Scoring method

Mission

It is our mission to provide the most accurate online intelligence test in the world. This test is free of charge for anyone using it for non-professional or personal reasons.

Vision

This test is aimed at giving everyone in the world a fair, transparent, and accurate method to determine their intelligence in a standardized way.

Fair

It is our goal to create a level playing field for everyone in the world:
  • by providing a culturally fair test with questions that use the simplest language and academic knowledge. We believe it is impossible to create a test free of cultural bias, but we try to make it as fair as possible.
  • by giving people the opportunity to take the test as often as they’d like. Some people have performance anxiety, had a sleepless night, or just need a bit more practice than others to feel comfortable when taking the test. We don’t think these things should be reflected in an intelligence score. We believe everybody should have enough chances to reach their ‘true’ intelligence score potential.
  • by providing open access to more questions across vast numbers of categories than anyone in the world could ever memorize. Plus, everybody worldwide has the same access to the questions that are used in the test.

Transparent

We believe that personal intelligence is a delicate subject. Therefore, we want to be clear that an intelligence score will always reflect subjective choices about what intelligence is. It is our goal to be clear about these choices:
  • by providing open access to all the questions that are used to determine someone's intelligence. After every test, we provide the metadata from the questions that have been answered by the candidate (average ratio correct and average time needed from all other users who answered the questions).
  • by being open about our scoring method. It is our goal to make the scoring method as simple as reasonably possible, to ensure that everybody understands how they can reach their own best results.
  • by giving free access to test results to all non-professional users. We believe that everybody should be able to see the results of any test they have taken. Therefore, we will never ask for a payment from individual users for them to see their test results.

Accurate

We make it our mission to provide the most accurate online intelligence test in the world:
  • By constantly calculating the predictive value of every question for every user to give the most accurate estimate of someone's intelligence in as little time as possible.
  • By testing on an immense scale. Our test has been taken by more than 2 million users from all over the world. This method provides a very accurate data compilation of all the individual questions and test results that are used to determine the intelligence score.
  • By constantly updating all the content. We add new questions and categories while removing questions that have a lesser predictive value on a daily basis.

Scoring method

Goals of the scoring method

It is our goal to provide a clear scoring method that:
  • Is easy to understand
  • Rewards correct answers
  • Rewards working quickly
  • Discourages guessing

Adaptive Test

The level of the questions in the test is adjusted to your level during the test. If you answer the first question correctly, the next question will be a little more difficult. If you give the incorrect answer to a question, the next question will be a little bit less difficult.

Data per question

For each question, we track the average time that it takes people to answer the question (average time) and the percentage of people who answered the question correctly (percentage correct).

Points earned per correct answer For every question you answer correctly you earn:

Average time * (1 - percentage correct) points

This ensures you earn more points for answering a difficult question correctly than for answering an easy question correctly. It also makes sure you earn more points for questions that take the average user a longer time to answer.

‘Points’ lost per correct answer
For every question you answer incorrectly you lose: Average time * percentage correct points

This ensures you lose more points for answering an easy question incorrectly than for answering a difficult question incorrectly.

‘Raw score’
You get 20 minutes to earn as many points as possible. The total amount of points you have earned during the test is called your “raw score”.

Percentile score

We compare your raw score to the raw scores of all of the other users who finished the test in order to calculate your “percentile score”. This calculation reflects the percentage of users who finished the test and had a lower raw score than yours.

For example, a percentile score of 67% means that you performed better (earned more points) than 67% of all people who ever finished the test.

IQ score

From the percentile score, we calculate your IQ score. We make this calculation of your IQ score with a median of 100 and a standard deviation of 15. This is the same calculative method used by other IQ testers, such as the Wechsler Adults Intelligence Scale.

Item Response Theory

Since not all users answer the same questions and some questions will be answered more or less frequently by users with a higher or a lower intelligence level, mathematical models are used to give the most accurate estimation of the difficulty of the questions. For the sake of clarity, we use the term “average percentage correct” without mentioning the Item Response Theory models that are used to calculate this percentage.

Limitations to online intelligence tests

We are well aware that there are certain limitations to the accuracy of online intelligence tests. It is our goal to be open about these limitations while limiting their impact as much as possible.

The main limitation of online intelligence tests and the main solutions we see:

People don’t take the test themselves We provide free access to our intelligence tests for people to be able to assess their own intelligence. This means each person has the responsibility to take the tests without help from others. Otherwise, you are just fooling yourself. So, we encourage each person to take the test without help. If people want to help you with your test, encourage them to take their own test.

People memorize the test

Most intelligence test publishers keep their questions “secret”. They do this to limit the number of people who have knowledge about the questions before they take the test. We don’t believe in this strategy. In our digital age, it is very difficult to keep such content secret from a bigger public. Some people will have more knowledge about the content of the questions than others, creating an uneven playing field.

We provide open access to all of our questions to everybody. We have created over 100,000 questions in more than 20 subcategories with thousands of different patterns and parameters. This makes it virtually impossible to memorize the questions. Of course, there is a certain advantage to studying the different types of questions, but this advantage is the same for everybody. Plus, we add new content daily to limit the possibilities for an individual to memorize the test.

People are nervous / had a bad night / have performance anxiety

People can take the intelligence test as often as desired. If you had a restless night, you can always just take a test again tomorrow. You can do this until you reach your optimal score. We believe this method leads to a score that comes closest to your “true intelligence”. Most people reach their optimal score after about three attempts.

People are not native speakers of the language of the test

Most intelligence tests contain questions that test the verbal reasoning capacities of candidates. Although we believe that someone’s verbal reasoning capacity can correlate with their intellectual capacity, we have not included these types of questions in our tests. We believe that verbal reasoning capacities are heavily influenced by different cultural backgrounds, education, social standing, and other aspects which end up creating an uneven playing field.

We have tried to limit the amount of text in the questions and used simple language as much as possible. Of course, people will need a basic understanding of the test language, but not more than that is necessary to complete a test.

People who take the test are not random/average

In order to calculate the individual IQ scores, the results of the intelligence tests are made by comparing all of the test scores of everybody who has ever taken the test on BRGHT. It is reasonable to assume that people who believe they will achieve good results are more likely to take the test than those who believe they will not see good results.

Therefore, it is reasonable to assume that the population we use to calculate the intelligence score is not random. However, we have chosen to not make artificial changes to the score. In order to calculate the percentile score and the intelligence score, an individual “raw score” is compared to all other raw scores. This way, everybody can clearly see what their own scores mean and all scores can be compared to each other.

Comparing your own IQ score to those of other people, whose IQ scores are based on our tests, is very accurate. However, comparing your score to other IQ scores that are based on other tests might be less accurate.