SOC150

evaluation research

test the effectiveness of a social policy or program

selective observation

noticing social patterns one has experienced or wishes to find

overgeneralization

assuming social patterns exist based on limited observation

data collection methods

detect social regularities

data analysis techniques

interpret social regularities

theory

explain social regularities

variable

characteristic or measure of a social phenomenon that can take different values

exploratory research

understand and topic, examine new areas of inquiry, look at extent of problem/behaviour, test feasibility of doing an extensive study

descriptive research

describe/define a topic, make observations and detailed documentation about a phenomenon

explanatory research

look for explanations or causal relationships, answer why and how questions

inductive approach

start with empirical observations, look for patterns, make theories

deductive approach

start with theory, develop hypothesis/model, collect and analyze data to test hypothesis/model

hypothesis

assumption about the relationship of 2+ variables

sampling

process of selecting cases to be analyzed

sampling frame

list of all members of a population, which are used to select cases for the sample

nonprobability sampling

likelihood of person being selected for the sample is unknown

probability sampling

likelihood of person being selected for the sample is known

repeated sampling

sharpens precision, more credible samples across polls should improve accuracy

simple random sampling

researcher gives all members of a population equal chance of being in the sample

systematic sampling (selection interval)

selecting every kth case to be in the sample

stratified sampling

dividing population into 2+ mutually exclusive subgroups and then drawing a sample from each subgroup

proportionate stratified sampling

cases chosen from strata are in line with numbers in larger population

disproportionate stratified sampling

sizes of subgroup samples are not in line with relative sizes in the population (oversampling and weighting)

pillars of science

logic and observation

levels scientific research operates on

theoretical and empirical

probabilistic relationship

when two variables go together with some degree of regularity

relational statement

connects 2+ variables (one variable gives us information about the other)

empirical paper

presents data collection/analysis of existing data

theoretical paper

focuses on advancing or critiquing theory with little to no empirical data

review article

synthesizes existing data, may or may not have a strict systematic protocol

systematic review

uses explicit and replicable methods to identify/select relevant research

gray literature

sources not published through traditional commercial or academic avenues

acquiescence bias

tendency for participants to agree with statements regardless of how they feel

random error

unpredictable variations/fluctuations in measurement

systematic error

bias skewing data in one direction