PSYC 102 - Research Methods

What is a hypothesis?

A testable prediction about the relationship between variables;

ex. "If people have higher levels of fitness/activity, they will become sick less frequently than people with lower levels of fitness/activity"

What is a population?

Everyone we are interested in studying;

i.e., students, senior citizens, women, men, etc.

What is a sample?

The subgroup we actually study

What is a subject(s)

The people or animals in the sample that we are studying

Advantages/Disadvantages of a random sample

Advantage

- Usually representative of the population

Disadvantage

- Inconvenient

- Expensive

Advantages/Disadvantages of an availability sample

Advantage

- Convenient

- Inexpensive

Disadvantage

- Not representative of population

What are some bad sampling techniques?

Self-selected samples and anecdotal evidence

What is the problem with self-selected samples?

- Subjects that respond often have strong opinions

- NEVER representative

What is naturalistic observation

The careful observation of subjects without direct intervention/interference

Advantages/disadvantages of naturalistic observation

Advantage

- Most natural environment and behavior

Disadvantage

- Reactivity; observation changes behavior

- Observer bias; see what you expect to see

- Can't determine motivation behind actions

Advantages/disadvantages of surveys

Advantage

- Can get a large sample

- Inexpensive/easy

Disadvantage

- wording of questions can change answers

- Social desirability bias; people give socially appropriate answers

- Hard to get a representative sample

What is a case study?

An in-depth investigation of an individual subject.

Advantages/disadvantages of case studies

Advantage

- Get lots of detailed information on "typical" or unique case

Disadvantage

- Not representative

What is a correlational study

Investigation of the relationship between two variables;

As A increases, what happens to B? Does it increase, decrease, or stay the same?

What is a positive correlation

As A increases, B increases

Strong positive; R = 1

What is a negative correlation

As A increases, B decreases

Strong negative; R = -1

What is the R score if there is no correlation (no relationship between A and B)

R = 0

What does it mean when someone says "correlation is NOT causation" ?

While a correlational study can determine that there is an ASSOCIATION between A and B, it cannot:

- Say if A causes B or if B causes A

- Rule out a third variable that could be causing the change in both variables. This is known as the Third Variable Problem

Advantage of correlational studies

Sometimes an experiment just isn't ethical

What is an experiment?

The investigator manipulates a variable under carefully controlled conditions and measures the change in other variables

Advantages/disadvantages of experiments

Advantage

- can establish casual relationship (A causes B)

Disadvantage

- Artificial and unrealistic

- ethical issues

independent variable

variable manipulated by the researcher

dependent variable

variable measured by researcher

extraneous variable

Anything other than the independent variable that could change the dependent variable

Experimental Group

Subjects who receive treatment to be studied

Control Group

Subjects that are treated EXACTLY like the experimental group EXCEPT for the independent variable (treatment)

Random assignment vs. matched assignment (remember the card activity in class)

Random:

throw them into random groups-- hope its equivalent

Matched:

Figure out their baselines with pretests, then make pairs of them that are similar; one goes in control one goes into experimental

What are repeated measures?

One group of participants;

Same group tested under all conditions necessary for the study

This removes the need for random/matched assignment

demand characteristics

Subjects behave the way they THINK they are expected to

Placebo effects

Subjects experience effect because they expect to;

NOTE: effects are still real

Experimenter Bias

Experimenters find an effect because they expect to

Double Blind Design

Neither the researcher or the participant knows if the treatment or control is given; this controls for placebo effects and experimenter biases

The five ethical guidelines

- Informed consent

- Deception is not used (in most cases see other question for exceptions)

- Subject is not harmed (in most cases, see other question for exceptions)

- Subject is debriefed after study

- Confidentiality is maintained

How do we treat the control group in an experiment?

The EXACT same way EXCEPT for the introduction of the 'treatment' in the study;

Ex. Two groups are given identical pills under the same circumstances, except for the fact that one pill is a sugar pill/placebo.