Q1 - Collecting Data

Parameter

describes a population

• Ex. the mean age of a population

Statistics

refers to a sample

• the part of a sample that has blue eyes

Collecting data

finding population parameters/characteristics

Observational studies

studies where we do not interfere; gathering data from interviews, surveys, etc

• observational studies are not that good because you can’t survey EVERYONE

  • so, take a sample

Representative sample

a small amount of the population that is similar to the population; has similar traits that we are measuring

What can biased sampling methods do to an estimate?

They can cause us to overestimate or underestimate the result.

What is the best way to reduce bias in a sampling method?

randomization

Voluntary response

a sampling method where indivs self-select/willingly take part in a survey

• not rep of the pop

• potential bias and ppl more interested in the survey are more likely to participate

Convenient sampling

where people are selected based on how convenient they are for the researcher

• could be chosen b/c of geographical location

• biased b/c not rep of pop

Simple random sampling (SRS)

where any sample of a size n is chosen and every individual in the population has an equal chance of getting selected for a sample

Individuals

what is being sampled (people, places, things, etc)

Variables

what is being described/measured

• can be quantitative or categorical

Categorical variables

labels individuals into categories

Quantitative variable

a variable that takes # values that are counts/measurements

ex. years, grade lvl, phone #, etc

Discrete variables

a type of quantitative variable with a fixed set of possible values w/ gaps in btwn

Continuous variable

measured with an integer value on a # line, can be any number

Distribution of a value

what values the variable takes and how often it has these values (shape, center, distribution)

what are the relationships btwn multiple values

Stratified Random Sampling

selects a sample by choosing an SRS from each stratum and combining all the SRS’s into one

• the strata are often HOMOLOGOUS; basically the different groups are the same within themselves but compared to each other they are different

  • kind of like political parties; they have the same belief in one party that differs from a diff party

• works best when a group of indivs is similar to each other but they are much different from another group

Strata

a group of indivs in a population who SHARE TRAITS (remember: homologous) that are associated with the variables in the study

Cluster sampling

where we randomly choose clusters and include each member of a selected cluster into the study

• HETEROGENEOUS; each group contains people/indivs with different traits.

  • kind of like where in 1 homeroom there are people from many different grades

Cluster

a group located near each other, not grouped together because of their shared traits

• THEY HAVE DIFFERENCES, THEY JUST ENDED UP BEING NEXT TO EACH OTHER

Systemic random sampling

where we select a sample from an ordered arrangement of a population by randomly selecting one of the first “k” indivs

• like every 5th person who walks into a door gets counted in the survey