Background to my interest in teaching children with autism

[Now, I am going to try to simplify a bit of what Doug has written, because I found the original just a bit too complicated for my little brain. RR]. According to Wikipedia, “In statistical hypothesis testing, a type I error is the incorrect rejection of a true null hypothesis (a ‘false positive’), while a type II error is the failure to reject a false null hypothesis (a ‘false’ negative’).” Googling Type 1 and Type 2 errors, we also find the following Khan Academy explanation: “The easiest way to think about Type 1 and Type 2 errors is in relation to medical tests. A type 1 error is where the person doesn’t have the disease, but the test says they do (false positive). A type 2 error is where the person has the disease but the test doesn’t pick it up (false negative).” That is, in clinical [psychological/psychiatric] practice, identifying/labelling ‘normals’ as ‘abnormal is what is known as a Type 1 error; labelling “abnormals’ as ‘normal’ is known as a Type 2 error.
In clinical work, both Type I and Type II errors are risky. If I make a Type I error, I will have increased the cost of treatment greatly, by treating more people who don’t need treatment than I treat of those who do. If I take the risk of making a Type II error, I miss treating many of those who need it – a problem which is made greater if the problem results in pain or loss not only to the person but also to others – as where the problem is a criminal one which risks victimization of others.
The best way to reduce the risks of making either Type I or Type II errors is to employ more than one ‘screen’ or more than one type of measure of the dimension in question through which people are processed prior to making an identification decision about an individual’s differences.
Many psychologists seek to accomplish these multiple ‘screens’ by using more than one test providing information about each dimension in a battery of tests. This is probably the most practical solution, but it still runs the risk that errors of measurement are, for example, at least partially related to the day or time when the assessment is undertaken – so that the errors affect all the measures taken at that time. Because of this (also relatively rare) risk, it is slightly better to use ‘successive screens’ – that is, observations or measures taken at different times, and preferably based on different types of data so that errors of measurement can at least be recognized.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s