Validity

I'm Dr African.

Sorry, and I'm a cognitive neuropsychologist at Goldsmiths University of London.

What that means is that I try to study how our mental functions such as memory and

face recognition work by studying people who've got brain

damage and you've got disruption in those systems.

So, for example,

I study people with memory disorders to understand how our memory works.

And I study people with a rare disorder called presto magnesia,

whereby they can't recognise faces

to try to understand how you and I recognise faces.

Now what we're going to do is to look at the issue of validity.

No validity, as it says on the tin,

is to do with how appropriate or realistic or valid your results are now.

Ultimately, when we do any form of experimentation,

we're trying to explore a particular issue such as, um,

is bullying brought on by watching violent video games or something like that?

There's a question there

and our experiment.

However we work it out or design, it is going to be to answer this particular question

Now. The issue of validity is, are we actually doing that correctly?

Our our results going to be valid and do they represent

the type of thing that we want to look at?

So the first type of validity is face validity.

Face validity is effectively what it says on its face or it's tin.

Is that what it's looking at?

So if we do a study on, say, the influence of violent video games on bullying,

the results that we get are they actually representing that particular issue?

Or have we ended up doing an odd study where, in fact, were, um,

looking at poverty or the impact of poverty on people's behaviour?

So face validity is basically about Does it actually

measure the thing that it's saying it does?

So if I'm doing a study on memory,

is the type of test that I'm doing to do with memory,

or am I actually somehow measuring people's I Q.

By mistake.

So the point here is that for face validity,

it needs to measure what it says it's Measuring.

Another form of validity is concurrent validity.

So how does this particular study or test

relate to other tests that currently measured the same thing?

So using the memory example again,

if I was developing a new test of short term memory.

For example,

if it's got concurrent validity,

it means that the results from my study relate

to other concurrent tests of short term memory.

So what you want to do is to make sure

that you are measuring what you're say you're measuring,

which is the face validity,

and that it relates at some level to the other

tests in the similar in the same or similar field.

So you want to have rough equivalence so

that you're not measuring something completely different.

Another really important type of validity is known as ecological validity

and ecological validity is something that's become

really quite important in psychology because ultimately,

what ecological validity is about?

Is that what we're measuring as experimenters in

a laboratory or in a classroom setting?

Does it actually relate to the outside world?

Now, for example, I'm as blind as a bat

my eyes that is so bad that if I didn't have my contact lenses on,

I wouldn't be able to see anything.

So when I go to an optician,

optician gets me to look at a chart of letters on green or red

background and asked me whether it's clearer

on the red background or green background.

Now I don't go around the outside world

looking at things on red and green background.

But importantly,

the diagnosis that they give the result of my eye test

and what they help me with in terms of

glasses or contact lenses has got ecological validity,

because with my contact lenses right now, I can see things quite clearly.

So there is ecological validity to the type of testing that the opposition is doing

in terms of my real world behaviour.

With some aspects of psychology, there is good ecological validity.

So, for example, some tests of memory,

short term and long term generally they do

measure what's happening in the outside world.

For example, there are other areas,

such as how the front of the brain measure the front of the brain,

controls human behaviour and allows us to plan, make decisions, be creative

that really lack ecological validity.

So, for example,

there are famous patients who've had tumours removed from the front

of their brains who have got major problems in everyday life.

And some of you may have come across a

case called Phineas Gauge who had a bizarre accident in

the States in the 1800 where a poll flew through

his face and caused a lot of brain damage,

and he ended up with frontal lobe brain damage.

Now we have patients like Phineas Gauge who have damaged in the same area of

the brain caused by brain surgery who can't do things in the outside world,

such as make decisions about where to go this afternoon or financially,

what to choose between five different options.

And yet they pass all of the standard test of frontal lobe functioning.

So what we have there is a lack of ecological validity.

We know that the patient in the outside world has got major problems,

but the test currently are saying that there's nothing wrong with them,

and that's a lack of ecological validity.

So overall validity is about the fact that we want our results

to bear some semblance

to what is actually happening in the outside world.

Otherwise, there's no point in doing this research,

and there are all sorts of consequences of

not having a valid result because, for example,

you might end up with a theory that's completely wrong or

were still in clinical cases like the patients with brain damage.

You might end up saying to them,

There's nothing wrong with you even though

there's something really wrong with them.

Civility is an extremely important thing,

and we have to make sure that our studies are experimentation

and our results have good validity within them.

The next thing we're going to look at is reliability.