SOC 101 Lecture Notes - Lecture 9: Methodology, Statistical Significance, Operationalization
Sociology 101
Introduction to Sociology
Halabi
GCC
Four Main Methodological Traditions in the social sciences:
1. Statistical
2. Ethnographic/Qualitative
3. Comparative-Historical,
4. Experimental
1. Statistical Methods
- heavily used by functionalists.
- preferred positivist methods need quantitative method, thus stats.
- Clear correlation b/w functionalism positivism statistical methods.
Correlation: Statistical methods are most concerned with establishing relationships between
variables
● i.e. T test, regression.
● Correlation =/= causation.
I.V: race/ethnicity
- age
- immigration status
DV: racial profiling
Try to see if all those variables cause the DV.
● Maybe just immigration status has a correlation (then ethnicity has no statistical
significance)
Assumption: if variables are related, there is a good possibility that one variable affects
another
Operationalization: numerical measurement of social phenomena is necessary for statistics
● Must turn concept TESTABLE.
● i.e. Fertility Rate: number of births per 1,000 people
● does level of industrialization effect fertility rate? Industrialization: more difficult to
operationalize
● per capita GDP, electricity use per capita, percentage of the workforce working in
factories
Benefits of Statistics
- There are clear methodological rules that anyone can follow
- Statistics provide a means of testing competing theories
- estimates of causal impact
- insight into generalizability
Document Summary
Four main methodological traditions in the social sciences: statistical, ethnographic/qualitative, comparative-historical, experimental, statistical methods. Preferred positivist methods need quantitative method, thus stats. Clear correlation b/w functionalism positivism statistical methods. Correlation: statistical methods are most concerned with establishing relationships between variables i. e. t test, regression. Try to see if all those variables cause the dv. Maybe just immigration status has a correlation (then ethnicity has no statistical significance) Assumption: if variables are related, there is a good possibility that one variable affects another. Operationalization: numerical measurement of social phenomena is necessary for statistics. Industrialization: more difficult to i. e. fertility rate: number of births per 1,000 people operationalize. Per capita gdp, electricity use per capita, percentage of the workforce working in factories. There are clear methodological rules that anyone can follow. Statistics provide a means of testing competing theories estimates of causal impact insight into generalizability. Reliability: data on fertility rate commonly of poor quality.