📖
Surveys and Experiments
Created @December 3, 2022 5:26 PM
Class MCRS
Type Class
Stage?
Reviewed
Column
Materials
Reading 1
Reading 2
SURVEYS
Meant to capture public opinion at anytime
Formatted, rather than open-ended
Because we can collect data from a LARGE sample easily using surveys, the
POPULATION VALIDITY OF THE DATA INCREASES, HENCE, THE EXTERNAL
VALIDITY INCREASES.
Research Questions for SURVEY DESIGNS:
- “To what extent?”
- Correlational design
CON: WE CANNOT RULE OUT ALTERNATIVE EXPLANATIONS
Surveys and Experiments 1
, TYPES OF SURVEYS:
Cross-sectional Survey Design: Measured at ONE POINT OF TIME
- Not expensive
- Cannot determine causal conclusions
Longitudinal Design: To predict something OVER TIME
- Expensive
- Can determine causal conclusions
{BUT SINCE A SURVEY DOES NOT HAVE A HIGH LEVEL OF INTERNAL
VALIDITY, NONE OF THE DESIGNS CAN HELP DETERMINE CAUSALITY,
THEY CAN ONLY GIVE AN IDEA OF CAUSALITY.}
Types of Longitudinal Designs:
TREND STUDY: To study a trend over time. EX: How people feel about gun
control? (can be measured every 5 years). YOU ARE NOT ADDRESSING
SAME PEOPLE, HOWEVER, SAME POPULATION ADDRESSED.5
COHORT STUDY: Examines changes in a POPULATION across time.
- Data is collected from different individuals within the same
cohort over time.
- Often used in developmental studies
- PRO: Gives more information than a Trend Study
- CON: Difficult to examine cause and effect because same
people are NOT studied repeatedly.
PANEL STUDY: Examines changes in INDIVIDUALS over time.
- Data is collected from SAME individuals from the SAME
population.
- Can be cross-sectional or longitudinal. However, panel
studies are done in a longitudinal manner more often.
CROSS-LAGED PANEL SURVEY: Independent and dependent variables
measured at different points of time.
Surveys and Experiments 2
Surveys and Experiments
Created @December 3, 2022 5:26 PM
Class MCRS
Type Class
Stage?
Reviewed
Column
Materials
Reading 1
Reading 2
SURVEYS
Meant to capture public opinion at anytime
Formatted, rather than open-ended
Because we can collect data from a LARGE sample easily using surveys, the
POPULATION VALIDITY OF THE DATA INCREASES, HENCE, THE EXTERNAL
VALIDITY INCREASES.
Research Questions for SURVEY DESIGNS:
- “To what extent?”
- Correlational design
CON: WE CANNOT RULE OUT ALTERNATIVE EXPLANATIONS
Surveys and Experiments 1
, TYPES OF SURVEYS:
Cross-sectional Survey Design: Measured at ONE POINT OF TIME
- Not expensive
- Cannot determine causal conclusions
Longitudinal Design: To predict something OVER TIME
- Expensive
- Can determine causal conclusions
{BUT SINCE A SURVEY DOES NOT HAVE A HIGH LEVEL OF INTERNAL
VALIDITY, NONE OF THE DESIGNS CAN HELP DETERMINE CAUSALITY,
THEY CAN ONLY GIVE AN IDEA OF CAUSALITY.}
Types of Longitudinal Designs:
TREND STUDY: To study a trend over time. EX: How people feel about gun
control? (can be measured every 5 years). YOU ARE NOT ADDRESSING
SAME PEOPLE, HOWEVER, SAME POPULATION ADDRESSED.5
COHORT STUDY: Examines changes in a POPULATION across time.
- Data is collected from different individuals within the same
cohort over time.
- Often used in developmental studies
- PRO: Gives more information than a Trend Study
- CON: Difficult to examine cause and effect because same
people are NOT studied repeatedly.
PANEL STUDY: Examines changes in INDIVIDUALS over time.
- Data is collected from SAME individuals from the SAME
population.
- Can be cross-sectional or longitudinal. However, panel
studies are done in a longitudinal manner more often.
CROSS-LAGED PANEL SURVEY: Independent and dependent variables
measured at different points of time.
Surveys and Experiments 2
