1. Drawing Statistical Conclusions

Question 1

Difference between Randomized Experiments and Observational Studies in Statistics

Answer 1

In Randomized Experiments, we randomly assign subjects to groups and then compare the outcomes. This method helps us infer cause-and-effect relationships.

Observational Studies, on the other hand, involve observing subjects without influencing them. These are less reliable for inferring cause-and-effect because of potential unseen factors.

Question 2

What is a Confounding Variable in Statistics?

Answer 2

A Confounding Variable is something that might affect the outcome of a study but is not the main focus of the study.

It’s related to both the group being studied and the result you’re looking at. It can make it hard to tell if the outcome is really caused by what you’re studying or something else.

Question 3

Do Observational Studies Have Value in Statistics?

Answer 3

Yes! Even though Observational Studies can’t definitively prove cause-and-effect, they are still valuable. They can suggest trends, help form hypotheses, and are sometimes the only option for ethical reasons.

Question 4

What Role Does Observational Data Play in Statistics?

Answer 4

Observational data can help in establishing causation indirectly, especially when randomized experiments are not possible due to ethical reasons. They can provide evidence towards causal theories and suggest directions for future research.

Question 5

How Can Observational Studies Claim Cause and Effect?

Answer 5

Through consistency across different populations and times, changes in response with different levels of the explanatory variable, and a logical reason explaining the cause-and-effect relationship.

Question 6

What is Statistical Sampling?

Answer 6

It’s like picking a few apples from a big tree to judge the quality of all apples on the tree. You’re taking a small group (sample) from a larger group (population) to learn about the whole group.

Question 7

Difference between Population and Sample in statistics?

Answer 7

Population is the entire group you’re interested in, like all the fish in a lake. Sample is a smaller group you actually study, like catching a few fish to study.

Question 8

What are Parameter and Statistic in statistics?

Answer 8

A Parameter is a number that describes a feature of the entire population (like the average weight of all fish in the lake). A Statistic is a number that describes a feature of your sample (like the average weight of the fish you caught).

Question 9

What are the differences in sample designs?

Answer 9

Imagine studying birds vs. kids. Studying birds in a controlled setup (randomized experiment) lets you make cause-effect conclusions. Observing kids in their natural setting (observational study) might include other influences.

Question 10

Why is randomization important in sampling?

Answer 10

Randomization in sampling is like drawing names from a hat. It makes sure everyone has an equal chance to be chosen, making the sample unbiased

Question 11

What are different types of random samples?

Answer 11

Simple random sample is like picking names blindly from a list. Stratified random sample is like dividing the list into groups (like age groups) and then picking names from each group.

Question 12

What makes a sample representative?

Answer 12

A representative sample accurately reflects the whole population. It’s like making sure the few apples you pick truly represent all apples on the tree.

Question 13

What are the differences between an observational study and an experiment, in terms of structure of the study and interpretation of the results?

Answer 13

In an observational study, researchers watch and record the variables they are interested in without interfering. This approach is hands-off. In an experiment, researchers apply a specific treatment to the variables they are interested in to see how they react. This method is hands-on.

Question 14

What explanation is the purpose of randomization of subjects to treatment groups in an experiment?

Answer 14

To distribute covariates (characteristics of subjects not
necessarily measured) equally across the treatment groups

Question 15

Where do we start with determining whether a study is appropriate?

Answer 15

An institutional review board.

Question 16

Explain the difference between confidentiality and anonymity.

Answer 16

Confidentiality involves keeping participant identities private from the public, while researchers still know who they are.

Anonymity is when the researchers do not know the participants’ identities.

Question 17

What is a clinical trial?

Answer 17

A clinical trial is a study where researchers test new medical treatments, like medicine, diets, or medical tools, etc, to see if they are safe and effective for people.

Question 18

How do we test whether d (the treatment effect) is O?

Answer 18

Set up a formal hypothesis test where the hull is d = O and the alternative is al
not equal to 0.

Question 19

Definition of the p-value

Answer 19

The probably of observing by random chance a result as extreme or more extreme than was what was actually observed under the assumption that the null hypothesis is true.

Question 20

The null hypothesis always has….

Answer 20

The assumption and the equal sign

Question 21

How should we interpret the p-value?

Answer 21

A sufficiently small p-value is evidence that the null hypothesis is false because it reflects the probability that what we observed would happen very rarely if the assumption (the null
hypothesis was true.

Question 22

Convenience Sample(good or bad sample method)

Answer 22

(Bad)choosing subjects who are around or those that are convenient to talk to.

→ Sampling the Fist 30 people that enter the library and ask then about their study times

→ sampling people you see in the mal abut shopping habits

Question 23

Voluntary Response Sample(good or bad sampling method)

Answer 23

(Bad)
People who chose themeselves responding to a general appeal
→ Wite in or callin opinion polls

Question 24

Why are these sampling methods bad?

Answer 24

Because they create bias

Question 25

What’ is bias?

Answer 25

-prejudice in favor of or against one thing, person group, ete. The design of a staistical study ill show bias if it stematicaly favors certair outermes lie, if it consistently overestimates or underestimates

Question 26

Voluntary Respnse Bias

Answer 26

people are eager to volunteer when they have a strong opinion on the matter

Question 27

Response Bias

Answer 27

anything in the survey that influence responses → respondents lying, trying to please the interviewer, unwillignes to reveal facts, leading or confusing questions

Question 28

Nonresponse Bias

Answer 28

when a large amount of those samples do not respond
→ you were selected, but you chose not to respond/partiapate (didn’t answer the door or phone or mail back questionnaire)

Question 29

Under-coverage bias

Answer 29

Sampling in a way that leaves out a certain portion of The population that should be in the sample
→ telephone polls, registered voter lists, etc.

Question 30

Simple Random Sample (SRS)-

Answer 30

every experimental unit and every possible sample has the same chance. of being picked

→ Flipping coin, roling die, Calcuate random generators, names out of hat, Table of kardon Digits

Question 31

Stratified Randon Sample

Answer 31

divides the population into Smaller groups (Strata) with something in common and then apply a SRS to each strata
→ gender, grade, age, ete.

Question 32

Systematic Random Sample

Answer 32

sample selected at a randon starting point but with a fixed periodic interval

→ sampling a large neighborhood. Start by picking a house at random and the survey every 10th house from there.
→ Selecting every 5th person that walks though the door

Question 33

Cluster Sampling

Answer 33

popuation is brokendown into groups, then All members of one or more graps are taken as the sample

→ Sample all students in a grade
→ Choose a bus at random and survey everyone on the bus

Question 34

Placebo

Answer 34

A Placebo is a substance that is made to resemble a drug (medicine) but does not contain an active drug, and are used as a control in drug testing.

Question 35

response variable

Answer 35

A response variable measures on outcome of a study

Question 36

Explantory Variable

Answer 36

Helps to explain or influences changes in a response variable

Question 37

Lurking variable

Answer 37

A Lurking variable is one that is not expanatory or response, but can influence how we interpret the relationship between them

Ex. Should ice cream be blamed for murder.Studies shown that when ice cream sales increases, murder rate of homicide increases. However the lurking variable is the Sun because when it’s hot weather, sales go up and murder rate increases.

Question 38

Confounding

Answer 38

confusing one variable (thing) for the cause of something happening. In other words, the response is partially do to a third factor or variable.

Question 39

What are the three types of confounding?

Answer 39

1.) Confounding variable

2.) Placebo effect

3.) Blinding (single blind or double blind)

Question 40

Confounding variable

Answer 40

3rd variable that we know about that could cause the response.

Question 41

Placebo Effect

Answer 41

an effect that occurs from a fake treatment because the patient believes that the effect should occur

Question 42

Blinding (single Blind or Double Blind)

Answer 42

Single-blind: the subjets are unaware of which treatment they are receiving but the researcher knows.

Double-blind - neither the subjects nor the researcher knows the treatment a subject received