Hello, Anderson Smith, again.

And now we're gonna talk about observation studies.

Observation studies are better than other descriptive measures.

Like for example with surveys and polls,

we have to rely on self-report,

where the person's actually telling us about things that happened in their past

or their estimate of something about them, themselves.

While with observation studies,

we can actually see what is going on in the natural habitat where the behavior occurs,

because we are observing the behavior actually as it actually occurs.

And it has been used both with humans and with animals.

Very commonly used with animals because we want to see what the behavior is,

not in cages in some kind of laboratory,

but out there where the behavior of the animal is actually occurring.

Now we're gonna talk about,

in this section of the course,

three different kinds of observational methods -- observational method with animals,

observational methods with humans,

and then something called participant observation,

which is a little more complex,

not used as much in psychology as other social sciences,

but all three are observational methods.

Now with naturalistic observation with animals,

there are two very well-known examples of psychologists who've studied

animals in their natural habitat to understand the behavior of those organisms.

There's Jane Goodall who studied chimps in Tanzania.

And here's a picture of Jane with one of her animals,

even though this is obviously a representation of the animal,

not a real one, the chimpanzee.

And then there's Dian Fossey who studied the animal gorillas,

the mountain gorillas in Rwanda.

Both psychologists had studied the animals in their natural habitats.

Here's a picture of Dian Fossey.

And here's a picture of her grave.

She was actually murdered in her cottage

located very close to the habitat where the gorillas live.

We still don't know what that murder was caused by,

but she's no longer with us.

But both Dian Fossey and Jane Goodall did

a lot to help us understand these animals that are very close to us,

in an evolutionary sense.

For example, it used to be said before Jane Goodall's research,

that humans were the only animal that actually used tools.

We developed tools to help us in our behaviors.

But she showed that chimpanzees use tools all the time.

In one case, she showed that they would take a branch,

pull off all the limbs,

make a sharp stick,

and they'd dig down with that stick into

the termite mounds to dig out the termites and eat them off the stick,

the first popsicle, I guess,

but showed that they do in fact use tools in lots of ways to help them do things.

So, better understand the chimpanzee behavior and with Dian Fossey,

better understand gorilla behaviors by observing the behavior when it actually occurs.

Now, it's a descriptive study.

There's an observer watching that but the observer needs to be

very careful in making these observations.

First, they have to minimize the impact of being present.

Just the fact that they are there could cause the chimps or

the gorillas to do things differently they would do if she's not there.

So first, she has to develop this rapport so they sort of habituate to her,

adapt to her, and would engage in the behaviors without

really her interfering with those behaviors.

She also has to not change the behavior in

any way by being there or change the environment in any way by being there.

She wants the environment to be natural,

and she wants the behavior of the organisms in that environment to be natural.

Now, the researcher, the observer has to also be aware of

the possibility of reactivity on the part of the animals being observed.

That is, they react and behave differently

than would be expected if the observer was not there.

Now you'll see with humans,

that's a real problem because if I'm observing someone doing something,

they might behave very differently trying to please me as an observer.

And that could also be true of chimps and gorillas.

They just don't behave in the natural way because the observer is there.

So the observer has to be aware of these effects of having

impact on the actual behavior and the environment in which

the behavior occurs and try to minimize that.

They have to also be aware of observer bias.

They might bring into the observation

certain biases about what the animals are supposed to be doing,

and that will somehow cloud their recording of what the behaviors are.

Reactivity, observer bias, two problems with observational techniques.

Now, reactivity has even been shown in cockroaches.

That Zajonc showed that cockroaches

behave differently if other cockroaches are around than if they're alone,

by themselves as a cockroach.

So again, just having other organisms around, especially an observer,

which is a very different kind of organism from the normal grouping,

can have effects on what the behavior you actually see.

Now let me give you examples of that.

Here's an example of Bob Rosenthal and Fode done in the 60s,

and they used bright and dull rats.

Two groups of rats ran in a single maze.

Researcher was told that rats were bred to be fast

in a maze or told the rats were bred to be slow in a maze,

even though they were the same rats who were randomly selected to be in the two groups.

Interesting enough though, just an observer thinking that

the rats were smart or the rats were dull,

the two groups of rats differed in their maze learning,

again showing somehow experimenter bias.

We don't really know what the experimenter did to create that bias,

but somehow the bias was there and the animals ran the maze differently.

The well-known example is Clever Hans.

In the early part of the 20th century,

a horse named Hans was trained by William von Osten to do simple mathematic tricks.

He trained the animal for over four years,

and he got Hans to appear that he was at about

the fifth grade level in doing mathematical problems.

He would answer by tapping his hoof.

He would say, "What is two plus two, Hans?"

And Hans would go one, two, three, four.

He wouldn't say that, he would tap his foot, his hoof.

And everybody was amazed by Hans' mathematical abilities.

Many scientists tested Hans.

And here's a picture of Hans,

you see the horse,

and they're doing a math problem.

And many scientists tested Hans,

and the scientist would say,

"What is three times two?"

And he would tap his foot six times.

So it wasn't just Mr. Osten that was creating the problem,

it was anyone could do this.

Clever Hans.

And then a psychologist named Oskar Pfungst in 1911 showed that Hans could not solve

mathematical problems by having the scientist whisper one

number and then having another scientist whisper the other number.

So one scientist would go up to the horse and say, "Five."

And the other would go up and say, "Plus two."

And Hans would not be able to tap his foot seven times.

Hans wouldn't know what to do.

He couldn't perform.

So Hans was clever, but he wasn't clever in mathematics.

He was clever in detecting subtle nonverbal cues that

the person giving the numbers did while Hans was tapping his hoof.

The problem was solved.

It was shown that he didn't have math problem by using what's

a very common design, an experimental technique,

it's called a blind design,

where neither observer knew exactly what the problem was.

One observer was simply set to say "Five" and the other was to say

"Plus two" and neither one knew that seven was the correct answer or not,

and Hans didn't either.

So again, examples of reactivity,

that there could be ways in which the behavior of the animal is

changed simply by something that the observer does in

the environment or something that other organisms do in the environment to

help with that particular behavior being observed.

So the key here is coding of behaviors.

As I'm measuring something,

I'm observing something, I have to have some code of what it is on being observed.

And what we tend to use is multiple observers.

So then we can look at the degree to which the observers

agree that the behavior is what we say it is,

and that's called interrater reliability so reliability is the important construct here.

And then, also construct validity.

Are we measuring the same thing?

When we say we're measuring one thing,

do we agree on what we're measuring,

and do we agree on what's being measured?

What is being measured is construct validity,

and whether we're measuring the same thing is interrater reliability.

Thank you.

Lesson 2B: Naturalistic Observation (Animals)

Descriptive Research Methods in Psychology

与讲师见面