SOC 380 Midterm

logical aspects of science

patterns in observations and comparison to logical assumptions

Begins with theory > hypothesis > observation > confirmation

begins with observation > pattern > tentative hypothesis > theory

gaining knowledge to understand social processes

answers a question in the real world to address concrete problems

cultural knowledge shared with those around us

knowledge based on shared cultural understanding
knowledge based on the status of the discoverer

innacurate observations, overgeneralizations, selective observations

most daily observations are casual and semiconscious

assuming few observations indicate evidence for a general pattern

confirmation bias

human behavior is a product of personal experiences and historical forces (self- and social-consciousness)

factors to consider: social historical context -> social norms, institutions -> insight into social relations

- research answers questions of HOW, with the goal of documenting how particular processes unfold or how something is defined
- explore a problem that isn't clearly defined
- **no hypothesis

- describing and quantifying phenomena
- describes trends, variations, and patterns
- no hypothesis

- research documents causes and effects of social phenomena
- questions WHY
- independent and dependent variables
- hypothesis

government data, administrative records, surveys, participant observation, interviews, focus groups, ethnography

surveys

census, admin records, gov data

numerical data
+easier to compare and summarize data
+consistent
+easy to replicate and analyze
- the why is missing
- not in-depth

textual data
+ flexible
+ development of theory
- small n (number of people)
- time consuming
- not easy to generalize
*ethnographic, focus groups, interviews, participant observations
**researchers must acknowledge their own perspectives

field work, observation in small groups over time period
+ in-depth
- results difficult to replicate

set of questions given to group to illicit conversation and understanding
(usually exploratory)

participants drive conversation

Wollcot styles
1. active participant: has a job to do in the setting in addition to the research
2. privileged observer: someone who is known and trusted and given access to info
3. limited observer: asks questions and builds trust but doesn't have public role other than researcher

overarching, broad categories

shared definition and meaning

more refined meaning-- different aspects of a concept

observation we choose to consider as a reflection of what we wish to study

population studied

between concepts and dimensions

between dimensions and indicators

testable statement of relationship

systematic explanation for the observations that relate to a particular aspect of life

1. macro: states, gov, populations
2. meso: groups, organization, local context
3. micro: individuals, interactions, dyads, family

world view (examples: positivism and postmodern)

knowledge can be confirmed or refuted through observation

significant skeptcism of claims about facts or principals- social reality is relative

structural functionalism
marxism
conflict
rational choice symbolic interactionism

-macro
-Durkheim
- how structures and institutions work together for betterment of society (stability)

- macro
- lens of struggle between powerful and less powerful; those with more power use societal resources to maintain position
- proletariat vs bourgeosie

- macro
- Weber
- society as structure defined by ongoing struggles between groups for power, resources, and status

-micro
-Mead
-words and action convey meaning
-subtance of society in individual interactions

Institutional Review Board
-human subject research
-protect rights and welfare of human participants
-rights of participants
*respect: informed consent and voluntary
*beneficiance: to do good and minimize risks
*justice: fair and no one systematically prevented from participation

-measurement quality must be precise and accurate
-reliable
-valid

ability to achieve the same result
-precision
*test-retest: same person more than once
*intercoder reliability: double coding (two or more people)
*split-half: separate items into two lists randomly, have same person take both tests, calculate the average

more reliable: ask questions people have experience with, relevance

*qualitative

empirical measure relfects meaning of the concept
internal validity
*quantitative

face validity: does this make sense

criterion-related validity: concurreent- correlate w/ well-established measure of same concept /// predictive- how well does measure predict something else

content: coverage more complete

construct: items to measure connect to underlying concept

data mining: income, numbers, etc.

1. predicitive: how well does it correlate with measure predicts
2. concurrent: does it correlate with well-established measure of the same concept

operational definition matches actual measurement

belief you can effectively achieve goal
scale: not at all true, hardly true, moderately true, exactly true

how representative is group
how real is the study

how will this be measured

vague concept to specific question

consider: definitions, enough info for others to recreate

variable controlled by researcher; affects dependent

effect

how knowledge is created

acknowledge and understand influence of researchers' ideas/perceptions

1. grounded theory
2. phenomenology
3. ethnography
*ALL: data collection stop when "saturated" (nothing new); data is coded, or translated into standardized categories; writing memos

generate and discover theory
*inductive inquiry: data generates theory
*data collected and analyzed at the same time (theoretical sampling: usually interview or focus groups)
*questions core phenomenon, causal conditions, strategies, consequences
*used when there's no/limited theories
*analyze patterns and interpret human understanding

lived experience
-socially constructed
*people describe personal experience
1. develop understanding of experiences
2. describe meaning of experience
3. interview people who are hetereogeous
4. focus on what people have in common

-identifies impact, searches for cause and effect
-does not seek to generalize to broader population
-critique: lacks reproductability, generalizable, potential researcher bias

understand world as researcher participants experience
-can be observer or participant
-can be in-depth interviews

1. choose question
2. discovery process
3. what will end product look like?
4. what do you want to say?
5. what do results look like?
6. is it an explanation for phenomena?

two or more methods used to expand, inform, confirm, cross-validate, or corroborate findings within a study
*leverage the things that each are good at
*capital letters denote priority (QUAL vs QUANT)
*sequential exploratory and explanatory designs

both styles at the same time
-equal weight to qual and quant (separate until the end)
-triangulation
-difficult if results don't agree

direct comparison and contrast of quant and qual results for validation and corroboration

QUAL helps inform quant
1. explore issue qualitatively first
2. can help identify theories to test, relevant questions, creating surveys, etc

QUANT --> qual
-qual helps explain QUANT results