Methods in Psychology (unit 2)

psychology

• relies on individual experience/knowledge

• can be generalized

Limits of Intuition

we think we know more than we do

Overconfidence

our tendency to seek out info that confirms out previous beliefs (ignore info that goes against it)

Confirmation Bias

• an explanation

• integrates principles (current understandings)

• not a guess, based on scientific observations/evidence

• predicts behaviour/events

Theory

• testable prediction

• prompted by theory

• test enables us to accept/reject/revise theory

Hypothesis

description of property (concrete), measurable

Operational Definition

tells us we must:

1. develop a theory

2. make falsifiable hypothesis

3. test hypothesis by observing world

SCIENTIFIC METHOD

• studies one person in depth

• looks at behaviour

• issue w: don’t tell us everything about anyone else, cant generalize

Descriptive Methods: Case Study

• study animals/people in natural environment

• issue w: accurately records behaviour, but lack of control of external variables

Descriptive Methods: Naturalistic Observation

• questions random sample of people

• self reported opinions/behaviours

• difficult: wording (bias), sampling

• issue w: self reported makes unreliable, reliability of questions can be difficult to measure

Descriptive Methods: Survey

• examine relationsjip b/w 2 variables

• correlation doesnt equal causation

• data displayed on scatterplot (shows strength of relationship)

• values (R) ranges from +1 to -1 (no correlation r = 0)

• issue w: third variable problem

Descriptive Methods: Correlation

2 variables change in same direction (both increase/decrease)

Positive Correlation

2 variables change in opposite directions (one increases, other decreases)

Negative Correlation

• issue w correlation studies

• both variables influenced by some other variable not being measured

Third Variable Problem

• only way to determine cause + effect

• manipulating/controlling variables

• independent + dependent variables

Research Methods: Experimentation

I: thing being manipulated (cause)

D: thing being measured (effect/outcome)

Independent v Dependent

• randomly assign participants into groups

• allows for control

Random Assignment

• done to avoid changes in participants behaviour due to cues given by experimenter

• participants unaware of which group they’re in

Blind

• participants + those conducting experiment are unaware of which patients get the real treatment until after

Double Blind

single # represents info about overall data

3 types

• mean

• median

• mode

Measures of Central Tendency

• average

• sum of set of scores divided by # of scores

Mean

score that appears most often

Mode

• score divides data in half, same # of scores above/below median

• must be in order from lowest to highest

Median

reasons to use animals

• replacement (no alternative to animal use)

• reduction (use smallest amount of animals possible)

• refinement (experiment must be modified to reduce discomfort/pain)

Research on Animals

key ethical principles

• informed consent

• freedom from coercion

• protection from harm

• risk benefit analysis

• deception

• debriefing

• confidentiality

Research on Humans

D: “belief” → develop theories of body functions
E: “experience” → examine sick people (knowledge of world gained from observing)

Dogmatists v. Empiricists

• feature of operational definition

• specified operations are considered good indicators of specified properties

Construct Validity

aspects of observational setting that cause people to behave as they think someone wants/expects

overcome:

• natural observation

• privacy/control (kept anonymous)

• unawareness (don’t know what is being observed)

demand characteristics? how to overcome?

graphic representation shows # of times that the measurement of a property takes on each possible value

Frequency Distribution

• frequency of measurements is highest in middle

• decreases symmetrically in both directions

• BELL CURVE

Normal Distribution

how each measurement differs from mean/centre

Standard Deviation

• conduct an experiment

• has manipulation + random assignment

how to get around 3rd variable problem

• trait of an experiment

• establishes casual relationships

Internal Validity

researches conclude there’s a casual relationship when there’s not

Type I error

researches conclude there’s not a causal relationship when there is

Type II Error

ethics board contains:

• 1 non scientist

• 1 person not affiliated w institution

If study is federally funded, what is required