research methods and data analysis midterm

theory

ideas about how the world works

epistimology

the science of knowing

methodology

the science of finding out

multicausality

there’s multiple causes of things

paradigm

theoretical frameworks that structure what we observe and how we interpret it

importance of theory

prevents flukes, explains why patterns make sense, suggests avenues for social action, suggests new paths of research, determines the why of social patterns

paradigm shift

‘shift’ in the basic concepts, assumptions, and methods of scientific inquiry

macrotheory

theories of large portions of society

microtheory

theories of smaller portions of society (small groups, social networks)

mesotheory

not concerned with social life, attempts to generalize theories

social theory

systematic explanation for the observed patterns of social life (how the world is, not how it should be)

deduction

general to particular

induction

particular to general

research ethics

consensus norms regarding proper conduct in research

pillars of research ethics

voluntary participation, no harm to participants, and informed consent

belmont report

fundamental ethical principles involving human subjects (respect for persons, benefits, and justice)

informed consent

participants must be informed of all details of the study before consenting

anonymity

response cannot be linked to subject

confidentiality

researcher can identify subject, but won’t

deception

withholding certain info that could invalidate the study

ethical considerations in analysis and reporting

selective reporting, data fraud, plagiarism, carelessness of analytical methods

institutional review boards

assure risks are minimized at research institutions

objectivity and ideology

weber-sociology need to remain unencumbered by personal values, marx disagreed. research should have an intentional impact

political considerations of social research

conflicts of interest, ‘politicization’ of science, scientific censorship

3 purposes of social research

exploration, description, explanation

exploration

satisfy initial curiosity, test feasibility, develop methods for a pilot study

description

what is observed

explanation

explain why observed patterns exist

ideographic explanation

understanding of causes of outcomes in single/small number of cases (case studies)

nomothetic explanation

identify factors that cause ______ in general

dependent variable

affected role (being caused)

independent variable

causal role (doing the causing)

nomothetic causality

x causes y, if you change x then y changes too

correlation

x is associated with y

time order

x before y

nonspuriousness

there’s no alternative explanation

reverse causality

y causes x

omitted variables bias

leaving out a variable to show causation or a connection

necessary causes

condition must be present for effect to happen

sufficient causes

condition guarantees the effect

unit of analysis

what/who is being studied

ecological fallacy

assumption that something about the unit says something about individual making up that unit

reductionism

reducing the explanation to limited/low-order concepts

cross-sectional study

study done at one point in time (snapshot)

longitudinal study

data collection at different points in time

trend study

change over time

cohort study

subset studied over time

panel study

examine same units at each point in time

mixed methods

used to strengthen weakness of one method with the strengths of another

conceptualization

clarifying what things you are studying

concepts

theoretical marker representing class of things

indicator

observation that is reflective of a concept/variable

dimension

specific aspect of a concept

operationalization

process by which concepts are measured empirically

variables

logical set of attributes

attributes

characteristic/quality of something

nominal

no intristic order

ordinal

intristic order

binary (dichotomous variable)

two exhaustive categories

interval

rank ordered and equal distances between adjacent attributes

ratio

interval measures has a ‘true zero’ point

mutually exclusive

each observation can only take on one value for given variable

exhaustive

all possible values of a measure must be included

precision

‘fineness’ or resolution of the attributes measured

accuracy

extent to which attribute accurately reflects social reality

measurement validity

degree to which a variable’s operationalization accurately reflects the concept

construct validity

degree to which construct relates to other variables as expected

content validity

does the measure capture all content domain of the construct?

face validity

measure construct of interest

criterion-related validity

check performance of measure against an external criterion

measurement reliability

the extent to which operationalizations of the concept show consistent results

test-retest reliability

tests across time should show consistent results

internal reliability

consistent responses across similar items

inter-rater reliability

consistent rating by different observers