Anderson Smith again. And now,

we're talking about observational techniques that are

used in descriptive methods in psychology.

We know that observational techniques are better than surveys

and self-reports because we can look at the behavior as it actually occurs.

And as we said, it's been used with both animals and humans.

We've already talked about animal observation,

and now we're gonna talk about naturalistic observation with humans.

Now, the advantages of naturalistic observation

are – we can study behaviors that are difficult to see in a laboratory.

For example, what do children do in an actual classroom?

What do they do on the playground?

How's behavior different in crowds than is in a small group, so individuals?

So we can look at behaviors in situations where they exist.

And it's very hard to reproduce an actual classroom that's not contrived in a laboratory.

I'll look at what crowd effects are in a laboratory,

but we can do it when they're actually occurring.

It also maximizes what's called ecological validity, that is,

what I'm measuring can actually generalize to the population.

It is an actual behavior that would occur.

And by studying it in the natural habitat,

it's more likely to generalize than if I'm studying in a laboratory which tends to be

contrived and where subjects tend to do what they think they're expected to do.

Now, the disadvantages are – we can only study what.

Remember, we can't get causal effects by looking at observation,

unless we make manipulations.

If we make manipulations, as we'll see later,

that can produce causal effects but not just descriptive measures themselves.

Another disadvantage is, as we said with the animal observational methods,

if it's an overt observation,

where the observer can actually be detected by the people being observed,

then it's open to reactivity.

The fact that having an observer there can somehow produce reactive effects,

effects that would be different from when the observer wasn't there.

It can also lead to observer bias.

The fact that if the observer has particular beliefs about what the behavior should be,

that can influence what I actually observe.

And also, observer differences can lead to different conclusions.

That is, we might be observing the same thing,

but my conclusion will be different from another observer.

So we have to be careful about

these possible disadvantages when using naturalistic observation.

Now I'm gonna give you two examples of naturalistic observation in humans.

First, beer drinking and college students.

There's a picture of that.

And the second would be attachment styles in children.

That is the bonding that occurs between the caregiver and a very young child.

And that was done through naturalistic observation.

OK. Beer drinking in college students.

This was a study by Scott Geller at Virginia Tech,

and it was done back in the 80s.

And he actually used naturalistic observation to observe beer drinking in

college students to see about what the behaviors are like,

and he found things that were really very different what people expected.

He had 243 students drinking beer at six locations,

and he simply had observations of that.

There were 139 students (29 females,

110 males) at Virginia Tech student center.

And in Virginia, they do sell beer in the student centers there.

I went to University of Virginia where they were selling beer.

And then, 104 students (27 females, 77 males) at five bars,

drinking establishments, around campus.

One or two graduate research assistants would simply go to that place,

whether it's a student center or a bar,

and they would sit at a table and simply observe

a nearby table of people while they were drinking.

And they would measure things like –

by the way the legal age for drinking beer in Virginia is 18

so we're not looking at anything that's illegal –

and they would measure things like gender of the people sitting at the table,

time of arrival, time of departure,

time that they started drinking each container that they had

beer in it (They didn't use alcohol, just beer.),

when they finished that container,

what the container type was,

were they drinking from a cup, a bottle,

or a pitcher, and the number of people at the table.

And all of these were measured and recorded by the observer.

Now, what did they find?

Well, first, the reliability of the observers, the inter-rater reliability,

which is a very important aspect of looking at observation was amazing,

both in the total ounces consumed,

the total time in the bar,

and then the things which were very obvious like gender,

container type, and number of people were all above 98 percent.

So clearly, reliability is not a problem here.

The results? Well first,

males drank more than females.

I think we would expect that.

But interestingly, males left the bar sooner than the females.

Females stayed at the bar longer,

which means that they had to drink slower,

they drank less, and they stayed longer.

So their consumption was in a much slower rate than males.

Students drank more in groups than when alone.

And that's, by the way, something we find very commonly among teenagers

and the sort of early 20s is that because the frontal lobes are the last to develop,

when you're with your peers,

you tend to act a little bit differently than you would if you were not with your peers.

So in groups of peers,

they drank more than if they were alone.

And simply, students drank more from pitchers than they did from cups or from bottles.

This just shows the three kinds of containers and the ounces of beer consumed;

and as you can see if a pitcher is used,

the consumption is much,

much over three times more than you would see with cups, for example.

And it varied by age,

excuse me, by gender.

You look at cups,

and females drank more from cups than bottles but just a slightly amount more.

But with the pitchers where people drank more,

and both males and females drank more from pitchers,

the males drank much more from pitchers.

So this has led to rules like at Virginia Tech and

other colleges that if you have parties that serve beer,

you can't use pitchers.

You've got to use cups to drink from.

The second naturalistic observation study

I want to talk about is attachment styles in children,

and this is a research of Mary Ainsworth who was at the University of Virginia.

And the study was done back in the late 70s called pattern--or book

called "Patterns of Attachment: A Psychological Study of the Strange Situation".

And the strange situation was actually

a natural place where she could observe the behaviors as they occurred,

the attachment between the mother and the young child.

So the strange situation was a room with recording devices and cameras in the room,

a room that had two chairs lots of toys on the floor.

A videotaping took place,

so she could then look at the videotapes to observe their behaviors.

And they looked at the observation of the child.

Here's an example of this room.

The mother would come into the room.

The mother plays with the child,

and this is what happening here.

The mother sits down in one of the chairs.

A stranger enters the room and sits down.

The mother is present.

The stranger attempts to play with the child.

And then, the stranger sits down again,

and the mother leaves the room and then eventually comes back into the room.

And the stranger tries to play with the child both before and after,

with the mother present and with the mother absent.

So we have all of these kinds of situations,

all naturally observed through the videotape.

And from this, she developed styles that the different children

showed in watching this particular scenario through observation.

So based on the observation of the child,

she classified children in one of four attachment styles, types of bonding.

The first was secure,

and a vast majority of the children had secure attachment with the caregiver,

mostly the mother, but caregiver.

And the secure attachment style means when the mother came back into the room,

the caregiver was actually acknowledged,

and the child was relieved when the mother came back into the room.

Nineteen percent showed an avoidant attachment style

– just doesn't seem to acknowledge the caregiver at all,

doesn't matter whether the mother's present or absent,

the child does what the child wants to do and doesn't show

necessarily a response when the mother leaves and comes back,

where the secure child shows distress when the mother leaves and shows relief

when the mother comes back or the caregiver comes back.

The third is the anxious/ambivalent attachment style

where the infant is distressed when the caregiver leaves, but when she returns,

the infant rebuffs her – is sort of mad with her,

doesn't show the sort of secure attachment that would

show in the 64 percent when the mother returned to the room.

And in 3 percent, she couldn't really tell.

It was just totally disorganized;

there wasn't any kind of way in which

the child reacted that I could categorize in that case.

So secure, avoidant, anxious/ambivalent, and disorganized.

And she could classify all the children into these particular categories.

Now, she also showed through some international research.

Here's the American style, mostly secure,

then you see the avoidant and then smaller amount of anxious attachment.

But if you look at Germany and you look at Japan,

you get two very different relationships of

the avoidant to the anxious/ambivalent bonding styles.

In every case, secure was at much higher percentage,

and in fact in Japan,

even a higher percentage,

but there was this difference in

the other two categories of attachment style in children.

So we've seen two different observation, naturalistic observation,

using, in this case,

children, one where we look at attachment style.

And then, a little bit older children,

we look at drinking while they're in college.

Again, using observation to draw conclusions about behavior when it actually occurs.

So we can use observations to get accurate objective descriptions of behaviors.

But remember, the observations should not be obtrusive,

and the observer should do everything they can to avoid

observer bias and to avoid possible reactivity. Thank you.

Lesson 2C – Naturalistic Observation (Humans)

Descriptive Research Methods in Psychology

Meet the Instructors