Psychology as a Science

In this discussion we will explore:

limits of common sense science - a method for understanding methods of science
description correlation experimentation

evaluating data with statistics sources of error and bias in research

How do we know things?

We Just Know It Intuition or Common Sense OR Tried and True Science

Can we trust our intuition?

Science vs. Common Sense

Why dont we rely on intuition (common sense)?

1. Hindsight Bias tendency to believe, after learning an outcome, that one would have foreseen it.

To find out something has happened, makes it inevitable I Knew It All Along Phenomenon

Why dont we rely on intuition (common sense)?

2. Overconfidence we tend to think we know more than we do

Humans are usually more confident than correct. Try these anagrams: WREAT ETRYN GRABE OCHSA

Why dont we rely on intuition (common sense)?

Water
Entry Barge Chaos

Once people know the target word, hindsight makes the words obvious. We believe that we should have seen the solution in a shorter period of time.

Science vs. Common Sense

Common sense and intuition often tell us about psychology

e.g., suppose a study tells us that separation weakens romantic attraction common sense may tell us - out of sight, out of mind or common sense may say the opposite absence makes the heart grow fonder

Common sense can be inconsistent and based on hindsight

Science vs. Common Sense

Science helps build explanations that are consistent and predictive rather than conflicting and postdictive (hindsight) Science is based on

knowledge of facts developing theories testing hypotheses public and repeatable procedures

Scientific Inquiry/Research

Facts are what need to be explained

objective - viewable by others based on direct observation reasonable observers agree are true

Theory is a set of ideas that

organizes facts makes predictions about observations (new facts)

Hypothesis
prediction about new facts (often implied by theory) can be verified or falsified

Scientific Inquiry/Research
As a check on their bias, psychologists report their findings with

Operational Definitions (of Concepts):

a statement of the procedures used to define research variables
Operational Definitions allow others to replicate the study (repeating the essence of the study to get similar results).

Research Methods in Psychology

Setting

- field vs. laboratory Methods of data collection

self-report vs. observational
Research

plan or design

descriptive correlational experimental

Descriptive Research
Describes a set of facts Does not look for relationships between facts Does not predict what may influence the facts May or may not include numerical data

Example: measure the % of new students from out-of-state each year since 1980

Descriptive Research

Types of descriptive research

1. Case Study
In-depth study of one individual

2. Naturalistic Observation
Recorded observation of people or animals in their natural environment focus on a specific behavior

3. Survey

Surveys

Wording Effect - need to watch for bias, order of wording, too general
(attractiveness, intelligence)

Sampling
False Consensus Effect tendency to overestimate others agreement with us

Creates the need for a representative sample of the population

Surveys

Sampling Contd.
Population all the cases in a group, from which samples may be drawn for a study (e.g. study on studying habits population would be all students) Random Sample sample in which every person in a group has equal chance of participating

Correlational Resarch

Collects a set of facts organized into two or more categories

measure parents disciplinary style measure childrens behavior

Examines the relationships between 2 or more categories

e.g., more democratic parents have children who behave better

Correlational Research

Correlation CANNOT prove causation

Do democratic parents produce better behaved children? Do better behaved children encourage parents to be democratic?

May be an unmeasured common factor

e.g., good neighborhoods produce democratic adults and well behaved children

Correlational Research

Scatterplots graphed cluster of dots, each of which represents the values of two variables
See pg. 27 for examples

Correlation Coefficient a statistical measure of relationship

Statistical measure of the extent to which two factors vary together, and thus of how well either factor predicts the other.

Correlational Research

Direction of a Correlation

(slope of the points) Positive Correlation as one variable goes up, the other variable tends to go up (so as one goes down, the other goes down) max. +1.00 Negative Correlation as one variable goes up, the other tends to go down, the inverse is also true min. -1.00 No Correlation correlation is 0.0

Strength of Correlation
(amount of scatter) The higher the correlation coefficient is (without regard to sign) the stronger the correlation is.
The stronger the correlation is, the better one variable can predict the other.

Correlational Research
Correlations make visible relationships that we might otherwise miss. They also restrain our seeing relationships that actually do not exist illusory correlations perception of a relationship where none exists (e.g., superstitious beliefs) Believe likely to notice and recall

Perceiving Order
As humans we want to make sense of our world, so we look for meaningful patterns. Prob Random sequences often dont look random leads to illusory corr.
If someone flipped a coin six times, which of the following sequences of H and T would be most likely? HHHTTT HTTHTH HHHHHH

Correlational Research
A little note about correlation

CORRELATION DOES NOT PROVE CAUSATION!!

Experiments

Direct way to test a hypothesis about a cause-effect relationship between variables

One variable is controlled by the experimenter
e.g., democratic vs. authoritarian classroom

The other is observed and measured

e.g., cooperative behavior among students

Experimental Variables

Independent variable
the controlled factor in an experiment what the experimenter manipulates hypothesized to cause an effect on another variable

Dependent variable
the measured facts/data hypothesized to be affected

Independent Variable

Must have at least two levels

categories - male vs. female numeric - ages 10, 12, 14

Simplest is experimental vs. control

experimental condition gets treatment (e.g., gets the new drug) control condition does not (e.g., gets the placebo) (created when using a double-blind procedure)

Experimental Design

Levels may differ between or within people Within-subject experiment - different levels of the independent variable are applied to the same subject Between-groups experiment - different levels of the independent variable are applied to different groups of subjects

Experimental Design

Random sample - every member of the population being studied should have an equal chance of being selected for the study Random assignment - every subject in the study should have an equal chance of being placed in either the experimental or control group Randomization helps avoid false results

Experimental Design
Validity extent to which a test measures or predicts what it is supposed to Internal Validity controlled for confounding variables (if random assignment is violated, can destroy internal validity) External Validity generalizability (if random sample is violated, can destroy external validity)

Sources of Bias

Observer-expectancy effect
researcher has expectations that influence measurements

Subject-expectancy effect
subject knows design and tries to produce expected result

Blinding
minimize expectancy by removing knowledge about experimental conditions

Blinding
Single-blind study - when subjects are kept uninformed as to the treatment they are receiving Double-blind study - when both subjects and experimenter are kept uninformed about aspects of the study that could lead to differential expectations

Double Blind Procedures

Double Blind Procedures allow researchers to check for a treatments actual effects apart from the placebo effect - experimental results caused by expectations alone; such as a research participants enthusiasm for it or the healing power of belief

Sources of Bias
Confounding Variables Any variable besides the independent variable that affects the dependent variable in one group but not the other. (e.g., placebo takes a study skills seminar, non random assignment)

Research Settings

Laboratory
a setting designed for research provide uniform conditions for all subjects permits elimination of irrelevant factors may seem artificial

Field research
behavior observed in real-world setting poor control over conditions measures may be more representative of reality

Data-Collection Methods

Self-report - procedures in which people rate or describe their own behavior or mental state
questionnaires rating scales

on a scale from 1 to 7 rate your opinion of on a scale from 1 to 100 how hot is ...

judgements about perceptions

Data-Collection Methods

Observational methods - researchers directly observe and record behavior rather than relying on subject descriptions
naturalistic observation - researcher records behavior as it occurs naturally tests - researcher presents stimuli or problems and records responses

Data-Collection Methods

Cross-Sectional Studies study in which subjects of different ages are compared at a given time
Longitudinal Studies study in which subjects are followed and periodically reassessed over a period of time

Data-Collection Methods

Meta-Analysis combine and analyze data from many studies; it determines how much of the variance in scores across all studies can be explained by a particular variable