Greetings.

I will continue to study the Psychodiagnostics and Psychological Assessment.

Today we are completing the second theme.

We will talk about the rest psychometric characteristics

of the psychodiagnostics tests which describe them evaluation.

There are representativeness and norm.

So, what is the psychological test representativeness?

This rule was borrowed from mathematics.

It is associated with such concepts as the sample and the general population.

The sample is those responders who are tested during the development

and practical approval of the test.

Mathematical calculation were used to evaluate

how create some characteristics were measured.

The general population is all potential testees

to whom this method can be applied in principle.

And the results remain reliable.

Well, representativeness specifically shows

that this test can be applied to this type of testees.

From a mathematical point of view,

this means that there is a normal distribution for each of the scales.

If a normal distribution for each of scales isn't possible in principle

the condition of having the presentation is considered the stability of the distribution.

And I want to remind you that the normal distribution is depicted by Gauss' curve.

Its psychological meaning lies in the fact that on the x axis we note the value

of the scale and on the y axis the number of people who have this value.

This is an average value of the scale at the highest point of Gauss' curve.

And there are the areas of low and high values at the edge of a curve.

The minority of testees have this type of value.

Thus, we can see that the representation allows us to classify the test results

and to allocate high, average, and low levels at upper scale.

And we've came to review the latest psychometric characteristics.

It is the norms.

As I already mentioned, the norm shows borders of average values.

Availability of norms is mandatory for standardized tests

which have scale with a clear quantitative measurement.

In this case, the interpretation of particular testees results is based on

the correspondence of results and norms borders precisely.

What norms are there?

Broadly, there are social-cultural and statistical norms.

A social-cultural norm is a certain level of a psychological characteristic severity

which is considered necessary or appropriate in society.

A social cultural norm is an average intensity of a scale in most cases.

For example, having a very low level of anxiety is not good.

Such kind of people is usually referred to as 'thick skinned', and 'bored people'.

It is also really uncomfortable if you have a high level of anxiety.

It provokes a permanent distress

and a lot of trouble was in the life of the man and his environment.

But to have an average level of anxiety intensity, it is just right.

A statistical norm is an average range of values measured on a scale

of a psychological characteristic.

In this case, the norm is approximately of a values indicator

to a level that characterize the average person.

Moreover, the method of quantiles is used to calculate statistical norms.

You can remember the chart quantum from the course of physics.

It is a bit or a little element.

Accordingly, in psychology,

a method of quantiles means a separation of a scale into parts.

Each separated part is characterized by an equal expiration

of a measured property or characteristic.

There are several types of quantiles.

Percentile Ranks indicates a value below which a given percentage of observations,

or testees, in a group of observations, false.

For example, the 20s percentile is the value, or score,

below which 20 percent of observations might be found.

The second most widespread form of quantiles is a quartile.

In this case, a scale is separated into four parts.

The first part means a low level of values.

The second and the third parts, an average level of values.

The fourth part, high level of values.

And the last type of quantile is a decile.

By its name, it is clear that a scale is separated into 10 parts.

But, in psychodiagnostics, this type of scale

is the most popular under the name of 'sten score'.

Psychologists are begining to use a sten score

for tests with a large number of questions and scales.

Questions in the scales more often are distributed unevenly.

There are various test items in various scales.

As a result, there is a problem of comparing values at different scales.

For example, we have two scales in a test, and our testee has 15 on each scale.

But, you know that there are only 20 items in the first scale

and 40 items in the second scale.

A sten score was created exactly for such cases.

It is always measured from one through 10.

The values one, two, three, are considered low level of values; from four to seven,

an average level of values; and eight, nine,10, a high level of values.

First of all, we could collect the usual scale values.

Sometimes, we call them 'raw values' in a psychodiagnostics jargon,

then we use a special table for transforming these raw values to a sten score.

There are always 11 columns in this table.

There are name of scales in the first column.

All the other columns consist of raw values,

intervals which correspond to one or another sten value.

Let's go back to our example and try to transform all 15 points off of our scale to sten.

We can see that on the fifth scale 15 points are located

in the column eight in the interval from 15 to 16.

And, on the second scale,

15 points are located in column four in the interval from 13 to 16.

So, we can be sure in our interpretation after using a sten scale.

Now we understand that we have a high level of values in the first scale,

and an average on the board of lower level on the second scale.

Okay.

Now we end up to start is the most difficult module of our course.

Good luck and see you in the third module.

It will be dedicated to psychodiagnostics of intelligence.