A good hypothesis is

testable, clear and specific, based on a theory or prior experience, predicts a relationship between variables, falsifiable

Null hypothesis (H0)

there is no effect, difference, or relationship

Alternative hypothesis (H1 or Ha)

there is an effect, difference, or relationship

bad hypotheses may be

too vague, untestable, based on opinions, missing variables, impossible to measure

experimental

researcher manipulates independent variable

non-experimental

researcher observes variables without manipulation

Criteria for Causal Inference

1. Covariation --> variables are related 2. Temporal precedence --> cause comes before effect 3. Elimination of alternative explanations --> control for confounding variables

Independent Variable (IV)

variable manipulated by researcher Ex: amount of caffeine

Dependent Variable (DV)

Outcome being measured Ex: test performance

Control Variable

variables kept constant Ex: same testing room

Control Group vs. Control Variable

Control Group: group that does not receive treatment Control variable: specific factor held constant

Scales of Measurement

Nominal --> categories (gender, eye color, major) Ordinal --> Ranked categories (class rank, satisfaction ratings) Interval --> Equal intervals, no true zero (temperature in celsius) Ratio --> equal intervals with a true zero (height, weight, age, income)

Why scales matter

appropriate statistics, types of graphs, conclusions you can make

Test-retest reliability

consistency across time

inter-rater reliability

consistency between observers

internal consistency

items on a measure are consistent with each other

Face validity

looks like it measures what it should

content validity

covers all parts of a concept

construct validity

actually measures the theoretical construct

criterion validity

predicts or relates to a relevant outcome

Simple Random Sampling

Strengths: least biased, most representative Weaknesses: difficult to obtain

Convenience Sampling

use whoever is easiest to reach Strengths: fast, cheap Weaknesses: high bias, low generalizability

Stratified Sampling

Population divided into groups, then sampled Strengths: better representation Weaknesses: more complicated

Cluster Sampling

Randomly select groups/clusters Strengths: efficient Weaknesses: clusters may differ from population

Systematic Sampling

Choose every nth person Strengths: easy Weaknesses: can introduce bias if patterns exist

Institutional Review Board (IRB)

protect participants, ensure ethical research

Informed Consent

Participants must know purpose of study, risks, benefits, their rights, ability to withdraw

Debriefing

Occurs after participation. researches explain true purpose, any deception used, participant questions

Common ethical concerns

lack of informed consent, excessive risk, coercion, privacy violations, lack of confidentiality, harm to participants

Ethical Grey Areas

Situations where no answer is perfectly clear Ex: Deception --> acceptable only is necessary for study, risk is minimal, participants are debriefed Observation in Public --> is privacy expected? was permission needed? Using Existing Data --> Can individuals be identified? Is confidentiality protected?