Multi-Step Numerical Problems: How to Solve Them in Aptitude Tests
Multi-step numerical problems are among the most challenging question types in a numerical reasoning test. These problems require you to chain together two, three, or even four separate calculations before arriving at the correct answer, testing not only your mathematical ability but also your capacity to plan and maintain accuracy under time pressure.
Employers like Google, Deloitte, Unilever, JP Morgan, and the Civil Service all include multi-step problems in their recruitment assessments. Understanding how to approach them systematically is essential if you want to score well and advance past the screening stage.
What Are Multi-Step Numerical Problems?
A multi-step numerical problem is any question that requires more than one calculation to reach the final answer. Instead of reading a single value from a table and performing one operation, you need to extract multiple data points, apply several operations in sequence, and carry intermediate results forward to the next step.
A typical example might ask you to calculate total revenue from a sales table, determine the percentage increase between two years, and then apply that growth rate to project a future figure. Each step feeds into the next, and an error at any stage cascades through the entire solution.
These problems appear frequently in assessments from major test providers including SHL, Cubiks/Talogy, Kenexa, and Aon. They are especially common in tests designed for roles in finance, consulting, data analysis, and management. The Civil Service numerical reasoning test, for instance, regularly includes questions where candidates must combine data from two separate charts before performing a final calculation.
💡Multi-step problems test your ability to plan a sequence of calculations, not just your raw arithmetic skills. Candidates who approach them with a clear method consistently outperform those who dive straight into the numbers.
Why Employers Use Multi-Step Problems
Employers include multi-step problems because they simulate real workplace analytical thinking. At Deloitte, consultants analyse financial data involving multiple variables. At JP Morgan, analysts work through models where one incorrect assumption propagates errors across an entire spreadsheet. Google expects employees to reason through data-driven problems that rarely have single-step solutions. The Civil Service requires policy analysts to cross-reference multiple datasets.
These questions also differentiate candidates at the top of the scoring range. Most candidates handle single-step calculations accurately, but multi-step questions separate those who maintain precision across a chain of operations from those who lose accuracy as complexity increases.
If you are preparing for a numerical reasoning test, expect multi-step problems to make up a significant portion of the harder questions.
Types of Multi-Step Numerical Problems
Multi-step problems come in several common formats. Recognising the type of problem you are facing helps you plan your approach before you start calculating.
| Problem Type | What It Involves | Example Scenario | Typical Step Count |
|---|---|---|---|
| Percentage chain | Applying successive percentage changes to a base value | Calculate revenue after a 12% increase followed by an 8% decrease | 2-3 steps |
| Ratio and proportion | Using ratios to find missing values, then applying further calculations | Split profit by a 3:2 ratio, then calculate tax on each share | 2-3 steps |
| Data extraction and comparison | Reading values from tables or charts and combining them | Compare total sales across regions using data from two separate tables | 3-4 steps |
| Currency or unit conversion | Converting units before performing the main calculation | Convert USD revenue to GBP, then calculate per-unit profit margin | 2-3 steps |
| Weighted average | Combining different quantities with different weights | Calculate the average salary across three departments with different headcounts | 3-4 steps |
| Compound growth | Applying a growth rate over multiple periods | Project company revenue three years forward at 7% annual growth | 2-3 steps |
Understanding these categories helps you anticipate what the question is asking before you read the answer choices. For more practice with related question types, see the guide on data interpretation questions explained.
💡Most multi-step problems fall into a handful of recognisable patterns. Learning to identify the pattern quickly saves you valuable planning time during the test.
A Step-by-Step Method for Solving Multi-Step Problems
The most reliable way to handle multi-step problems is to follow a consistent method every time. This approach works regardless of the specific question type or test provider.
Step 1: Read the entire question first. Many candidates start calculating as soon as they see a number. Read the full question, including answer choices, so you understand what the final answer should look like. Is it a percentage, a currency amount, a ratio, or a raw number?
Step 2: Identify every piece of data you need. Scan the tables, charts, or text for the specific values required. Multi-step problems often include irrelevant data. Ignoring it saves time and reduces confusion.
Step 3: Map out your calculation chain. Before calculating, sketch a brief plan: "Find Year 1 revenue, calculate 15% increase, subtract costs, divide by units." This roadmap prevents you from losing track mid-calculation.
Step 4: Execute each calculation carefully. Work through your plan one step at a time. Write down each intermediate result with its units. Do not try to hold multiple numbers in your head simultaneously.
Step 5: Verify your final answer. Do a quick sense check. Does the number look reasonable? If you calculated a 300% profit margin for a retail company, something has gone wrong.
This method becomes automatic with repetition. Candidates who follow a structured process finish faster than those who jump between steps randomly, because they avoid costly mistakes that force restarts.
Common Mistakes and How to Avoid Them
Errors in multi-step problems are rarely about not knowing the maths. They are almost always process errors: mistakes in how you organise and execute your calculations. Here are the most frequent pitfalls and how to avoid them.
Premature rounding. Rounding an intermediate result might seem harmless, but the error compounds with each step. A small rounding difference in step one can shift your final answer enough to select the wrong option. Keep full precision until the final answer.
Misreading the question. Multi-step questions sometimes ask for the difference between two values, but candidates calculate only one and select it. Always re-read the question after completing your calculations.
Skipping units. When a problem involves mixed units like thousands and millions or different currencies, failing to convert properly leads to answers off by orders of magnitude. Write the unit next to every intermediate value.
Carrying forward the wrong value. In a chain of three calculations, it is easy to use the result from step one instead of step two. Labelling each intermediate result clearly prevents this.
Running out of time. Multi-step problems consume more time than single-step questions. Set a mental time limit of around 90 seconds per question and move on if you exceed it.
💡The majority of errors in multi-step problems come from poor process, not poor maths. A systematic approach with clearly labelled intermediate results eliminates most mistakes.
For a deeper look at calculation pitfalls, read about common numerical reasoning mistakes.
Worked Examples: Putting the Method Into Practice
Seeing the method applied to realistic problems helps cement the approach. Here are two examples at different difficulty levels.
Example 1: Percentage Chain (Two Steps)
A company reported revenue of 4.2 million pounds in 2023. Revenue grew by 12% in 2024 and then declined by 5% in 2025. What was the 2025 revenue?
- Step 1: 2024 revenue = 4,200,000 x 1.12 = 4,704,000
- Step 2: 2025 revenue = 4,704,000 x 0.95 = 4,468,800
The answer is 4,468,800 pounds. Note that a 12% increase followed by a 5% decrease does not produce a net 7% increase. This is a common trap.
Example 2: Data Extraction and Comparison (Three Steps)
A table shows that Department A has 45 employees with an average salary of 38,000 pounds, and Department B has 30 employees with an average salary of 42,000 pounds. What is the overall average salary across both departments?
- Step 1: Total salary cost for A = 45 x 38,000 = 1,710,000
- Step 2: Total salary cost for B = 30 x 42,000 = 1,260,000
- Step 3: Overall average = (1,710,000 + 1,260,000) / (45 + 30) = 2,970,000 / 75 = 39,600
The answer is 39,600 pounds. A common mistake is to simply average 38,000 and 42,000 to get 40,000, which ignores the different department sizes.
Start practising multi-step problems today to build speed and accuracy with these question types.
Time Management Strategies for Multi-Step Questions
Time pressure is the defining challenge of numerical reasoning tests. Unilever, Google, and JP Morgan all use tightly timed assessments where every second counts. Multi-step problems consume more time per question, so you need a deliberate strategy for managing your time across the full test.
Triage your questions. On your first pass, answer the single-step questions you can solve quickly. Flag multi-step problems for a second pass to collect easy marks first.
Set per-question time limits. If your test has 20 questions in 25 minutes, allow up to 90 seconds for multi-step problems and aim to complete simpler questions in 60 seconds to create that buffer.
Use estimation to eliminate options. Before calculating precisely, estimate the ballpark answer. If the options are 12,400, 15,800, 23,600, and 31,200, and your estimate suggests around 16,000, you can immediately eliminate two options.
Know when to guess and move on. If you have been working on a problem for over two minutes without progress, select the most likely answer and move on. One missed question is better than three missed questions.
For additional speed-building techniques, explore fast calculation strategies.
💡Strong time management in multi-step problems is about strategic choices, not just fast arithmetic. Triage, estimate, and set firm time limits to maximise your total score.
Frequently Asked Questions
How do I know which step to do first in a multi-step problem?
Identify the final answer the question asks for, then work backwards. Determine what intermediate values you need. For example, if the question asks for profit margin, you need profit, and to get profit you need revenue and costs. This reverse-engineering approach builds a clear sequence and prevents unnecessary calculations.
Should I round intermediate results in multi-step calculations?
No. Keep intermediate results as precise as possible throughout the calculation chain. Rounding at early stages introduces compounding errors. For instance, rounding 14.7% to 15% in step one might shift your final answer enough to select the wrong option. Only round your final answer when the question explicitly requests it.
What if I get stuck on one step of a multi-step problem?
Move on to the next question and return later if time permits. Spending three or four minutes on a single problem is rarely worthwhile when other questions of equal value might take 60 seconds. Flag the question and revisit it after completing the rest of the test.
How many steps do multi-step problems typically involve?
Most involve two to four distinct calculations. Assessments from providers like SHL and Kenexa rarely exceed five steps, though each step may require reading data from tables or charts. The underlying arithmetic at each step is usually straightforward: addition, subtraction, multiplication, division, or percentage calculations.
Are multi-step problems harder than single-step problems?
They are more time-consuming but not necessarily harder mathematically. Each individual calculation is typically straightforward. The challenge lies in maintaining accuracy across multiple operations, managing your time, and avoiding process errors like carrying forward the wrong intermediate value.
Can I use a calculator for multi-step numerical problems?
This depends on the test provider and employer. Many assessments from SHL and Aon provide an on-screen calculator. Others, particularly those used by JP Morgan and Deloitte, may restrict calculator use to test mental arithmetic. Always check your test invitation for the specific rules.
Start Preparing for Multi-Step Numerical Problems
Multi-step numerical problems reward preparation more than any other question type. The candidates who score highest have practised enough to recognise common patterns, execute a reliable method under pressure, and manage their time across the full test.
Get started with the complete numerical reasoning package at Assessment-Training.com and build the skills that employers at Google, Deloitte, JP Morgan, and the Civil Service are looking for.