Methods Final - Lecture

Dr. Paul S26

methods of knowledge acquisition

ways of knowing about the world including intuition/anecdote, authority, and empiricism

intuition/anecdote

knowledge without a known source based on instinct, hunches, or “friend-of-a-friend” information

common sense

belief based on agreement between experience and how we think the world works, often incorrect or oversimplified

emotion

using feelings as evidence for knowledge, often difficult to override even when incorrect

mysticism

gaining knowledge through insight or altered states of consciousness

limitations of intuition/anecdote

limited experience, cannot be fully trusted, insufficient information for certainty, varies across cultures and time

authority

believing something is true because a respected source says it is

limitations of authority

source reliability issues, personal bias, mistakes, discourages questioning, can suppress new ideas

empiricism

knowledge gained through direct observation and real-world evidence

scientific approach

systematic observation where findings can be verified by others

science is not defined by subject matter

science is defined by method, not topic

science is not defined by tools

use of equipment does not make something scientific

data over intuition/authority

scientific conclusions rely on evidence rather than belief or opinion

science is adversarial

findings are challenged and require peer review

testable problem

a question that can be answered using scientific methods

non-scientific question

a question that cannot be tested or measured empirically

falsifiability criterion

a scientific theory must be able to be proven wrong

falsifiability question

“what evidence would disprove this theory?”

hypothesis

statement describing relationships among variables

prediction

specific expected outcome based on a hypothesis

independent variable

variable manipulated by the experimenter

dependent variable

variable measured to assess the effect of the independent variable

control variable

factors held constant to prevent them from influencing the dependent variable

extraneous variable

any variable other than the independent variable that could affect the dependent variable

experimental group

group that receives the manipulation of interest

control group

group used for comparison that does not receive the manipulation

construct

a variable that exists but cannot be directly observed

examples of constructs

intelligence, love, aggression, happiness

operational definition

defining a variable in terms of how it is measured or manipulated

likert scale

a self-report scale typically ranging from 1–7 used to measure attitudes or feelings

steps in research design

selecting and planning how to manipulate the independent variable including number of levels and type of manipulation

levels of an independent variable

different conditions or values of the independent variable

between-subjects design

different participants are assigned to each level of the independent variable

requirements for an experiment

must manipulate at least one independent variable and ensure group equivalence

random assignment

assigning participants to conditions by chance to ensure groups are equivalent

treatment effect

differences in performance between groups due to the independent variable

control by matching

ensuring groups are similar by balancing characteristics before assignment

matching

creating mini-groups of similar participants and randomly assigning within those groups

rule of 30

recommendation to have about 30 participants per group

more conditions problem

increasing number of conditions requires more participants

within-subjects design

same participants are exposed to all levels of the independent variable

advantage of within-subjects

groups are automatically equivalent since same participants are used

within-subjects efficiency

requires fewer participants than between-subjects designs

statistical sensitivity

within-subjects designs are more sensitive to detecting small effects

carryover effects

effects of one condition influencing behavior in another condition

practice effect

improvement due to familiarity with the task

fatigue effect

decline in performance due to tiredness or boredom

general carryover effect

performance influenced by previous conditions overall

counterbalancing

controlling for order effects by varying the sequence of conditions

complete counterbalancing

using all possible orders of conditions

systematic counterbalancing

using a subset of orders, such as a Latin square, to balance conditions

latin square design

each condition appears once in each position and follows every other condition equally

random counterbalancing

using random order of conditions for participants

subjects vs participants

humans are participants, non-humans are subjects, though the terms are often used interchangeably

power analysis

statistical method used to determine required sample size

sample size determination

depends on desired accuracy and statistical power

precision of results

increases with larger sample sizes

non-experimental research

research that describes relationships without manipulating variables

limitation of non-experimental research

cannot determine cause-and-effect relationships

survey research

collecting self-report data from many participants using questionnaires

open-ended questions

allow participants to respond freely without fixed options

advantage of open-ended questions

less researcher bias and richer data

disadvantage of open-ended questions

difficult to score and analyze

restricted questions

questions with fixed response options

closed-ended questions

questions with fixed response options

advantage of restricted questions

easy to score and higher response rates

disadvantage of restricted questions

limited depth of responses

partially open-ended questions

fixed responses with an “other” option for additional input

survey delivery methods

ways surveys are administered, including in-person, telephone, and internet surveys

face-to-face interviews

direct interaction allowing clarification and probing

advantages of face-to-face interviews

can clarify questions and gather more detailed responses

disadvantages of face-to-face interviews

expensive, time-consuming, and may create interviewer bias

telephone surveys

surveys conducted by phone

advantages of telephone surveys

less expensive than in-person and may reduce interviewer bias

disadvantages of telephone surveys

low response rate and should usually be short

internet surveys

surveys conducted online

advantages of internet surveys

low cost, large and diverse samples, fast data collection

disadvantages of internet surveys

sampling issues, possible multiple responses, and uncertainty about who is responding

survey design issues

problems that affect quality of responses

leading questions

questions that bias participants toward a certain answer

double-barreled questions

questions that ask about two things at once

question order effects

earlier questions influence later responses

response sets

tendency to answer similarly across items regardless of content

good survey practices

keep questions short, neutral, consistent, and clearly labeled

archival research

using existing data collected for other purposes

statistical records

large datasets collected by organizations, such as census or weather data

content analysis

analyzing existing records or media to identify patterns

meta-analysis

combining results from multiple studies to determine overall effects

naturalistic observation

observing behavior in natural settings without manipulation

participant observation

researcher becomes part of the group being studied

concealed observation

participants are unaware they are being observed

advantages of naturalistic observation

high external validity and realistic behavior

disadvantages of naturalistic observation

ethical concerns and lack of control

systematic observation

structured observation of specific behaviors

contrived setting

artificial environment created by researcher

reactivity

participants change behavior because they know they are being observed

qualitative research

non-numerical data such as descriptions or narratives

archival data advantages

inexpensive and useful for large-scale patterns

mixed design

includes both between-subjects and within-subjects variables

factorial design

study with more than one independent variable