Video Notes Review: Sampling, Bias, Randomization, and Experimental Design (Fill-in-the-Blank)

Fill-in-the-blank flashcards covering key concepts from the notes on sampling, bias, randomization, stratification, blocking, blinding, replication, and experimental design.

Confounding Variables can cause __.

different measures

We use sample statistics to estimate __.

population parameters

We need samples that fairly represent __.

populations

A good practice is to ask: How was data collected? Is it representative of the __?

population

A bad sample can affect our __.

measurement

Randomization increases the chances of wider representation and effectiveness of __.

experiments

Larger samples give better __.

information

Observational studies can be retro (past) or pro (future) __.

perspective

Prospective studies are greater in __ and control but take longer.

accuracy and control

Be careful when naming a bias. You are better off describing the flaw than incorrectly __ the bias.

naming

Undercoverage bias occurs during sampling while __ bias occurs after.

nonresponse

Increasing the sample size will not reduce bias, just __.

error

To accurately use a simple random sample you must link numbers to __.

individuals

Simple random sampling provides equal likelihood of being __.

selected

To be in a stratum, individuals must have a common __.

attribute

You must be able to explain why __ is the best method.

stratification

A single __ should be able to represent the entire population.

cluster

Randomization occurs when choosing from each stratum, or choosing which __.

cluster

Stratifying is more precise than clustering, but takes longer and is only sometimes __.

needed

The larger the group, the more preferred __ is over stratifying.

cluster

If a list is structured, you cannot use __ sampling.

systematic

Accuracy is being __ while precision is being more consistent.

correct

Experiments __ while observations do not.

influence

Observations can only prove __ while experiments can suggest cause and effect.

associations

Random sampling is important for generalizing observations for and reducing __ in experiments.

confounding

Observations use __ while experiments use blocks.

strata

Any experimental design question needs the __ variable identified immediately.

response

You should always be asking if there is a possibility for cause and effect between the two __.

variables

Explanatory (factors) affect __ variables.

response

Confounding variables can affect both __.

explanatory and response variables

Random assignments should create groups that are all the same other than the __ imposed.

treatment

Randomization cannot eliminate confounding variables, only __ them.

reduce

You must specify how you are randomly __ groups in an experiment and which is receiving the treatment.

assigning

In blinding, both groups experience the __ effect.

placebo

To double blind, the evaluator must believe both groups have the same __ or lack thereof.

treatment

Completely randomized designs give everyone an equal chance of being assigned to treatment or control regardless of __.

characteristics

Randomized block designs only have randomization occurring after __.

blocking

Randomized paired design assigns two people together and only one is randomly given the __.

treatment

Be careful not to mix up __ and blocking.

stratification

Stratifying splits into groups and samples some from __.

each

Blocking splits into groups and assigns different treatments to __.

each group

Replication increases __.

validity

Do not refer to random sampling when doing an experiment when you mean __.

random assignment

Assignment determines if selected individuals are treatment or __.

control