Intelligence / IQ Tests In Psychology
Intelligence Tests
One of the problems in using mental age to measure intelligence is that as a person’s chronological age (CA) increases, so does their mental age (MA).
This makes it of little value in comparing intelligence between different types of people, because it is always rising and is therefore unstable.
The IQ Test
In an attempt to create a more useful and stable way to measure intelligence, the German psychologist William Stern created a ratio based on the comparison of mental age and chronological age. He called this the intelligence quotient, or IQ.
Note : It is called a “quotient” because the resulting score is created by division.
Calculating IQ Test Score
To calculate a persons IQ he used the following formula.
IQ = (MA / CA) x 100
For example, if a child has a CA of 10 and a MA of 10, then we can calculate their IQ with the following calculation :
10 / 10 = 1
1 x 100 = 100
IQ = 100
A score of 100 on an IQ test is regarded as an average level of IQ.
If another child is tested who has a CA of 8 and a MA of 10, then we can calculate their IQ with the following calculation :
10 / 8 = 1.25
1.25 x 100 = 125
IQ = 125
In this IQ Test score the child is said to have an above average level of intelligence.
If another child is tested who has a CA of 9 and a MA of 8, we can calculate their IQ with the following calculation :
8 / 9 = 0.89
0.89 x 100 = 89
IQ = 89
In this this IQ Test score the child would be said to have a below average level of intelligence.
IQ Test Score Distribution
IQ is a random variable, which means that IQ test scores will vary in a population according to the laws of chance.
The result of this is that IQ test scores taken from a random selection of individuals will tend to create a bell-shaped (aka Gaussian or Normal) curve.
At one end you will have people with the lowest IQ test scores. In the middle, the people with an average IQ test score, and at the other end the people with the highest IQ test scores.
As you can see from the distribution of IQ test scores most scores are clustered in the middle, which means most people have an average IQ. Fewer people will have lower IQ test scores, and fewer people will have higher IQ test scores.
If we apply this to a population, we can generalise by saying that on average around 2% have an IQ lower than 69, 50% have an average IQ (90-109), and 2% have an IQ of 130 or above.
Regardless of what method is used to measure intelligence, an intelligence test must be both valid and reliable.
Valid Intelligence Tests
A valid test is a test which measures what it is suppose to measure. So a valid intelligence test would measure intelligence and not something else like your personality.
However in order to determine the validity of a test, it is necessary to compare the scores obtained by that test with an outside criterion.
Outside Criterion
An outside criterion is a measurement instrument which is independent of the intelligence test. One example of an outside criterion is a grading marking scheme or grade point average.
If an intelligence test is valid, then we would expect people with higher intelligence test scores to obtain higher grades, and people with lower intelligence test scores to obtain lower grades.
Correlation Coefficient
The relationship between the test score and the outside criterion is determined with a correlation coefficient, which measures how closely two variables are related.
Therefore the higher the correlation coefficient, the closer the relationship will be between intelligence scores and grades and the more reliable the intelligence test is said to be.
Reliable Intelligence Test
A reliable test is a test that gives the same or similar results when that test is repeated. This is again measured using the correlation coefficient, which determines how reliable that test is.
Types Of Tests
When measuring the validity and reliability of a test, we can come to four different conclusions as to the type of test it is.
1 – Neither Valid Nor Reliable
The test is not valid (i.e. it tests something it is not suppose to), and the test is not reliable (i.e. does not give repeatable results).
2 – Valid But Not Reliable
The test tests what it is suppose to test, but does not give repeatable results.
3 – Reliable But Not Valid
The test gives repeatable results but does not test what it is suppose to test for.
4 – Valid And Reliable
The test tests what it is suppose to test for and gives repeatable results.
The Stanford-Binet Intelligence Scale and the Wechsler Scales are considered to be valid and reliable measures of intelligence.
Summary Of IQ Testing
• IQ scores are calculated by dividing mental age by chorological age.
• IQ scores allow intelligence scores to be compared between different age groups of people.
• 100 is an average IQ score.
• IQ is a random variable, and results in a bell shaped curve when a population is sampled at random.
• Most people have an average level of intelligence, with fewer people having below or above average intelligence.
• A valid test tests what it is suppose to test.
• A reliable test gives repeatable results over time.
• How reliable or valid a test is can be determined using a correlation coefficient, which determines how closely two variables are related to each other.
