Research Methods Unit 1 Terms

micro

study of social influences on people, small group interactions (typically less than 10 people)

meso

study of larger groups like organizations and social networks

macro

study of large-scale, broad social systems like the education system, political system, the economy, and comparisons of geopolitical units

ecological fallacy

using data or measurement concepts from group level data to draw conclusions about individuals

quantitative data collection

data collection that distills people’s responses and reactions to numbers that can be statistically analyzed

qualitative data collection

data collection involves thorough descriptions of people’s lived experiences and/or written/visual media

mixed methods data collection

a study that has both quantitative and qualitative aspects as main focus

triangulation

use of mixed methods to study the same research question to ‘hone in’ on findings about that question

helps to balance those issues to generate a thorough accounting of the research question at hand

basic research

research that seeks to answer a fundamental intellectual puzzle about our social world

no immediate goal to produce something or evaluate a program

overall intent is to understand the phenomena better/more completely

applied research

research that addresses a concrete, specific, localized real-world problem or evaluate a policy/program

goals are immediate in nature to a specific issue

scientific method for class

1. identify an important question that needs an answer

2. two parts

   1. construct a hypothesis/prediction about the answer to that question, or:

   2. construct a way to generate a better understanding of that question

3. gather data that allows for the assessment of the accuracy of that prediction

deception

deceiving participants about some facet(s) of the study

theory

a set of related arguments consisting of a series of logically related statements that explicate some element of social life

does not try to explain everything about social life

textbook: well-reasoned supposition about a social phenomenon that moves logically and systematically from one point to related points and to a conclusion or expectation

empiricism

the world can be subjected to observation to gather data about social phenomena

theory without empirical evidence is just speculation in search of proff

purposes of theory

1. to explore: theory guides researchers studying social phenomena by pointing researchers to important research questions and ways of considering the world

2. to describe: capturing thick, rich, detailed data of as many elements of a social phenomena as possible to generate an understanding of how and why it occurs

3. to explain: predict a process through which a social phenomena unfolds, pose hypotheses that can then be empirically tested that would then support or not the theory

inductive (ideographic) theory

start with a research question to guide data collection

1. empirical social reality —> 2. theory (moving up)

deductive (nomothetic) theory

start with a research question and use theory to guide hypotheses

1. theory —> 2. empirical social reality (moving down)

paradigms

very abstract set of assumptions about the way the world is

concept

an idea that can be named, defined, and eventually measured in some way

importance of a strong definition

neither true nor false

theoretical statements

a conditional relation among abstract, general concepts

domain

the phenomenon to be explained by the theory

scope

condition(s) under which a theory is meant to apply (i.e. explain the domain)

given scope conditions, X —> Y because of T

X = casual concept (cause)

Y = resultant concept (effect)

—> = statement of relation

T = tentative explanation for why X and Y are related

growth of theory - elaboration

using existing conceptual ideas or models as a foundation to develop new theoretical insights

growth of theory - proliferation

the deliberate creation and development of multiple competing and even contradictory theories to foster the growth of knowledge

growth of theory - competition

creating incentives for researchers to seek priority in making discoveries, obtain funding, and gain prestige

growth of theory - integration

combining concepts, variables, and methods from different theories or research to create a more comprehensive and robust theoretical framework

nomothetic/deductive causality

establishing that change in the causal concept creates change in the resultant concept

nomothetic causality = internal validity

3 criteria for internal validity - 1. association/correlation

a statistically significant correlation between two variables

3 criteria for internal validity - 2. temporal order

a change in the causal variable must come before a change in the resultant variable

effects do not spurn changes in causes

3 criteria for internal validity - 3. nonspuriousness

is X the true cause of Y, or is some third vairble, Z, the cause of both

Z = spurious cause

temporal order: Design

cross sectional: data are collected at one point in time

• a snapshot

longitudinal: data are collected at two or more points in time

• repeated cross sectional study

  • panel study

when is Z a non-problem?

if we know what it is

establishing non spuriousness with control variables

when using survey data, varibles are “statistically controlled”

a 3rd varible is held constant so that the relationship between the causal and resultant concepts’ varibles can be examined without it affecting that relationship

statistical control (establishing nonspuriousness)

1. the relationship between X and Y disappears

   1. suggests a spurious relationship…therefore not causal

2. the relationship between X and Y persists (even if weakened)

   1. suggests that the relationship is not spurious…therefore stronger evidence for causality

mediation

there is an intervening variable, Z, between X and Y

BUT, when Z is a mediator it creases a causal link between X and Y

create a causal connection between the cause and effect

moderation

variables that alter the relationship between the cause and effect

affect the nature of the relationship between the cause and effect

conceptualization - from theory to hypothesis

precisely defining ideas/meaning of each concept and turning them into variables

conceptual: a definition in abstract, theoretical terms that gives meaning to all aspects of a concept relevant to the theory

operalization - from theory to hypothesis

linking the conceptual definition to a specific set of measurement techniques or procedures

turns concepts into something you can measure

from conceptualization to operationalization

conceptualizations matter for how we think of and measure that thing

concepts must be:

• exhaustive: all elements of that concept are represented

  • mutually exclusive: only elements of the concept is part of the measure and not elements of a different concept (conceptual boundaries)

units of analysis

level of social life that a research question is about

individuals or some type of aggregation

• aggregations: counting or averaging individual-level data in some grouping to capture group-level data

reductionist fallacy

using data or measurement concepts from individual-level data to draw conclusions about groups

levels of measurement: categorical

variables have a finite set of possible values that are fixed and distinct from one another with unknown differences between them

levels of meausrement: categorical

nominal varibles: varibales with states that are parallel and cannot be ranked or orderred; attributes are only exhausitve and mutually exclusive

levels of measurement: categorical

ordinal varibles: variables with categories that can be ordered in some way but have unknowable differences/distances between them

levels of measurement: continuous

varibles that could have an infinite set of possible values that exist on a continuum from low to high with meaningful and identifiable differences between them

levels of measurement: continuous

interval varibles: varibales with a continuum of values with equal meaningful distances (or intervals) between them but no true zero

levels of measurement: continuous

ratio varibles: interval varibles that do have a true zero

writing a hypothesis in sentence format

depends on how the varibles are hypothesized to relate to each othe rnad the type of varibles they are (categorical or continuous)

if IV and/or DV is categorical then direction of association rules do not apply

measurement error

true differences

error

• systematic measurement error: something consistently wrong with how your measure is working

• random measurement error: error happening at random in data

face validity

jusdement by the scientific community that an indicator really measures the concept

content validity

is the full content of the concept’s definition represented in the measure

construct validity

how well do multiple indicators of a concept connect together (hang together) to reflect a single, underlying factor

divergent validity

is it the case that indicators of one concept that should not be associated with other concepts

reliability across time

stability reliability

test-retest reliability: strong positive correlation between scores on a measure administed to the same sample at two points in time

• intra- rater or intra-observer reliability when talking about raters or observers

reliability across indicators

inter-item reliability: strong correlation among all items in a composite measure

• suggests a homogenous set of items with a strong correlation

reliability across observers

intercoder or interobserver reliability: strong correleation in data coding produced by different coders or observers

reliability across populations

representative reliability: consistency of response across populations of people

• subpopulation analysis

improving reliability

clearly conceptualize all concepts to eliminate noise

each measure should indicate one and only one concpet/feature of the concpet

increase level of measurement

use multiple indivators

use pretests, pilot studies, replication