Unit 3 Notes

Question 1

What is the goal of blocking in experimental design?

Answer 1

Reduce/control variability

Example: Create 4 groups based on grade level.

Question 2

What does voluntary response bias give too much emphasis too?

Answer 2

people with strong opinions, especially negative opinions.

Example: Divide population based on gender.

Question 3

What is critical in selecting an unbiased sample?

Answer 3

use of randomization

Example: Stratify by age, gender, ethnicity.

Question 4

True or false: Accuracy refers to how close a set of measured values are to the true value.

Answer 4

TRUE

Accuracy is about hitting the target.

Question 5

What does precision measure in data collection?

Answer 5

Measure of variability

Precision is about consistently hitting the same location on the target.

Question 6

What is critical in being able to generalize a sample to the population?

Answer 6

use of randomization in selecting a sample

In the US, this occurs every 10 years.

Question 7

What is the difference between stratified sampling and cluster sampling?

Answer 7

• Stratified Sampling: SRS within each group
• Cluster Sampling: SRS of groups

Stratified involves SRS within each group, while cluster involves SRS of groups.

Question 8

What is the purpose of a control group in an experiment?

Answer 8

Needed for comparison to treatment group

Helps to measure the effect of the treatment.

Question 9

When are differences in observed responses statistically significant?

Answer 9

when it’s not likely that they can be explained by chance variation

Ensures groups are similar other than treatment.

Question 10

Random assignment of subjects to treatment groups is important in handling what?

Answer 10

unknown and uncontrollable differences.

One subject receives one treatment, the other receives the second.

Question 11

What is sampling variability?

Answer 11

Unavoidable error when taking a sample of a population

Describes the probability of how likely we are to have a certain error size.

Question 12

What is the difference between random assignment and random sampling?

Answer 12

random assignment refers to what’s done with subjects after they’ve been selected for a study while random sampling has to do with how subjects are selected for a study.

Important to identify in studies.

Question 13

What is the difference between an observational study and an experiment?

Answer 13

• Observational Study: Measures variables without influencing them
• Experiment: Applies treatments to determine effects

Experiments establish cause-and-effect relationships.

Question 14

Randomized block design

Answer 14

experiments in which randomization only occurs within blocks.

This can lead to unrepresentative samples.

Question 15

Randomized paired comparison design

Answer 15

experiments where subjects are paired and randomization is used to decide who in each pair receives which treatment.

Goal is to remove bias and control variables.

Question 16

What is the purpose of replication in experiments?

Answer 16

To redo the same experiment with more than one experimental unit and get similar results

Enhances reliability of findings.

Question 17

What’s the goal of blinding and double blinding?

Answer 17

remove bias and control all variables

Example: Telephone surveys ignore those without access to phones.

Question 18

What is the goal of blocking?

Answer 18

reduces/controls variability

Results in low response rates.

Question 19

What does stratification reduce?

Answer 19

variability

Happens after choosing a sample.

Question 20

What does it mean when there’s a low accuracy?

Answer 20

it means there’s bias in the selection method

Subjects are told that the treatment works.

Question 21

What is the significance of sample size in studies?

Answer 21

Larger samples lead to more accurate reflections of the population

Important for generalizability.

Question 22

What is proportional sampling?

Answer 22

Sizes of random samples depend on the proportion of the total population represented

Example: Surveying students based on demographics.

Question 23

What does randomization in assignment to treatment groups increase?

Answer 23

the chance that the treatment groups are as similar as possible other than which treatment each group receives.

Question 24

Advantages and Disadvantages of Retrospective Studies

Answer 24

Advantages: Smaller scale, quicker, less expensive

Disadvantages: Researchers rely on previous recordkeepers, inaccurate
accounts, biases

Question 25

Advantages and Disadvantages of Prospective Studies

Answer 25

Advantages: Provide more accurate data, less susceptible to recall error from subjects. Disadvantages: Expensive, time consuming, requires following many subjects over a long period of time.

Question 26

Sampling frame

Answer 26

lists all the members of your target population

Question 27

Simple Random Sample (SRS) Advantages

Answer 27

its simple and makes it easy to interpret the data collected, requires less prior knowledge of the population other than knowing the sampling frame, allows us to make generalizations from the sample to the population, and is unbiased since it gives statistics that are centered around the true parameter value when repeated many times.

Question 28

SRS Disadvantages

Answer 28

needs a list for all subjects, can be difficult to execute if population is large, time-consuming, and important groups may be left out from the sample.

Question 29

Stratified random sample advantages

Answer 29

makes sure that every stratum is represented, collecting data is more manageable by looking at subgroups, cost-effective, have reduced variability making the estimates more precise, more information, insights, and differences can become more apparent among groups.

Question 30

STRS disadvantages

Answer 30

prior knowledge about the population and variables may not be available or accurate, may be difficult to classify every member of the population into a stratum, should know the size of each stratum, evaluating the results may be more difficult, hard with larger populations, planning is needed to ensure that subdivisions are logical in the context of a study.

Question 31

Cluster sampling advantages

Answer 31

clusters are chosen for convenience, allows for a larger sampling size, appropriate when the construction of a complete list is difficult or expensive to organize.

Question 32

Cluster sampling disadvantages

Answer 32

difficult to mirror the full diversity of the population, analysis can be difficult if the clusters are different size, provides less precision with a given sample size, could be biased if the population doesn't have natural clusters or clusters that are representative of the population.

Question 33

systematic sampling advantages

Answer 33

simpler and quicker, practical and reasonably random, can end up with a stratified sample if similar members in the list are grouped together.

Question 34

systematic sampling disadvantages

Answer 34

randomization is weak, determining the fixed interval for sample selection can lead to inaccurate results if the population size isn't known, if it has a periodic structure, bias can occur.