot at one moment, while longitudinal studies track the same subjects across multiple time points.

  4. In a randomized controlled trial, what does stratified randomization help ensure?

    Answer: Balance of important prognostic factors across treatment arms

    Stratified randomization ensures that key baseline variables (e.g., age, disease severity) are evenly distributed between treatment groups.

  5. Which HIPAA identifier, if removed, is NOT sufficient to de-identify health information under the Safe Harbor method?

    Answer: Geographic data smaller than a state

    Under HIPAA Safe Harbor, geographic subdivisions smaller than a state (except first 3 digits of ZIP if population >20,000) must be removed.

  6. A receiver operating characteristic (ROC) curve is used to evaluate a diagnostic test. An AUC of 0.50 indicates what?

    Answer: Test performance no better than random chance

    An AUC of 0.50 means the test cannot discriminate between positive and negative cases better than a coin flip.

  7. What is the purpose of a data dictionary in clinical research?

    Answer: To define variables, coding schemes, units, and permissible values in a dataset

    A data dictionary is a reference document that standardizes variable definitions and coding to ensure consistent data interpretation.