AP Stat - Unit 1: Vocab Review

Observational Study

treatments ARE NOT assigned by the researcher

Experiment

treatments ARE randomly assigned by the researcher

Response Variable

a variable that measure the OUTCOME of a study (must be measurable/ quantifiable)

Confounding Variable

a variable that relates to both the independent and dependent variable

Population

entire group of inviduals we want information about

Census

collect data from every individual in the population

Sample

A subset of inviduals from the popualtion we collect data from

Sampling Frame

the physical list og the population used to select the random sample

Simple Random Sample

Every group of n individuals in the population has an equal chance of being selected as the sample

• How to take an SRS

  • State how you will number the population

  • State the method you will use (random number generator)

  • State the range of numbers used and how many you will select. Also, communicate how no repeated numbers are allowed.

Systematic Random Sample

Select a random starting place on the sampling frame. Then select every kth item.

• k= total/sample needed

Stratified Random Sample

Group individuals so that within a group individuals are similar with a respect to a confounding variable. Then, take a randome sample from each group.

• Take a sample that is porportional to the population size

  • EX: if the group is 50% of the total population, take 50% of the group

Cluster Random Sample

Group individuals located close to one another. Randomly select some clusters as a whole to be the sample. Ignore clusters not chose.

Bias

when a sample is very likely to underestimate or overestimate the population valueyou are trying ot estimate

Selection Bias (Undrcoverage Bias)

occurs when some members of the population are less likely to be chosen or cannot be chosen in a single smaple.

• Sample was poorly selected because :

  • No Randomness Was Used

Convenience Sampling

the sample is selected by what is convenient for the researcher

Voluntary Response Sampling

This is where the individuals choose to respond. The people who respond are often similar to each other with respect to their views/opinions.

Nonresponse Bias

occurs when several individuals are chosen for the sample can’t be contacted or refuse to participate

Response Bias

occurs when there is a systematic pattern of inaccurate answers

Leading Question

People know what answer you want so they tell you that answer

Confusing Questions

People don’t understand the question, so they give an inaccurate answer

Experimental Units/Subjects

The individuals on which the experiment is done:

• Subjects = Humans

• Units = Anything non-human

Explanatory Variable

A variable whose levels are manipulated intentionally (independent variable)

Levels

The subcatergories/values a factor breaks down into

Treatments

All combinations that can be made when each level of a factor is combined with each level of the other factors

What makes a well-designed experiment?

1. Comparison

2. Random Assignment

3. Control

4. Replication

Comparison

Must compare two or more treatment groups

What can one of the groups be for a comparison to occur?

A control/ placebo group

Random Assignment

Experimental Subjects/Units have to be randomly assigned to treatments

Control

Control potential confounding variables by keeping all other variables constant for all groups. Could use blinds to accomplish

Single Blind

Either the subjects or the members of the research team, but not both, don’t know which treatment the sibject is receiving

Double-Blind

Neither the subjects nor the members of the research team who have contact with the subjects kniw which treatment the subject is receiving. One (or more researchers who don’t have contact wih subjects will know so that data can be analyzed)

Replication

Must have more than one experimental unit/subject in each treatment

Completely Randomized Design

• Most basic design

• Take whole group of subjects and use a random method to assign them into treatment groups

How do you make a completely randomized experiment?

• Number the subjects from 1- end of sampling frame

• Use a random number generator to choose n unique random numbers from sampling frame

• Assiign chosen numbers into treatments

• Track data

Randomized Block Design

• separate subjects into groups first

• randomly assign subjects to treatments from those groups

How would you conduct a block design experiment?

• Block the subjects based on the confounding variable (ex. age: 20 -30, 30 - 40)'

• Number the subjects in the first group by sampling frame

• Use a random number generator to generate n uniquerandom numbers from sampling frame

• Assign given number into treatment groups

• Repeat for other group(s)

Mathced Pair Design (Similar Subjects)

• two subjects are paired together such that they are similar with respect to the blocking variable

• one subject receives treatment 1, the other subject receives treatment 2

Matched Pair Design (both - both)

• Treatments are done in random order so that each subject does both treatments

What do observational studies reveal?

Association between variables, not causation.

What do experiments reveal?

Causation between variables, due to controlled conditions.

You can only generalize the findings of a study to the population from which the sample was selected from.

Observational Studies

Results can be generalized to others who are similar to the volunteers

Well-Designed Experiments