Spokeo·Data Scientist·Online Assessment (OA)·Intermediate

Intermediate

May 2026

Summary

Spokeo data scientist interview had a HackerRank notebook task using the Titanic dataset. Pretty classic EDA stuff but the open-ended nature of it meant you had to decide what mattered and actually explain your thinking, not just produce output.

Questions Asked (1)

Given a Titanic passenger dataset, perform an exploratory data analysis: compute survival rate broken down by sex, average age per passenger class, and the correlation between age and survival. Include labeled visualizations and a brief discussion of what the data shows.

Product Analytics & MetricsData Modeling

Author's notes

The EDA itself wasn't hard but I spent too long on the correlation piece.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by outlining a structured EDA plan: data loading and cleaning, then compute the required statistics (survival rate by sex, average age per class, correlation between age and survival). Create clear visualizations (bar charts, box plots, heatmap) with proper labels and titles, and conclude with a brief interpretation of the findings, highlighting any surprising patterns or data quality issues.

Pro tip: Always check for missing values in 'Age' and 'Cabin' before computing statistics, and explicitly state how you handle them (e.g., imputation or exclusion) because it affects the results and shows rigor. Also, consider that correlation between age and survival might be non-linear, so supplement with visualizations like survival rate by age bins.

1. Data Loading and Inspection

Load the dataset and inspect its structure, data types, and missing values. Identify key columns: Survived, Sex, Age, Pclass.

2. Data Cleaning and Preprocessing

Handle missing values (e.g., impute Age with median or mean, drop or impute Cabin) and ensure data types are correct. Document your decisions.

3. Compute Required Statistics

Calculate survival rate by sex (groupby Sex, mean of Survived), average age per passenger class (groupby Pclass, mean of Age), and correlation between Age and Survived (point-biserial correlation).

4. Create Visualizations

Generate labeled plots: bar chart for survival rate by sex, box plot or bar chart for average age per class, and a heatmap or scatter plot for age vs. survival. Ensure all axes, titles, and legends are clear.

5. Interpret and Discuss Findings

Summarize the results: e.g., females had higher survival rates, first-class passengers were older on average, and age may have a weak correlation with survival. Discuss potential reasons and limitations.

Key Points to Mention

Data cleaning: handling missing Age values and their impact on analysis
Survival rate by sex: likely higher for females, illustrating 'women and children first'
Average age per class: first class older, third class younger, reflecting socioeconomic differences
Correlation between age and survival: likely weak, but check for non-linear patterns (e.g., children prioritized)
Visualization best practices: clear labels, titles, and appropriate chart types
Discussion of confounding factors: sex, class, and age interactions

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.