7 - Nature of quantitive research

features of quan research

focus on numbers

• deductive view

• positivism

• objectivism

• epistemological position

• ontological position

ontological position

belief about the nature of reality + social entities

more than presence of numbers

main methods of data collection

questionnaire + surveys
quan context analysis
secondary daya anlysis
structured obs

can include multi method research

imp of tech in surveys

collection of large amount of data

self completion questionnaires

emergence of big data

main preoccupations of quan data

measurement

causality

generalization

replication

causality

why things are the way they are

causes of variation

imp of causal inference = experimental over cross sectional

generalisation

sufficient external validity

• probability sampling

replication

goal of minimizing biasis = replication

explicit about procedure = high validity

operalization

selection of research design

hypothesis

informed specilation based on knowledge of existing research literature

null hypothesis

predicts there will be no change

alternative hypothesis

there will be an effect

process of quan research

1. theory

2. research q

3. hypothesis

4. research design

5. device measures of concepts

6. select research site

7. select research subjects / respondents

8. administer research instruments

9. process data

10. analyse data

11. findings / conc

12. write up

presence of both idn + deductive

translating concepts into measurments

concept

building block of theory + point around which social research is conducted

• conc on common features of elements of social world

main reasons for the focus on measurement in quan research

allow us to identify fine diff between ppl’s charac

consistent device for identifying such diff - reliability

provides the basis for more precise extimates of the degree of relationship

how do we measure

devising indicators

multiple-ideicator measures

• likert scale

Indicators

for concepts that are less directly quan

more ambiguous than measures

reasons for using multiple-idicator measures

single might incorrectly classify ind

single q might be too general / limited in scope

multiple q allow researcher to make mush finer distinctions

Dimensions of concepts

concepts as made up of dif dimensions

need to measure all - multi-dimentional ‘profile’

reliability

consistency of measurement

ex. stepping on a scale

• stability

• internal reliability

• inter-rater reliability

inter-rater reliability

degree to which ind agree about a coding of an item

stability - test

test-re-test

internal reliability

applies to multiple indicator measutes

indicators might not relate to the same thing

validity

measure really measures that concept

face validity

measure reflects the content of the concept explored

concurrent validity

overlapping + predictive validity

construct validity

evaluates a measure bu how well the measure conforms to theorectical explenation

convergent validity

comparing a measure to a dif measure

critique of quan research

treatin social world as nature

• interpretivism

artificial measuring process

• low connection to everyday life

static view of social life

• constructivism

feministic critique

practice of quan - gap in the ideal typycal appraoch

reverse operationism

• deductive view - neglects inductive

reliability and validity testing

sampling