NUMERICAL REASONING TEST
Raising the bar
THE NEXT GENERATION OF NUMERICAL TESTING
Evidence shows there are two things that contribute to numerical performance: accuracy (answering correctly) and efficiency (time taken to do so).
The way that other tests on the market typically assess this is to set a time limit per question or ask candidates to complete as many questions as possible in a set amount of time.
When was the last time someone stood over your desk and requested a numerical answer in 30 seconds?
We believe that the adverse impact often seen in ability tests could be caused by the ‘time-limited’ presentation of the assessments, which puts candidates under stress. There is clear evidence from evolutionary and biobehavioural research that men’s and women’s responses to stress are different. We believe that the stress caused by ‘time-limited’ ability assessments favours men, and results in bias against women.
OUR APPROACH IS DIFFERENT
We determined that an innovative ‘time-recorded’ approach to ability assessment, as distinct from the traditional ‘time-limited’ approach, would allow us to overcome adverse impact, as well as providing a far more realistic work sample test for candidates.
We've developed an electronically-administered numerical reasoning test that is truly state of the art, built upon an in-depth literature and market review to get under the skin of numerical reasoning in a modern, commercial world.
Our review led to the definition of a unique model of numerical mindset, assessing the ability to:
- Identify relevant data to engage with and prioritise
- Comprehend a range of numerical information
- Interpret data presented in tables, graphs and modern visual formats (e.g. Excel)
- Manipulate numerical data (e.g. perform appropriate calculations)
- Decipher problems and make decisions, thus deducting the commercial or pragmatic meaning of the data.
A BETTER USER EXPERIENCE
Time recorded - not time limited! Our state of the science scoring algorithm, which accounts for the speed, accuracy and difficulty of each question, enables greater candidate differentiation than ever before.
Fair - eliminating the potential adverse impact of more traditional time-limited tests. No longer do you need to take it for granted that a numerical reasoning assessment will favour candidates from a specific demographic. Data on our test has shown no adverse impact across gender, ethnicity, and social mobility groups.
Adaptive - more precise measurement from fewer questions. With every question completed, the test gets closer and closer to identifying the candidate’s true level of ability. It continues to adjust the difficulty of the questions until the most accurate ability prediction is achieved.
Test security – a tailored set of questions. Candidates are presented with items from our ever-increasing item pool. This means that no items are over-exposed, increasing the security and integrity of the test. The chances of 2 candidates taking the same test is currently less than one in a billion!
Candidate experience – maximising test-taker engagement. A diverse set of questions and response formats provides a more relaxed and realistic test-taker experience.
TRY it for yourself
SHARE THIS PAGE