Sampling
the entire group of individuals we want information about
population
Observational study or experiment: Ninety residents of a retirement community were selected at random. Each member of the sample was asked a number of questions, including questions about exercise and blood pressure.
observational study
Some members of the population are less likely to be chosen or cannot be chosen for the sample
undercoverage
provides a baseline to compare the treatment against
control group
selecting individuals who are easy to reach
Convenience sampling
Give an example of a non-statistical question
How much money does Mr. Tagawa have in his wallet right now?
When people can’t be contacted or refuse to answer
nonresponse
changes are seen from a fake treatment, usually a psychological effect
placebo effect
only surveying people who chose to be in the sample
voluntary sampling
What type of sampling: From the population of all MHS students, randomly select 10 homerooms and survey everyone in those homerooms
cluster sampling
Example: A non-anonymous survey about cheating on tests might lead students to lie
response bias
the patient doesn't know if they received a real treatment
single-blind study
What type of sampling is this: From the population of all MHS students, randomly select 10 freshman, 10 sophomores, 10 juniors, and 10 seniors.
stratified sampling
What type of sampling: From the population of all MHS students, pick every 3rd student from an alphabetized list of all the students.
systematic random sampling
A study on teen social media uses analyzed data from 11,872 randomly selected adolescents (aged 13-15) in the United Kingdom. In the sample, 20.8% reported using social media 5 or more hours per day, on average. What is the largest population to which we can generalize this result?
All adolescents in the U.K. that are 13-15 years old.
outside factors that might be the real reason for a certain observation
confounding variables
How can you decrease variability in your samples?
increase the sample size
What is the goal of doing an experiment?
to understand the cause and effect relationship between two variables
The National Health and Nutrition Examination Survey is a long-term research program that uses random sampling methods to examine the health and nutrition among adults and children in the United States. One study of 2437 young adults (aged 18-25 years) found that the young adults who reported eating added sugars more than 5 times a week had a 73% higher risk of developing gum disease than those who reported eating no added sugars. Based on the study, is it reasonable to say that eating foods with added sugars causes gum problems? Explain why or why not.
No. The researchers didn't randomly assign young adults from the US to eat added sugars a specific number of times per week, so it is not reasonable to say that eating food with added sugars causes gum problems.
What is the purpose of using random assignment?
To create groups that are roughly equivalent