Sampling

•      Population-->

•      Sampling frame-->

•      Sample

 

Probability sampling

•      Simple random

•      Stratified random

•      Cluster

 

Nonprobability sampling

•      Convenience

•      Quota

•      Purposive

 

Types of surveys

•      Cross-sectional

•      Longitudinal

–   Trend

–   Panel

•      Cross-sequential

 

Advantages

•      Large samples

•      Non-observable behavior

•      Efficient

 

Disadvantages

•      Not causal

•      Reactivity

•      Response styles (social desirability, response acquiescence, response deviance)

•      Volunteer problem

 

Correlational study

•      Correlation coefficient

•      r

•      From -1 to +1

•      Magnitude vs. sign vs. significance

•      Positive/negative correlations


Examples

“High” Correlations

•      IQ test reliability     .93                    

•      IQ and first grade grades      .85-.90

•      traffic fatalities and indices of progress in 3rd world countries            -.72

 

Examples

“Moderate” Correlations

•      IQ in high school and college grades   .50-.55

•      Physical similarity of spouses     .40             

•      IQ of fraternal twins     .54

•      Aversive maternal and child behaviors  .55

 

Examples

“Low” correlations

•      Reading achievement and TV viewing                       -.05

•      GRE and grad school grades            .17

 

Examples--table

 

Third variable problem

•      The more ice cream being bought in a region, the greater the number of drownings.

•      The more active children are in Boy/Girl Scouts, the less likely they are to commit a street crime.

•      People who eat breakfast tend to live longer than people who skip breakfast.

 

Third variable problem

•      The more time fathers spend with their children, the less likely they are to sexually abuse them.

•      The larger a woman’s feet, the less trouble she has in childbirth.

Scatterplots

•      aka scatter diagrams

 

Three things that can bias correlations

•      Truncated range

•      Outliers

•      Small sample size

 

“Causal” relationships

•      Cross-lagged correlations

•      Path analysis

•      Structural equation modelling

 

Correlations: 4 sets of 3

•      Three parts to a correlation: sign, magnitude, significance

•      Show r’s: raw data, correlation coefficient, scatterplots

•      3rd variable problem

•      Possible biases: truncated range, outliers, small sample size