CDMRE MCQ common mistakes
CDMRE MCQ · Understand the traps and how to avoid them.
Selecting an answer because it describes a plausible marketing activity, without checking whether it actually answers the specific question asked (e.g., picking a tactic when the stem asks for the metric or the objective).
Under time pressure candidates pattern-match on familiar terms in the options rather than re-reading the stem's operative verb and object, and distractor options are deliberately written as true-but-irrelevant statements.
What to do instead
Before choosing, restate the stem in your own words as a question and eliminate any option that is true but does not resolve that exact question; verify the option answers the verb (identify, calculate, best explains, most likely).
Confusing research designs and their appropriate use-treating exploratory, descriptive, and causal designs as interchangeable, or recommending a focus group when the question requires quantifying incidence or establishing causation.
Qualitative and quantitative methods are learned as a list rather than as tools matched to research objectives, so candidates default to the method they know best.
What to do instead
Build a two-axis decision table (objective: explore vs. describe vs. test causation; data type: qualitative vs. quantitative) and, for each practice stem, first classify the objective, then the method, then check the option matches both.
Misreading sampling and validity concepts-assuming a large sample fixes a biased sampling frame, or treating a convenience sample as generalizable because the n is large.
Sample size and representativeness feel like the same idea, and exam stems often tempt with an impressive sample number attached to a flawed frame.
What to do instead
Apply the rule 'representativeness is about frame and selection method, precision is about size' to every sampling question; reject any option that claims a large biased sample is generalizable.
Applying classic consumer-behavior theory (e.g., hierarchy-of-effects, Maslow, high/low involvement) to a scenario where the described decision process contradicts the model's assumptions, instead of reading the behavior described in the stem.
Candidates memorize model names and definitions but not the boundary conditions under which each model applies, so they force-fit the most recognizable theory.
What to do instead
For each behavioral theory, memorize its preconditions (involvement level, purchase frequency, perceived risk, information search) and test the stem against those preconditions before selecting the theory name.
Treating SEO factors as equally weighted absolutes-choosing 'keyword density target' or 'meta keywords matter' style options, or assuming a technical fix alone guarantees ranking improvement.
Outdated or oversimplified SEO heuristics circulate widely, and questions asking about 'most likely impact' are answered with absolute-sounding rules rather than relative influence.
What to do instead
Frame SEO answers around relevance, authority, and technical accessibility as interacting factors; prefer options describing relative contribution and user intent over fixed numeric thresholds, and flag any option asserting a guaranteed ranking outcome.
Mixing up paid media metrics and their formulas-confusing CTR with conversion rate, CPC with CPA, or ROAS with ROI-and then selecting the option that computes the wrong ratio correctly.
The metrics share inputs (clicks, impressions, spend, revenue), so under time pressure candidates plug numbers into the wrong formula and the arithmetic looks right.
What to do instead
Write the formula for each metric (CTR = clicks/impressions; CPC = spend/clicks; CPA = spend/conversions; ROAS = revenue/spend; ROI = (revenue − cost)/cost) before computing, and sanity-check whether the resulting magnitude is plausible for the metric named.
Treating correlation as causation or ignoring confounders when interpreting an analytics scenario-e.g., crediting a campaign for a lift that coincides with seasonality, a site change, or a tracking change.
Dashboards present attribution and lift as causal, and candidates accept the reported number rather than interrogating what changed concurrently.
What to do instead
For any measurement stem, ask 'what else changed in the same window, and was there a control or holdout?' and prefer options that recommend controlled tests, holdouts, or confounder checks over options that assert the campaign caused the result.
Choosing a vanity or aggregate metric (impressions, page views, average time on site, blended averages) when the question asks for the KPI that best reflects the stated business objective.
Aggregate metrics are the most visible in standard reports, and candidates equate 'reported' with 'relevant' rather than mapping metric to objective.
What to do instead
For KPI questions, first name the business objective and decision the metric should inform, then select the metric tied to that outcome; reject options that cannot change a decision or that average across heterogeneous segments.
Ignoring segment-level differences and reporting a blended average that masks opposite trends in subgroups (Simpson's-paradox style errors), or drawing conclusions from too small or non-representative slices.
Blended dashboards are the default view, and candidates summarize the top-line number without disaggregating or checking subgroup sample sizes.
What to do instead
On any data-interpretation item, mentally disaggregate by the obvious segments (channel, device, new vs. returning, cohort) and check whether the conclusion holds; prefer options that recommend segment-level analysis and flag small-n or confounding.
Overlooking consent, lawful basis, and data-minimization constraints-recommending tracking, list acquisition, or targeting tactics that the scenario's privacy/regulatory context prohibits, or ignoring opt-in status.
Candidates evaluate tactics on effectiveness alone and treat compliance as a separate topic, so legally invalid but high-performing options look attractive.
What to do instead
Run every tactic option through a quick compliance screen (consent obtained? lawful basis? data minimized? opt-out honored? jurisdiction-specific rules?) and eliminate any option that requires data use the scenario has not authorized, even if it would perform well.