Business Analytics Quiz: Types Of Analytics

10 questions · exam conditions

0:00

Types Of AnalyticsQuestion 1 of 10

A retailer's weekly dashboard first reports that revenue fell by $8\%$ . An analyst then decomposes the change into store traffic, conversion rate, and average transaction value. The analyst finds that lower conversion accounts for most of the decline, while traffic and transaction value remained nearly stable.

Which classification best describes the analyst's decomposition?

Descriptive analytics, because it quantifies the size of the revenue decline and its component metrics

Diagnostic analytics, because it investigates which business factor most plausibly produced the revenue decline

Predictive analytics, because it uses component metrics to estimate revenue in a future reporting period

Prescriptive analytics, because it identifies the operational response that will restore the lost revenue

← Back to quizzes

Business Analytics Quiz

Business Analytics Quiz: Types Of Analytics

Practice Types Of Analytics in Business Analytics with focused quiz questions that help you check what you know, review explanations, and build confidence with test-style prompts.

What this quiz covers

This quiz focuses on Types Of Analytics, giving you a quick way to practice the rules, question types, and explanations that matter most for Business Analytics.

How to use this quiz

Try each quiz question before looking at the correct answer. Use the explanations to review missed ideas, then come back to similar questions until the pattern feels familiar.

All questions

Question 1

Which classification best describes the analyst's decomposition?

Descriptive analytics, because it quantifies the size of the revenue decline and its component metrics
Diagnostic analytics, because it investigates which business factor most plausibly produced the revenue decline (correct answer)
Predictive analytics, because it uses component metrics to estimate revenue in a future reporting period
Prescriptive analytics, because it identifies the operational response that will restore the lost revenue

8\%

decline) is already given by the dashboard. The analyst goes further, decomposing the decline into its drivers to pinpoint which factor caused it. Isolating conversion rate as the primary culprit is a classic "root cause" investigation. That's the hallmark of diagnostic analytics, making B the correct answer. A is tempting because the decomposition does involve quantifying component metrics, but descriptive analytics stops at measuring and summarizing what occurred — it doesn't investigate causality. Reporting the

8\%

drop is descriptive; explaining why it dropped is not. C is wrong because nothing in the analysis projects future revenue. Predictive analytics requires a forward-looking model (regression, forecasting, etc.), none of which appear here. D misidentifies the work as prescriptive, but the analyst never recommends a corrective action — no pricing change, no marketing response, nothing operational. Prescriptive analytics goes beyond diagnosis to suggest decisions. A useful memory trick: think of the four tiers as a doctor visit — descriptive is taking your temperature, diagnostic is identifying the illness, predictive is forecasting recovery time, and prescriptive is writing the prescription. On exam questions, watch for whether the analysis stops at identifying the cause or goes further to recommend action — that boundary separates diagnostic from prescriptive every time.

Question 2

A bank trains a model using customer tenure, transaction activity, service complaints, and product holdings. For every active customer, the model estimates the probability that the customer will close all accounts during the next $90$ days. Managers receive the probabilities but no recommended retention offers.

What is the primary type of analytics represented by the model's output?

Descriptive analytics, because the input variables summarize each customer's current relationship with the bank
Diagnostic analytics, because the model variables may be associated with customers' reasons for leaving
Predictive analytics, because the output estimates the likelihood of a specified future customer event (correct answer)
Prescriptive analytics, because managers can use the scores when deciding which customers to contact

Explanation: When a question asks you to classify a type of analytics, anchor yourself to the four-tier framework: descriptive (what happened), diagnostic (why it happened), predictive (what will happen), and prescriptive (what should we do about it). The key is identifying what the model's output actually does — not what the inputs describe or what humans might do with the results afterward. Here, the model produces a probability that a customer will close all accounts within the next 90 days. That output is forward-looking and quantifies the likelihood of a future event. That is the textbook definition of predictive analytics, making C the correct answer. Choice A is tempting because the input variables — tenure, transaction activity, complaints, product holdings — do summarize a customer's current relationship. But analytics type is determined by the output, not the inputs. The output is a future probability, not a historical summary. Choice B misreads "diagnostic." Diagnostic analytics explains why past events occurred (e.g., why did churn spike last quarter?). Identifying variables that correlate with churn risk is part of model building, not the purpose of the output itself. Choice D is the trickiest distractor. Prescriptive analytics goes one step further by recommending a specific action (e.g., "offer this customer a fee waiver"). The passage explicitly states managers receive probabilities but no recommended retention offers — the model stops at prediction, not prescription. Study tip: On analytics-type questions, always evaluate the output, not the inputs or downstream human decisions. If the output says "here's the probability of X happening," that's predictive — full stop.

Question 3

A distributor's system calculates days of inventory coverage from current stock and recent demand. When coverage falls below five days, the system automatically recommends expedited replenishment from a designated supplier. The threshold and supplier were established in advance by management.

Which classification best applies to the system's final recommendation?

Descriptive analytics, because the recommendation begins with a current inventory-coverage metric
Diagnostic analytics, because crossing the threshold explains the underlying cause of low inventory
Predictive analytics, because low coverage may indicate that a future stockout could occur
Prescriptive analytics, because a decision rule converts the metric into a recommended operational action (correct answer)

Explanation: When a question asks you to classify a business intelligence system, focus on what the system does with data at its final step — not what data it starts with. The four analytics types form a progression: descriptive (what happened), diagnostic (why it happened), predictive (what will happen), and prescriptive (what you should do about it). This system ends by issuing a specific operational recommendation — reorder from a designated supplier via expedited shipping — based on a pre-set decision rule. That is the hallmark of prescriptive analytics: translating a metric or condition into a concrete, actionable decision. Answer D correctly identifies this. The rule ("if coverage < 5 days, then recommend expedited replenishment") is exactly the kind of encoded management logic that makes a system prescriptive. Answer A tempts you because the process begins with a descriptive metric (days of inventory coverage). But classification depends on the system's output and purpose, not its inputs. Starting with descriptive data doesn't make the system descriptive. Answer B mistakes the threshold trigger for a diagnostic finding. Diagnostic analytics investigates why something happened — root-cause analysis. Crossing a threshold is a condition check, not a causal explanation. Answer C is the trickiest distractor: yes, low coverage implies a future stockout risk, but the system never forecasts a probability or models a future state. It skips prediction entirely and jumps straight to a recommendation. A useful rule of thumb: if the system tells you what to do, it's prescriptive. Prescriptive analytics always features decision rules, optimization logic, or policy-driven recommendations converting data into action.

Question 4

A retailer is choosing among no discount, a $10\%$ discount, and a $20\%$ discount. A demand model forecasts sales of $500$ , $620$ , and $680$ units, respectively. Unit contribution margins would be $40$ , $36$ , and $32$ currency units. Company policy is to recommend the option with the greatest forecast contribution.

After applying the policy, how should the recommendation be classified?

Predictive analytics recommending no discount, because its forecast contribution is $20{,}000$ currency units
Prescriptive analytics recommending the $10\%$ discount, because its forecast contribution is $22{,}320$ currency units (correct answer)
Diagnostic analytics recommending the $20\%$ discount, because its forecast sales volume is the highest
Descriptive analytics recommending the $10\%$ discount, because contribution is calculated from supplied figures

Explanation: When you see a question blending forecasting with decision-making, ask yourself: is the system describing the past, diagnosing causes, predicting outcomes, or recommending an action? That last category — using a model plus a decision rule to prescribe what to do — is prescriptive analytics. Here's how the math works. Multiply each option's forecasted units by its contribution margin:

No discount: $500 \times 40 = 20{,}000$
10% discount: $620 \times 36 = 22{,}320$
20% discount: $680 \times 32 = 21{,}760$

The

10\%

discount yields the highest forecast contribution at

22{,}320

currency units. The company policy then prescribes that option — making this prescriptive analytics. So B is correct. A is wrong on two counts: the contribution figure of

20{,}000

belongs to the no-discount option, which isn't the maximum, and calling it "predictive" misidentifies the analytics type. Predictive analytics generates forecasts; it doesn't make recommendations. C is wrong because choosing the highest volume (680 units) ignores margin, producing a suboptimal contribution of

21{,}760

. The policy specifies greatest contribution, not greatest volume. "Diagnostic analytics" is also wrong here — that type explains why something happened historically. D is wrong because descriptive analytics summarizes historical data; it doesn't optimize decisions from a forecasting model and a decision rule. A handy tip: remember the analytics hierarchy — Descriptive → Diagnostic → Predictive → Prescriptive — in order of complexity. Whenever a model plus a decision rule recommends an action, that's prescriptive, no matter how simple the math.

Question 5

A service center's monthly report states that customer satisfaction was $92\%$ , compared with $94\%$ in the prior month and a target of $95\%$ . It flags the target gap but contains no customer-level analysis, forecast, explanation, or recommended response.

Which type of analytics does the report primarily provide?

Descriptive analytics, because it summarizes performance and compares the KPI with benchmarks (correct answer)
Diagnostic analytics, because the difference from target explains why satisfaction underperformed
Predictive analytics, because the prior-month comparison establishes the direction of future satisfaction
Prescriptive analytics, because flagging a target gap communicates that corrective action is required

Explanation: Whenever you see a question about analytics types, anchor yourself to the four-tier framework: descriptive (what happened?), diagnostic (why did it happen?), predictive (what will happen?), and prescriptive (what should we do?). The key is identifying what the report actually does, not what someone could theoretically do with the information. This report does exactly one thing: it presents a current metric (

92\%

), compares it to a prior period (

94\%

) and a target (

95\%

), and notes the gap. That is pure summarization and benchmarking — the definition of descriptive analytics, making A correct. B is tempting because the gap feels like an explanation, but flagging that satisfaction missed target by

3\%

does not explain why it missed. Diagnostic analytics requires investigating root causes — for example, identifying which customer segments or service failures drove the decline. No such analysis exists here. C misreads "prior-month comparison" as forecasting. Looking backward at last month's number is historical reporting, not a statistical model projecting future outcomes. Predictive analytics requires forward-looking inference, not trend observation. D confuses implication with prescription. Yes, a rational manager might take corrective action after reading this report, but the report itself recommends nothing. Prescriptive analytics explicitly tells you what to do — optimization models, decision recommendations, or action plans. Flagging a gap is not the same as prescribing a solution. The key study tip: focus on what the output actually contains, not what could be inferred from it. Exam distractors frequently exploit the gap between what data implies and what analysis delivers.

Question 6

A forecasting team reviews a demand model after the quarter ends. It calculates the model's mean absolute error, counts the percentage of forecasts within tolerance, and compares these metrics with those from the previous model. The team does not analyze the causes of individual errors or generate new forecasts.

How should this review be classified?

Descriptive analytics, because it summarizes realized model performance using historical error metrics (correct answer)
Diagnostic analytics, because calculating errors identifies why the forecasts deviated from actual outcomes
Predictive analytics, because performance metrics derived from a forecasting model carry forward the model's predictive purpose
Prescriptive analytics, because comparing two models on error metrics is sufficient to select the model that should be deployed going forward

Explanation: Whenever you see a question about analytics classification, anchor yourself to this hierarchy: descriptive (what happened?), diagnostic (why did it happen?), predictive (what will happen?), and prescriptive (what should we do?). The key is identifying what the activity itself accomplishes, not what the inputs or tools are associated with. In this scenario, the team calculates MAE, measures the percentage of forecasts within tolerance, and compares those figures to a prior model. Every action is a summary of past performance — what already happened. That makes this descriptive analytics (A). The team is aggregating and presenting historical error data to characterize model behavior. Nothing in the passage involves discovering causes, generating new forecasts, or issuing a recommendation. B is wrong because diagnostic analytics requires investigating why errors occurred — tracing deviations back to root causes like seasonality, data quality, or model misspecification. The team here never examines individual errors or their sources; it only reports aggregate metrics. C is a subtle trap. Just because a forecasting model was used to produce the forecasts doesn't mean evaluating those forecasts retrospectively is predictive analytics. The review activity looks backward, not forward. The model's purpose doesn't transfer to the audit of its results. D misreads what prescriptive analytics requires. Comparing metrics is still descriptive. Prescriptive analytics would explicitly recommend an action — such as deploying Model B — and ideally optimize that decision using constraints or objectives. Comparison alone doesn't cross that threshold. Study tip: Watch for questions that conflate the tool used with the type of analysis performed. Always classify analytics by what the activity does, not what it involves.

Question 7

A manufacturer develops four outputs in sequence: first, a report of last year's product-return rates; second, an analysis of whether defect type and production shift are associated with returns; third, a model estimating next month's return probability for each shipment; and fourth, a system allocating limited inspection hours to shipments to minimize expected return costs.

Which sequence correctly classifies the four outputs?

Descriptive, predictive, diagnostic, prescriptive
Diagnostic, descriptive, predictive, prescriptive
Descriptive, diagnostic, prescriptive, predictive
Descriptive, diagnostic, predictive, prescriptive (correct answer)

Explanation: Business analytics recognizes four distinct types of analytics, each answering a progressively deeper question: What happened? (descriptive), Why did it happen? (diagnostic), What will happen? (predictive), and What should we do? (prescriptive). When you see a question like this, map each output to the question it answers. Walking through the passage: the first output reports last year's return rates — it summarizes historical data, making it descriptive. The second investigates whether defect type and production shift are associated with returns — it's explaining causes, which is diagnostic. The third builds a model estimating future return probability — forecasting is the hallmark of predictive analytics. The fourth allocates inspection hours to minimize costs — it's recommending an optimal action, which is prescriptive. That sequence — descriptive, diagnostic, predictive, prescriptive — confirms answer D. Answer A swaps diagnostic and predictive, placing "predictive" second and "diagnostic" third. Identifying why defects correlate with returns is causal explanation (diagnostic), not forecasting. Answer B makes the first output diagnostic, but a simple report of historical rates describes what happened — there's no causal investigation yet. Answer C is the most tempting trap: it places prescriptive before predictive. You can't optimize an action (prescriptive) without first building a forecast (predictive) — the model estimating return probability logically precedes the scheduling system that uses those probabilities. A handy memory anchor: think of the four types as a staircase — Describe, Diagnose, Predict, Prescribe (DDPP). Each step requires the one before it, so order always matters on questions like this.

Question 8

An airline model estimates passenger demand and operating profit under each of six possible fare levels for an upcoming route. The output provides conditional forecasts for every fare but neither ranks the fares nor recommends one. Executives will apply their own risk preferences before choosing.

How should the model's output be classified?

Descriptive analytics, because the six scenarios organize information about fares and operating profit
Diagnostic analytics, because differences among the scenarios reveal why historical passengers selected certain fares
Predictive analytics, because it forecasts outcomes conditional on several possible management decisions (correct answer)
Prescriptive analytics, because evaluating decision scenarios is sufficient to identify the action management should take

Explanation: When you see a question asking how to classify an analytics output, your first move should be to anchor yourself to the four-tier analytics framework: descriptive (what happened), diagnostic (why it happened), predictive (what will happen), and prescriptive (what should we do). The key is identifying what the model is doing, not just what its output looks like on the surface. Here, the model takes six possible fare levels as inputs and produces conditional forecasts — "if we set this fare, here is the expected demand and profit." That structure is the hallmark of predictive analytics: it generates forward-looking estimates based on specified conditions. No ranking or recommendation is produced; executives must still apply their own judgment. This confirms the answer is C. A is tempting because the output does organize information into a tidy table of scenarios. However, descriptive analytics summarizes historical data — what has already occurred. This model is projecting future outcomes, not reporting past ones. B misreads the scenarios as an explanation of past passenger behavior. Diagnostic analytics investigates root causes of historical events. The model says nothing about why passengers chose certain fares before — it forecasts what they will do under hypothetical conditions going forward. D is the most dangerous distractor. It conflates evaluating decision scenarios with prescribing a decision. True prescriptive analytics goes further — it ranks options, optimizes across constraints, and explicitly recommends an action. Because executives still apply their own risk preferences, no prescription has been made. Study tip: Ask yourself whether the model is making a decision or informing one — prescriptive analytics does the former; predictive analytics does the latter.

Question 9

A telecommunications company studies last quarter's customer cancellations. A regression analysis finds that cancellations were substantially more common among customers who experienced repeated service outages, even after controlling for tenure and plan type. Because the data are observational, the analyst does not claim that outages definitively caused every cancellation.

What is the strongest defensible classification of the analysis?

Descriptive analytics, because an analysis restricted to completed cancellations primarily serves to summarize historical activity
Diagnostic analytics, because it investigates factors associated with why cancellations were more likely to occur (correct answer)
Predictive analytics, because regression is a modeling technique most commonly applied to forecasting future outcomes
Prescriptive analytics, because identifying outage-related cancellations specifies which network investments to approve

Explanation: When you see a question about analytics classification, anchor yourself to the four-tier framework: descriptive (what happened), diagnostic (why it happened), predictive (what will happen), and prescriptive (what should we do). The key is matching the purpose and output of the analysis to the right tier — not the technique used or the data type. Here, the analyst is using regression to uncover factors associated with past cancellations — specifically, why certain customers were more likely to cancel. That investigation into underlying drivers is the hallmark of diagnostic analytics, making B the correct answer. The caution about causation doesn't disqualify it; diagnostic analytics works with association and correlation all the time. A is wrong because descriptive analytics merely summarizes what occurred — counts, averages, totals. This analysis goes further by examining why cancellations clustered around outage-heavy customers, which moves it beyond simple summarization. C is a classic trap: it conflates the tool (regression) with the purpose. Regression can be used descriptively, diagnostically, or predictively depending on the goal. Since the analysis looks backward at completed cancellations to find explanatory factors — not forward to forecast new ones — it is not predictive. D misreads the scope of the analysis. The study identifies an association; it does not evaluate investment options, model tradeoffs, or recommend specific actions. Prescriptive analytics requires an optimization or decision-recommendation component, which is absent here. Your strategy tip: never classify analytics by the technique alone — always ask what question the analysis is designed to answer. That question ("why?") is your fastest signal for diagnostic analytics.

Question 10

An online retailer randomly assigns customers to its existing checkout process or a redesigned process. The analyst estimates that the redesign increased completed purchases by $2.1$ percentage points and reports a confidence interval. The report evaluates whether the redesign caused the observed difference but stops short of recommending a rollout.

Which type of analytics is most directly represented by this analysis?

Descriptive analytics, because it reports the observed conversion rates for the two randomly assigned groups
Diagnostic analytics, because it tests whether a controlled change explains the observed conversion difference (correct answer)
Predictive analytics, because the estimated lift can serve as a forecast of next quarter's conversion rate improvement
Prescriptive analytics, because the statistical evidence of an effect indicates which checkout process the retailer should deploy

Explanation: When classifying analytics, the key question is: what is the analysis trying to do? Descriptive analytics summarizes what happened, diagnostic analytics explains why or whether a specific cause produced an effect, predictive analytics forecasts future outcomes, and prescriptive analytics recommends a decision or action. This scenario involves a randomized experiment where customers were assigned to two groups, a lift of

2.1

percentage points was measured, and a confidence interval was constructed to evaluate whether the redesign caused the conversion difference. That causal investigation — using controlled assignment to explain an observed outcome — is the hallmark of diagnostic analytics, making B the correct answer. A is tempting but incomplete. Yes, the analyst reports observed conversion rates, but the analysis goes further by testing a causal explanation through random assignment and inferential statistics. Merely reporting rates would be descriptive; testing whether a controlled intervention explains the difference is diagnostic. C misreads the purpose of the lift estimate. A

2.1

percentage-point finding describes what happened in this experiment — it doesn't model or forecast future conversion rates. Using the result as a forecast would be a separate predictive step not described in the passage. D is the subtlest trap. Prescriptive analytics doesn't just identify an effect; it actively recommends or optimizes a course of action. The passage explicitly says the report stops short of recommending a rollout, which disqualifies prescriptive. A useful rule of thumb: if the analysis concludes with "here's what caused it" but no actionable recommendation, it's diagnostic, not prescriptive. Watch for that distinction — it's a frequent source of confusion on analytics classification questions.