Test administration, measurements and scoring - Chapter 6 - Verhoeven - 2014

This summary of Test administration, measurements and scoring - Chapter 6 - Verhoeven - 2014 is written in 2020

Psychodiagnostic has to be seen as more than testing, making sure the client does the tasks right, tallying the scores and filling out the report.

Where do you need to think about during testing day?

On a testing day the test administrator has to be sure the candidate is aware of everything that is happening and that there are no unrealistic expectations or unlogical ideas. The administrator must also think about the client's feelings and perceptions. There are a lot of practical considerations the psychologist/test administrator has to think about, such as:

• Ensure that all testing materials are in order and that the instructions are very clear.
• You have to think about how to take the tests, because multiple testing hours or days can be very exhausting for candidates, and this can influence the results.
• The test administrator has to have a log book to be able to note down anything weird or noteworthy.

Once the tests are all in order you have to think about all the important norm groups you want to compare the client's raw scores with. Norm data is important to make good-quality decisions. Researchers use norm groups to make a cut-off score, and you can think about race, gender, age, work etc. when making a norm group.

How do you score the data of the test?

Tests initially give out raw scores that don't say much about anything on their own. You have to compare the raw score with the norm data. But when you compare one raw score with the norm group, you still don't know much. Because this is a single measurement which means that you don't know what will happen when you do the whole measurement another time. In psychology, it is often not possible to let the same person do the test a 100 times, that is why psychologist compare the score of an individual with the data collected from the entire population of people. The population data can give information about the reliability of the score of the individual, how greater the standard is, how wider the range within which the candidate could easily have scored. 

When you have the data of the individual and the data of the whole population of people who have done the same test you can determine the range in which the candidate's true score must lie. In practice, researchers often use a 90-95% confidence interval to look for the range of true score. 

What are standard scores?

The fundamental idea of psychology is that everything can be measured, meaning that all vague, not touchable concepts within or about a person can be measured, be quantified and observable. The thing that separates psychology from other sciences is that all the psychological topics must be looked at and researched independent of each other, which means that each variable has a measurement unit of itself. 

But this also means that in some situations within psychology you want to compare different measurement units and variables with each other, what means that you want to compare apples to oranges (what is quite difficult to do).

A very important standard score researchers often use is the z-score. The z-score always takes on the value of 0.0 with every variable. If you know the mean and the standard deviation of two total different variables you will be able with the z-score to make a useful comparison. An example: when the mean IQ is 100, and the standard deviation is 15, you can convert this to z-scores, what means that 100 (the mean) becomes 0.0 and the 15 standard deviation IQ points translates to 1.0.