1.1 Samples and Surveys

Methods of Collecting Data

Observational Studies + Experiments

Difference between Observational Study + Experiment

Experiment intervenes and is more powerful (ex: give certain drug to certain people/ has someone running for certain amount of time)

What is bias?

systematic favoritism (systematic: happens a lot) (favoritism: happens in certain direction)

What is the impact of bias

reduces credibility of results

Two challenges of getting good survey results

select good sample, collect good data

What are all good samples

Random - everyone has same chance and no favoritism

Types of random samples

simple R.S. and Stratified R.S. (and more)

Simple Random Sample

Examine entire population as it exists

Stratified Random Sample

Compare subgroups of population equally

2 Criteria for Biased Samples

sampling procedure must be used + same must represent entire population (aka truly random)

Design

wording of survey question can affect results, type of survey can affect results (ex: using land-line to survey students)

Implementation

timing of survey can affect results, nonresponse - respondents likely have stronger opinions (look for high PERCENTAGE of respondents), response bias - individual doesn’t give correct data (make it anonymous or confidential)