Research Methods

3 ways of knowing the truth

3 ways of knowing the truth

1. faith - religious beliefs

2. consensus - based on what experts say

3. science

empirical observatoin/data

must be able to see it, measure it, and replicate it

experiential reality (criteria for empirical observation)

personally experienced

agreement reality (criteria for empirical observation)

things are considered to be real because everyone agrees they’re real

science (criteria for empirical observation)

median between experiential and agreement realities

-must be logical

-must be empirical

native human inquiry (criteria for empirical observation)

predicting behavior

causal reasoning

hard sciences; future circumstances are caused by present ones (a causes b)

probabilistic reasoning

effects occur more often, but not always (more likely to happen, not 100%)

sources of secondhand knowledge

1. tradition - “things that everyone knows”

   1. authority (consensus) - relying on experts

errors in personal inquiry

mistakes made when finding research

inaccurate observatoin

(error) sloppy observers

-solution: make observation a conscious activity

overgeneralization

(error) only making a few observations and assuming all interviewed people represent the whole population

-leads to selective observation: only paying attention to info that supports hypothesis

-solution: predetermine the # of people that you’re going to interview

made up information

(error) making up what you want to hear

-leads to ex post facto hypothesizing: hypothesis isn’t supported

-when research switches hypothesis, it’s okay

-solution: don’t lie

illogical reasoning

(error)

-ex. gambler’s fallacy

ego involvement in understanding

(error) whenever researchers become so involved, that it becomes personal

-solution: replication

premature closure of inquiry

(error) stop researching too soon

-solution: never stop researching

mystification

(error) the belief that we can’t understand everything

-bad because when something is mystified, research stops

foundations of social science

1. theory - logical aspect of research (what you expect)

2. research method - observational aspect (what you observe)

   1. stats - compare what you observed with what you expected

social regularity

if there is social regularity, there’s predictability and we can study scientifically

cases against studying society

-triviality: why study information we already know

-exceptions: particular cases are not significant overall

-people interfering with research: compensate for this with large sample size

aggregates

people in the same place at the same time

attributes

a description; the specifics or details (69 years old)

variables

logical groupings of attributes (age)

-independent: has an effect on the dependent variable and manipulated in an experiment

-dependent: affected by the independent variable

positive correlation

variables go in the same direction

negative correlation

variables go in opposite directions

null/spurious

no correlation

reasons to research

1. to test theories that are already in existence

2. to explore unstructured interests

3. applied research: when you gather data for a client

   1. involuntary research: research done under external pressure (class projects)

theory

a very large thing that encompasses a lot of information

hypothesis

a part of a theory that is removed and subjected to measure

operationalization

game plan; how do you plan to test this

*must operationalize variables

deductive logic

system of logic;

-you go from the general to the particular

-you go from theory to observation

-you reason toward observation

-it tests theories

inductive logic

system of logic;

-you go from the particular to the general

-you go from observation to theory

-you reason from observation

-you develop theories

subjectivity

when you allow your personal feelings to get involved

*can’t do

objectivity

you do not allow personal feelings to get involved

*trying to reach

law

all cultures have it

sociology paradigm

structural-functional: society is structured and everything has a function

conflict: Marx; money and power

symbolic interaction: we create reality and reality is created for us

psychology paradigm

psychoanalytic: Freud; you do not have free will, you are controlled by your unconscious

-human nature is bad

behaviorism: you do not have free will, your environment controls you

-human behavior is neutral

humanism: you do have free will

-you’re a good person, and when left alone you’ll do great things

units of analysis

things we study;

1. individuals

2. groups - results from individuals averaged together into one measure

3. organization - businesses

   1. social artifacts - inanimate objects

ways of describing units of analysis

characteristics, orientations, actions

errors in mislabeling units of analysis

ecological fallacy: you say it’s individuals, but it’s actually groups (all you’ve looked at are groups) ex. everyone who wears glasses have a high IQ

reductionism: you limit yourself to only one range of causes

cross-sectional study

you study different groups of people at the same time

-advantages: quick and dirty research

-disadvantages: don’t control for cohort (time period/age) differences

longitudinal study

take place over time

trend study

type of longitudinal study; study changes with a general population over time

cohort study

type of longitudinal study; studies different age groups

panel study

type of longitudinal study; uses the same people over and over again

cross-sequential study

same people at different times

panel attrition

panel and cross sequential studies: subjects become test why’s (same inventory results in participants knowing what the study wants) and people either drop out or die

Kaplan-classes of things that scientists measure: direct observables

things we observe directly

Kaplan-classes of things that scientists measure: indirect observables

a check in a box in a questionnaire (or telling in an interview)

Kaplan-classes of things that scientists measure: constructs

*not on test

reliability

consistency

how to increase reliability with subjects

-ask only what people know

-ask about relevant things

-ask clearly

methods to increase reliability

-test retest method: make same measurement twice

-established measures

-research worker reliability: want your research workers doing the exact same thing

validity

quality of measurement

internal validity

measuring what you want to measure

external validity

how much do your results apply to the outside world