Data Analyst Interview Questions for Freshers

Prepare to describe how you made an answer trustworthy. These questions are written for freshers in India preparing for data analyst, MIS and reporting interviews. They are for practice, and each should be answered with an example from your own work. For every question we note what a strong answer shows and what a weak answer sounds like. The pattern across all of them is the same: interviewers want to know whether you check your numbers and whether you can explain them to someone who did not do the work.

Data understanding and quality

What do you check when you receive a new dataset?

Start with the question, the source, what one row represents and the permitted use. Then inspect types, missing values, duplicates, date formats and units. In a fictional orders table, check whether one row is an order or an item line before counting sales activity.

A strong answer shows a repeatable checklist and one real example of a problem you found. A weak answer jumps straight to "I would make a dashboard".

How do you handle missing data?

Investigate its meaning first. A blank return date on an open library loan is different from a missing date on a completed loan. Decide whether to exclude, retain, flag or impute based on the question, and document the effect on your totals.

A strong answer distinguishes "missing because not applicable" from "missing because not recorded". A weak answer says "I would fill it with the average" for every case.

How do you know your numbers are right?

Reconcile against a known total, recompute a sample by hand, and compare row counts at each step. Explain the denominator of every rate you report.

A strong answer names a specific check you ran. A weak answer relies on the tool having produced a number.

SQL

Why did a join increase the total?

Explain one-to-many and many-to-many relationships. A customer table with repeated IDs can duplicate every matching order. Compare key uniqueness and row counts before and after the join, then aggregate to the intended grain or repair the relationship.

A strong answer avoids solving it by adding DISTINCT, which can hide the fault or remove legitimate repeated values. A weak answer does exactly that.

What is the difference between WHERE and HAVING?

WHERE filters rows before grouping; HAVING filters groups after aggregation. Give an example: customers with more than three orders needs HAVING on a count.

A strong answer includes a short worked query. A weak answer recites the definition without an example.

When would you use a window function?

When you need a calculation across related rows without collapsing them, such as a running total, a rank within a category or comparing a row to the previous month.

A strong answer explains the partition and ordering. A weak answer confuses it with GROUP BY.

How would you find the second highest order value per customer?

Describe a window function with ROW_NUMBER or DENSE_RANK partitioned by customer and ordered by value, then filter where the rank is two. Mention how ties are handled.

A strong answer talks about ties. A weak answer forgets the partition.

Analysis and reasoning

Which chart would you choose?

Decide by the comparison. A line shows change over time; bars compare categories; a table supports exact lookup. Explain labels, units and why your choice helps the audience.

A strong answer starts from the question the reader has. A weak answer picks a chart because it looks sophisticated.

How would you explain an unexpected result?

Recheck definitions, transformations and source totals first. Then examine segments and possible explanations. Separate an observed pattern from a causal explanation. Sales growth may reflect a change in product mix rather than more customers. State what evidence would distinguish the explanations.

A strong answer begins with "I would first check whether the number is correct". A weak answer begins with a story about why it happened.

What is the difference between correlation and causation, in your own words?

Two things moving together do not mean one caused the other. Give an everyday example and then a business one, such as discounts and sales volume both rising in a festive month.

A strong answer names a confounding factor. A weak answer stops at the textbook definition.

Explain mean, median and when each misleads.

The mean is pulled by very large or very small values; the median is not. Order values in retail often have a long tail, so the median order value may describe a typical customer better.

A strong answer gives a case where you chose one over the other. A weak answer cannot say why the two differ.

Communication and working style

How do you prioritise a request?

Clarify the decision, deadline, required accuracy and available data. Offer a useful bounded answer if the full request is too broad. State assumptions explicitly.

A strong answer shows you push back politely and still deliver something useful. A weak answer says yes to everything.

How would you explain a finding to a manager who does not use data?

Lead with the observation in one sentence, then the evidence, then the limitation, then the suggested next step. Avoid tool names and jargon.

A strong answer is short and ordered. A weak answer walks through every step you took.

Tell us about a time your analysis was wrong.

Everyone has one. Describe the error, how you found it, how you corrected it and what check you added afterwards.

A strong answer shows ownership and a new habit. A weak answer claims it never happened.

How to talk about your project

Prepare a three-minute version and a thirty-second version. Cover the question, the input data, one quality problem you found, the calculation and the final finding. Bring a small worked example that demonstrates why your result is correct, for example a hand calculation for one month that matches your query.

Say what the data was. If it was synthetic, say so; that is normal for a fresher and interviewers know it. Say what you would do next with more time. If asked about a tool you have not used, connect the underlying concept to familiar work and describe how you would learn the interface. Do not invent experience.

Questions to ask the interviewer

• What does a typical week look like for a fresher in this team?
• Which tools does the team use for data, reporting and sharing results?
• How do analysts here check that a report is correct before it goes out?
• Who are the main people who consume the team's analysis?
• How is feedback given in the first few months?

Asking about data quality and audiences signals that you think like an analyst. Avoid asking only about leave policy or salary in the first round.

Free practice resources

Kaggle Learn's Intro to SQL and Advanced SQL are free and browser based. SQLBolt gives quick interactive repetition. Mode's SQL tutorial covers analytical SQL with worked examples. Use the PostgreSQL tutorial once you set up a local database.

How GullyMentor approaches it

On the data analytics track (/career-tracks/data-analytics) a mentor asks these questions about your own project, points out where an answer is vague and makes you repeat it until it is clear. Rehearsal with someone who has sat on the other side of the table is the whole point. We do not guarantee jobs, interviews or salaries.

Sources & further reading

• PostgreSQL SQL tutorial
• Kaggle Learn: Intro to SQL (free)
• Kaggle Learn: Advanced SQL (free)
• SQLBolt interactive SQL lessons
• Mode SQL tutorial

Examples and practice recommendations are educational illustrations. Tool documentation and programme inclusions should be checked for your intended use.