SRM: Unit 1-5 Notes

Key terms:

• Conceptualization: A researcher’s clarification of how they are defining a concept 

• Operationalization - Turning a conceptualization into a measure using specific, observable indicators 

• Valid - Measure has to actually measure the concept

• Reliable - Measure has to work every time, even in different contexts 

• Explanatory variable: The supposed CAUSE

• Outcome variable: The supposed EFFECT 

• Confounding variable 

General notes:

• Consider what questions are even being asked and by whom

• Key parts of a project’s validity

  • Measurement (internal validity)

    • Operationalization methods

      • Ethnography

      • Surveys 

      • Interviews 

    • Four types of measurement validity 

      • Face - measure seems reasonable/makes sense 

      • Content - all possible answers are included in measure 

      • Criterion - answers are comparable to answers that would be received from the ideal possible question (gold standard) 

      • Construct - measure is supported by other measures for the same concept

    • Two ways to test reliability

      • Test-retest reliability - asks the same question at two different points in time

      • Alternate forms - Asks the same question two different ways 

  • Generalizability (external validity)

    • Possible bias in sampling

      • Selection bias - not everyone has an equal chance of being selected for the sample 

      • Non-response bias - cases that are selected in the sample are not actually included in the sample 

  • Causality (Causal validity) - explanatory research questions 

    • Are X and Y correlated?

      • Types of causal explanations

        • Idiographic 

          • Focus on a single event and single set of people

          • Personal troubles 

        • Nomothetic 

          • Focus on patterns of events that happen to patterns of people 

            • Public Issues 

    • Time/temporal order - does X happen before Y

    • Non-spuriousness - there’s not something causing both X and Y to change at once

      • X doesn’t have to be the only cause of Y 

        • Only spurious if X isn’t the real cause of Y 

    • Mechanism - why could X cause Y?