Please enable JavaScript.
Coggle requires JavaScript to display documents.
Research Methods (Overview (Research process (1) Define topic, 2)…
Research Methods
Overview
Social science
Knowledge acquisition
Non-scientific
-
Method of authority: consulting internet, books or experts
-
-
Method of tenacity: holding on to ideas/beliefs due to it being long accepted as facts or habit/superstition
Scientific
-
-
-
Empirical (answers observable/manageable questions, confined by time/resources/instruments)
Status
-
Support
-
-
Scientific method: consists of standards/procedures allowing repetition and verification of research results
-
Theory
-
-
-
-
Structures how findings expressed for purposes of generalization (i.e. explanation of various findings already available)
-
Variety
Approach
-
Interpretive
Aim/Goal
Explore social reality as independent research topic/further research direction: to find out if something exists/occurs
-
-
-
Type
Qualitative
Meaning of behaviour (how individuals perceive their world), reject natural science, natural setting, words, inductive, describes development of process/identifies cultural patterns
Quantitative
Objective reality (external behaviour), natural science model, control/artificial setting, numeric, deductive, cause-and-effect, universal laws
-
Operation
-
Measurement & Instrument
Measurement
Level
Nominal
Category/Classification/Membership: named, distinct/mutually exclusive/exhaustive (e.g. gender, ethnicity)
Ordinal
Order: ranking, comparison (e.g. educational qualification, 1st/2nd/3rd)
Interval
Numerical distance between categories: difference (e.g. temperature in Celsius, Cartesian coordinates, date from arbituary epoch such as AD)
Ratio
Absolute zero: magnitude (e.g. mass, duration, length, age)
Reliability
Test-retest: compare scores from 2 successive measurements of same individuals and calculate correlation between 2 sets of scores
-
-
Split half: split items on questionnaire/test into half, compute separate score for each half, then calculate degree of consistency for a group of participants
-
-
Sampling design
Technique
-
Non-probability: control/flexible, no randomisation (sampling error cannot be estimated)
-
Quota: similar to stratified, but set numbers to be selected according to specific criteria
-
Snowball: recommendation by other people, for hard-to-acess respondents (e.g. homosexual, prostitute, battered wife)
-