What is data? Flashcards
How can we measure difficult to define concepts in IR?
Concept - Operational def - Indicator - Measurement
Once a concept has been identified we can begin working on an operational definition, moving from the abstract to the specific. Indicators of the operational definition can then be established - can be measured
Why is measurement important?
Important to prove/test the operational definition that has been decided on
What is content validity?
Examines the extent to which the indicator covers the full range of the concept.
What is construct validity?
Examines how well the measure conforms to out theoretical expectations by examining the extent that it relates to other theoretical relevant factors.
What is intercoder reliability?
Reveals the extent to which different coders come to the same coding decisions.
What is the test- retest method - in terms of reliability?
Applying the same test to the same observations at different times and comparing the results.
Steps for choosing cases.
1) Define your population of interest.
2) Outline clear rationale for deciding which particular case or cases you are going to look at - should be related to the theories or ideas you want to look at.
3) Pick the cases you want to look at according to your previously decided selection criteria.
What is quantitative data?
Research based on statistical analysis of carefully coded information for many cases or observations.
What is qualitative data?
Research that tends to be based on discursive analysis of more losley coded information for just a few cases.
What are the levels of measurement?
- Nominal: Numbers are assigned to a label/name - not literally numerical
- Ordinal: Numbers are assigned and do have meaning - levels of political interest 1-5.
- Interval/scale: No distinction between value and label e.g Age
How would you asses the validity and reliability of secondary data?
Does it contain errors, biases, exclusions or exaggerations?
Investigate the biases of the org/govt.
Check what practices and procedures were used to collect the data as well as how it is classified.
What is distinctive about big data?
Big data is essentially a huge N-sample collected from a very large quantity of people.
Distinctive because it offers a new wider perspective while being cheaper that other methods of data collection - public record.
Can present problems with reliability, validity and bias
Should data be made publicly available?
Makes research more transparent -
Verification, practice of making data available for secondary analysis so other scholars can re-analyse the data.
Replication: Carrying out a similar data collection exercise to see if the findings from the original study can be reproduced on an independent data set.
