Data Analytics Certification Practice Exam

Which of the following best describes structured data?

Structured data fits neatly into tables with rows and columns. Unstructured data, like images, video, or free text, doesn't follow a fixed schema, which is what the second option describes instead.

In Python, which data type stores a value like True or False?

bool is Python's Boolean type, used constantly in conditional logic and filtering.

Which Python data structure stores data as unique, unordered key-value pairs?

A dictionary maps keys to values (for example price: value), so you look things up by key rather than by position.

A dataset shows daily website visitors over a year. Most days see 200 to 400 visitors, but a few holiday sales days spike to 5,000+. Which measure of central tendency is most distorted by those spikes?

The mean is pulled toward extreme values. The median, the middle value once sorted, stays far more stable in a skewed distribution like this one.

What is the main purpose of exploratory data analysis (EDA)?

EDA happens early, getting familiar with distributions, missing values, and relationships before committing to a modeling approach.

Which of these is a categorical variable?

Categorical variables represent labels or groups, not numeric quantities. The other three are all numeric.

What does it mean to clean a dataset?

Data cleaning covers tasks like deduplication, handling missing values, and fixing formatting. It often takes up the largest share of a real analytics project.

Which Python loop repeats a block of code a specific, known number of times?

A for loop is ideal when you know how many iterations you need. A while loop suits conditions where the iteration count isn't known upfront.

Which of the following is an example of quantitative continuous data?

Height can take any value within a range, including fractions, which makes it continuous. Number of children is discrete, and T-shirt size and zip code are both categorical even though zip codes look numeric.

Which type of chart is best suited for showing how a single variable's value changes over time?

A line chart connects data points in order, which makes trends and changes over time easy to see. Bar charts compare categories, and pie charts show parts of a whole at a single point in time.

In hypothesis testing, what does the null hypothesis (H0) represent?

H0 is the default assumption of no real effect or relationship. A test checks whether there's enough evidence to reject it.

A p-value of 0.03 with a significance level (alpha) of 0.05 means:

0.03 is below the 0.05 threshold. The first option is a common misreading: a p-value isn't the probability that the null hypothesis is true.

What does a 95% confidence interval for a population mean actually tell you?

A confidence interval describes the reliability of the method over repeated sampling, not the probability for this one interval.

Which sampling method gives every member of a population an equal chance of selection?

Random sampling avoids selection bias by giving every unit an equal chance, which is key for results that generalize to the full population.

Per the Central Limit Theorem, as sample size increases, the sampling distribution of the sample mean tends to:

This is why many statistical tests can assume normality even when the underlying data isn't normal, as long as the sample is large enough.

Ice cream sales and drowning incidents are positively correlated. What's the most likely explanation?

This is the classic correlation isn't causation example. Hot weather is a confounding variable behind both.

In the simple linear regression y = 3 + 2x, what does the coefficient 2 represent?

In y = b0 + b1x, b1 is the slope. A one unit increase in x predicts a 2 unit increase in y.

In pandas, which method computes the average of a column for each group in another column (for example average sales per region)?

.groupby() splits data into groups, and chaining .mean() computes that statistic within each group.

What does a large standard deviation indicate about a dataset?

Standard deviation measures how spread out values are from the mean. A large standard deviation means the data points vary widely; a small one means they cluster close to the average.

In hypothesis testing, a Type I error occurs when you:

A Type I error is a false positive: rejecting a null hypothesis that was actually true. A Type II error is the opposite, failing to reject a null hypothesis that was actually false.

Multicollinearity in a multiple regression model refers to:

Highly correlated predictors make it hard to isolate each variable's individual effect, which inflates standard errors. It's typically flagged using Variance Inflation Factor (VIF), where values above 5 to 10 are a common red flag.

What's a common fix for multicollinearity?

Drop one of the correlated variables, combine them (for example via PCA), or use regularization to shrink their impact.

When is logistic regression the appropriate choice over linear regression?

Logistic regression is built for classification: predicting a categorical outcome, most often binary.

In a two-way ANOVA, what can you test that a one-way ANOVA cannot?

A two-way ANOVA adds a second categorical variable and lets you test whether the two variables interact, for example whether the effect of ad spend on sales depends on region.

An R-squared of 0.85 for a regression model means:

R-squared measures explained variance, not prediction accuracy. A high R-squared on training data can even be a red flag for overfitting.

What is the bias-variance tradeoff?

A high bias model misses real patterns. A high variance model fits training noise so closely it fails on new data. Good models sit between the two.

Ridge and Lasso regression both address overfitting by:

Both penalize large coefficients. Ridge shrinks them toward zero without eliminating them. Lasso can shrink some all the way to zero, which doubles as feature selection.

In a classification model, which metric matters most when false negatives are far more costly than false positives (for example missing a serious illness)?

Recall measures how many actual positives the model catches. When missing a true positive is costly, you optimize for recall, even at the expense of more false positives.

Why is feature scaling (such as standardization) often important before training many machine learning models?

Many algorithms, including distance-based models and anything trained with gradient descent, treat larger-scale features as more important by default. Scaling puts features on a comparable range so the model weighs them fairly.

Why do data scientists split data into training and test sets before evaluating a model?

Testing a model on data it has never seen estimates how it will actually perform in the real world. Evaluating only on training data risks a falsely optimistic result, since the model may have simply memorized that data.