UConn HDFS 2004W Midterm - Bellizzi

3 purposes of social research

exploratory, descriptive, explanatory

basic vs applied research

basic: sometimes justified in terms of gaining knowledge for knowledge's sake

applied: putting research into practice

non-scientific inquiry vs scientific inquiry

non-scientific: direct experience and observation, tradition, authority

scientific: method of understanding social phenomenon while trying to minimize error

errors --> solutions of human inquiry

inaccurate observations --> measurement devices add precision

overgeneralization --> replication is when you repeat a study to make sure the same results are produced each time in different samples

illogical reasoning --> use systems of logic explicitly

premodern view of reality

all knowledge was based on knowledge revealed by authority figures

modern view of reality

knowledge based on empiricism and logic

postmodern view of reality

there is no object reality to observe, all that is real is the images we get from our own points of view

two pillars of social science

logic and observation

methodology

the science of finding out

epistemology

the science of knowing

different components of a research article

abstract

introduction

methods

results

discussion

references

macrotheory

larger lens in which we view a social phenomenon

microtheory

very specific, focuses on individual level factors

idiographic explanation

An approach in which we seek to exhaust the idiosyncratic causes of a particular condition or event.

list ALL reasons

nomothetic explanation

An approach to explanation in which we seek to identify a few causal factors that generally impact a class of conditions or events.

list SOME reasons

theory

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

paradigm

broad lens in which we see the world

inductive theory

bottom up

logical model in which general principles are developed from specific observations

deductive theory

top down

the logical model in which specific expectations or hypotheses are developed on the basis of general principles

conceptualization

produces an agreed upon meaning for a concept for the purposes of research

describes the indicators that we'll use to measure the concept and the different aspects of the concept

operationalization

the development of specific research procedures that will result in empirical observations representing those concepts in the real world

observation

measurements in the real world

dimension

a specifiable aspect of a concept

Ex: for religion - belief, ritual, devotional

indicators

an observation that we choose to consider as a reflection of a variable we wish to study

the question we use to measure a construct

variable

logical groupings of attributes

education, income, age

attribute

characteristics or qualities that describe an object

college educated, high SES, 21

different types of variables

situational

participant

response

mediating

situational variable

a characteristic that differs across environments or stimuli

participant variable

A variable such as age, gender, or ethnicity whose levels are selected (i.e., measured), not manipulated.

response variable

measures an outcome of a study

mediating variable

variables that mediate the effects of a situational or participant variable on a particular response

qualitative data

nonnumerical data, exploratory research

quantitative data

numerical data, makes observations more explicit and makes it easier to aggregate, compare, and summarize data

different relationships between variables

positive linear

negative linear

curvilinear

no relationship

3 basic principles of Belmont Report

respect for persons

beneficence

justice

respect for persons

treat individuals as autonomous human beings, capable of making their own decisions

not using people as means to an end

capacity to understand and process information

freedom to volunteer or withdraw

beneficence

minimize risk of harm

maximize benefit of research

justice

selection of participants is equitable

fair procedures and outcomes in the selection of research subjects

burdens and benefits are shared equitably

ethics

standards or codes of behavior expected by a given profession or group

ethical issues social scientists should be aware of

voluntary participation

no harm to participants

anonymity and confidentiality

deception must be justified by compelling scientific concerns

researchers must be honest about their findings and research

voluntary participation

no one should be forced or coerced to participate

easier to accept in theory than to apply in practice

scientific goal of generalizability is threatened by this requirement

no harm to participants

revealing of information that would embarrass participants or endanger

no psychological harm

anonymity

the researcher cannot identify a given response with a given respondent

confidentiality

the researcher can identify a given person's response but promises not to do so publicly

debriefing

interviewing subjects to learn about their experience of participation in the project and to inform them of any unrevealed purpose

researchers must be honest

ethical obligation to the scientific community

obligation to make all limitations and shortcomings known to their readers

why do we use informed consent?

subjects in a study must base their voluntary participation on a full understanding of the possible risks involved

independent variable

presumed to cause or determine a dependent variable

dependent variable

assumed to depend on or be caused by the independent variable

nomothetic

tries to find key independent variables that account for the variations in a given phenomenon

idiographic

focuses on a complete, in-depth understanding of a single case

criteria for nomothetic causality

correlation

temporal relationship

non-spurious

consistency (replication)

unit of analysis

the major entity that you are analyzing in your study

individuals

social groups

formal organizations

social artifacts

null hypothesis

no relationship between two variables

hypothesis testing

make and test an educated guess about a problem/solution

necessary cause

a condition that MUST be present for the effect to follow

sufficient cause

a condition that if present, GUARANTEES the effect in question

experimental method

manipulating IV in a controlled environment

non-experimental method

observing behavior as it occurs naturally in environment

cross-sectional design

observations of a sample or cross-section of a population or phenomena that are made at one point in time

descriptive or exploratory

longitudinal design

permits observations of same phenomenon over an extended period

trend

cohort

panel

trend study

examines changes within a population over time

cohort study

examines specific subpopulations or cohorts as they change over time

panel study

examines the same set of people each time

unique

reliability

questions of stability and consistency

whether the same measurement process yields the same results

test-retest

inter-item

test-retest reliability

take the same measurement more than once

inter-item reliability

multiple items used to measure same concept are correlated

validity

the extent to which we are measuring what we hope to measure

face

content

criterion

construct

face validity

whether a measure appears to measure the concept it's intended to measure

criterion validity

applies to instruments that have been developed for usefulness as an indicator of a specific trait or behavior, either now or in the future

driving test has good predictive validity of driving performance

construct validity

degree to which a measure relates to other variables as expected within a system of theoretical relationships

self-esteem and depression

content validity

extent to which a measure adequately represents all facets of a concept

different indicators of depression

4 levels of measurement

nominal

ordinal

interval

ratio

nominal

placing of data into categories without any order or structure

numbers are arbitrary and do not put participants in order

1 - male 2 - female

ordinal

describing a variable with attributes we can rank order along some dimension

interprets order but not distance between values

how important is religion to you?

interval

distance between points on scale are equal

numbers represent equal intervals in different segments on a continuum

equal differences in scores = equal distances in variables measured

NO REAL ZERO

ex: IQ

ratio

similar to interval but HAS REAL ZERO

ex: income

why do social scientists use composite measures?

trying to measure something we can't physically observe or see

index

set of items (questions) that focuses on multiple aspects of a dimension or domain of behavior, attitudes, or feelings into a single indicator or score

sometimes called composites, inventories, or tests

often designed for exploring the relevant causes of behaviors or health outcomes

constructing an index

Select items for a composite index.

Examine empirical relationships.

Assign scores for responses.

Validate the index.

Handle missing data.

scale

a cluster of items (questions) that taps into a unidimensional or multidimensional concept of behavior, attitudes, or feelings

takes advantage of differences in intensity among items

superior to indexes

construction of a scale

assigning scores to patterns of responses

selecting items (constructing index)

face validity

unidimensionality - items should measure one concept

general or specific - could capture specific or multiple domains of a concept

empirical relationships (constructing index)

established when respondents' answers to one question is correlated with other questions in the index

if two items are empirically related, we can argue that each reflects the same variable, and both can be included in the same index

assign scores (constructing index)

decide the desireable range of index scores

validate the index (constructing index)

item analysis - internal validation

external validation - people who score a particular way on one index should appear similar to other measures of the same concept

ways to handle missing data (constructing index)

exclude cases with missing data from index and analysis

treat missing data as one of the available responses

analyze missing data to interpret the meaning

Thurstone scales

judges determine the intensity of different indicators

developed in 1929 for measuring a core attitude when you have multiple dimensions or concerns around that attitude

ex: prejudice - strong, intermediate, or weak indicator?

single score assigned to respondent based on pattern of response to several items

Likert scale

uses standardized response categories

five point bipolar response format

agree/disagree, approve/disapprove

Guttman scales

examines strong and weak indicators of some variable

uses set of items so that a respondent who agrees with any specific question in the list will also agree with all previous questions

semantic differential scale

deals with emotions and feelings

based on idea that people think in polar opposites

asks respondents to rank answers between two extremes

Kaplan's classes

direct observables

indirect observables

constructs

direct observables

things that can be observed simply and directly

indirect observables

things that require more subtle observations

constructs

based on observations that cannot be observed