Data Science Quiz

Statistical analysis, modeling, and the data-science workflow behind turning data into insight.

This category currently has 100 questions in the SERVBG quiz bank. Below are a few sample questions — the full interactive quiz shuffles through the whole set with instant scoring.

Sample questions

A two-sample t-test yields p=0.03 with significance level α=0.05. Which statement is correct?

The effect size is large enough to be practically significant.
We accept H1 with 95% confidence because p < α.
We reject H0; the probability of observing a test statistic this extreme under H0 is 3%.
There is a 3% probability that H0 is true.
There is a 97% probability that H1 is true.

In hypothesis testing, a Type II error (β) occurs when:

The test statistic falls within the critical region under H0.
H0 is true but we reject it.
The p-value exceeds the significance level and H0 is true.
The confidence interval contains the null parameter value.
H0 is false but we fail to reject it.

An A/B test is designed with 80% power to detect a minimum detectable effect of 2% lift. Halfway through, you see p=0.06 and stop early. The primary concern is:

Stopping early always reduces variance and is statistically valid.
The test is underpowered because power was set at 80%.
The minimum detectable effect threshold must be recalculated after peeking.
Power calculations are irrelevant once data collection begins.
Peeking inflates the Type I error rate beyond the nominal α.

You run 50 independent hypothesis tests at α=0.05. Using Bonferroni correction, the per-test threshold becomes:

0.0025
0.001
0.0005
0.05
0.01

The Benjamini-Hochberg (BH) procedure controls which error rate, and what does it guarantee?

Family-wise error rate (FWER); the probability of any false positive across all tests is ≤ q.
Posterior error probability; the Bayesian probability of false discovery is ≤ q.
False discovery rate (FDR); the expected proportion of rejected H0s that are false positives is ≤ q.
Generalized error rate; the proportion of true nulls incorrectly rejected among all hypotheses is ≤ q.
Per-comparison error rate; each individual test has Type I error ≤ q.

Start the full quiz ›