Course Outline | Intro to Psych home | EnvironmentalET home |

Psychology is a Science

Psychology is considered to be an empirical discipline. It is held to be based on objective observation. A central belief in psychology is that it is public knowledge--psychological truths are not based on secret insights, personal prejudices, or mystic experiences.

The Scientific Method

The scientific method is based on certain basic assumptions:

- What we perceive with our five senses represents objective reality.
- Objective reality functions according to certain natural laws that are observable, describable, and persistent.
- Every effect has a cause; and each event can cause others (everything happens for an observable reason).
- Through observation, manipulation and reason we can decipher the basic principles of the physical world.

Note that these are assumptions--they cannot be proved. Can they be disproved? Our everyday commonsense understanding of life follows these assumptions fairly closely.

The scientific method is sometimes summarized as:

question: "Why are some people able to quit smoking while others fail?"

hypothesis: "The age at onset of smoking is negatively correlated with success in quitting"

test (experiment): Administer a survey to past and present smokers--"At what age did you start smoking?" Note: This is NOT an experiment

theory: Combine the results of many specific studies into a general conclusion

This summary misses the complex nature of each step and misses some of the fundamental dynamic character of science. Science relies on:

Observation:

based on the 5 senses

done carefully to eliminate bias

should be reproducible (repeated observations are called "facts")

strictly speaking cannot establish causal relationships

Experimentation:

essential to establishing causation

manipulation of one factor at a time

hypothesis = an educated guess about an observation __that can be
tested__

makes use of deductive reasoning (drawing logical conclusions from a set of premises)

Theories:

a series of repeatedly supported hypotheses can be combined into a theory

based on inductive reasoning (forming a general principle from specific observations)

subject to experimental refutation

Natural Laws and Explanatory Concepts:

allow prediction (sometimes quite precise though not usually in psychology)

consistent through space in time

should be consistent with each other and with observation

The scientific method is cyclical:

How do we evaluate science?

It helps to ask the following:

- What are the actual observations underlying the conclusion?
- Can the observations be verified and repeated?
- Does the conclusion follow logically from the observations?
- Does the conclusion account for all of the observations?
- Is the conclusion supported by the members of the scientific community with the greatest competence to judge the work?
- Is the conclusion the simplest one that will account for the observations? (Occam's Razor)

Research Methods

General Considerations

Sampling

population = the entire group whose characteristics are being studied

It is not usually feasible to study the population as a whole

sample = a subset of the population

representative sample = a sample whose characteristics match those of the population

statistical analysis = procedures for extrapolating from observations on samples to conclusions about populations

based on probability (and algebra)

significant = unlikely to have occurred by chance

Relevant factors are called variables.

A relationship between variables is called a correlation.

positive = variable A increases along with variable B

negative = variable A decreases as B increases

correlation coefficient = a numerical indication of the degree of correlation (varies from -1 to +1)

Bias is a systematic source of error.

A central challenge in science, especially psychology is operationalization: How do we state our hypotheses in objectively verifiable form?

variables must be objectively, concretely, and usually quantitatively defined

characteristics must be reliably measured

Research Setting

field setting

more natural

easier to be invisible

difficult to control

laboratory setting

easier to control

more artificial

more intrusive

Measures of Behavior

self-report

cheap and easy

may not be true

questionnaire ("pencil and paper" instruments)

structured interview

direct observation

participant observation

unobtrusive measures

Non-Experimental Methods

These do not involve the manipulation or control of any factor, and thus cannot be used to determine causal relationships.

Sometimes called descriptive or correlational.

archival research

a review of existing records

cheap

limited to available data

retrospective studies

an examination of the history of a group of subjects (self-reported or directly observed)

longitudinal studies

current characteristics of a set of groups are examined; the group is then followed over time to see if the outcome is different for the different groups

surveys

fixed-alternative

open-ended

naturalistic observation

non-intrusive (like the entomologist watching butterflies)

may be difficult to actually stay out of sight

case studies

intensive examination and description of a single individual

common and probably helpful in clinical settings

may be dramatic and interesting

may not be representative

may not be readily generalizable

Experimental Methods

These involve systematic manipulation or control of one (or more) factors, called independent variables.

This prior manipulation allows us to draw conclusions about causation.

If A tends to occur with B, there are three possible reasons:

A causes B

B caused A

C causes both A and B

Only an experiment will allow us to distinguish between these three.

experimentation

In an experiment, we manipulate the variable of interest and control (keep the same) other variables.

In the simplest experiment we have a treatment group and a control group.

In the purest experiments, we change only a single variable.

If multiple variables are changed, we need to use statistical techniques to separate the effects of each.

If extraneous variables change from one test to the next, they may confound the results.

If there is no standardization of conditions we have an uncontrolled experiment.

Statistics

Samples vs populations

We usually cannot study the entire population (ie, all people).

There are usually differences between different types of subjects

We try to pick a sample (a smaller group) of subjects that represent the population: a "representative sample."

Random assignment

Matched controls

To extend our findings with samples back to the population as a whole we use statistics.

A set of techniques that use mathematics and probability theory to draw conclusions from observations

Descriptive statistics

These summarize observations

mean = average

median = the middle value

mode = the most common value

standard deviation = a number that expresses how spread apart a set of numbers is

95% of values fall within two standard deviations of the mean

correlation coefficient = a number that tells how much two variable go together

varies from -1 to +1

+1 = perfect positive correlation

-1 = perfect negative correlation

0 = no correlation

The "best fit" or "least squares" line (slope and intercept) can also be calculated.

4 3 5 4 6 5 6 6 6 2 |

mean: 4.7 median: 5 mode: 6 sd: 1.4 |

Inferential statistics

These allow us to make inferences beyond the observations

For example, suppose that within a particular group of 100 children, girls scored higher than boys on a particular test. Does that mean that it is likely that another group of girls will also score higher than another group of boys?

Hypothesis testing

From a given set of numbers (eg, scores) we can determine the probability that a certain distribution occurred by chance.

If the probability is less than some value, say 5% or 1%, than we say that the finding is statistically significant.

Examples of inferential statistics include t-value, F-value and the correlation coefficient (r)

Milgram's Famous Obedience Experiment (top)

Milgram wanted to answer the question "Can someone be coerced by an authority figure to hurt someone else?"

The Set Up:

On the face of it, the experiment looked at learning and had three participants: an experimenter, a teacher and a learner. The teacher and the learner were supposed to be experimental subjects who drew straws to determine roles. In fact, the drawing was rigged so that the one subject was always the "teacher." The learner was always the same person (a 47 year old accountant who had been thoroughly coached on his performance).

The learner was strapped into an electric chair. Both the teacher and the learner were told "Although the shocks can be extremely painful, they cause no permanent tissue damage."

The Lesson:

The teacher sat in a different room where he could hear but not see the learner. The teacher gave the learner a simple word association test. When the learner answered wrong, the teacher was supposed to administer a shock.

The Shock Generator:

Milgram made a very convincing looking gizmo (see picture on p 627 in Feldman) with labeled switches for giving shocks at levels from 15 to 450 volts.

The teacher was given a sample shock using the switch marked 45 volts.

The switches were grouped from "Slight Shock" to "Danger: Severe Shock" (the last two were listed with "XXX").

When the switch was flipped, a red light came on above the switch, there was a click and a buzzing noise.

The Experiment:

Each time the learner made a mistake, the teacher was told to move to the next voltage level. The teacher was supposed to announce the voltage level to the learner then flip the switch.

At various voltage levels the learner expressed discomfort then pain then agony. At 120V, he said "Hey this really hurts; at 270V he screamed and shouted "let me out of here." At 300 V he refused to answer anymore (which of course meant he got the question wrong). After 330V he stopped responding altogether (after complaining about his heart since 150V).

If the teacher did not want to continue, the experimenter said "The experiment requires that you continue" or "You have no choice, you must continue."

The experiment ended when the teacher refused to give any more shocks or 450V had been reached. Once the experiment was over, the set up was explained and the teacher and learner shook hands.

The Results:

26 out of 40 subjects went all the way to 450V. No one stopped before 300V.

The subjects in Milgram's experiment were very distressed (even though they continued) and were greatly relieved when they heard about the hoax.

Milgram's experiment has been confirmed many times over.

Factors that Influenced Obedience in Milgram's Experiment:

- Expectation
- Context
- Gradual escalation
- Displaced responsibility
- Separation from the learner

Factors that reduced obedience:

Milgram tried a number of variations, some of which reduced the likelihood that the teacher would obey.

Return to: | Top | Course Outline | Intro to Psych | EnvironmentalET |

ANTHONY G BENOIT

ROOM 201B

(860) 885-2386