2 - Sampling Designs Flashcards Preview

Biostats > 2 - Sampling Designs > Flashcards

Flashcards in 2 - Sampling Designs Deck (65):
1

 2 approaches for collecting data: 1. 2.


1. Observational study 2. Experimental study


2

The purpose is to describe and compare existing situations.

Observational study

3

The purpose is to study whether a treatment causes a change in the response.

Experimental study

4

When our goal is to understand cause and effect, ______ are the only source of fully convincing data.

Experiments

5

Explanatory variable

A variable that may explain or cause influences in another variable. AKA independent variable.

6

Explanatory variable AKA ______.

Independent variable

7

Lurking variable

A variable that is not among the explanatory (independent) variable or response (dependent) variable in a study, and yet may influence the interpretation of relationships among these variables.

8

Confounded results

Occur when the effects of an explanatory and lurking variable cannot be distinguished from each other.

9

Why do observational studies often fail to show causality?

Because of confounding (lurking) variables.

10

The individuals in an experiment are often referred to as _____ if they are human.

Subjects

11

The individuals in an experiment are often referred to as _____ if they are not human.

Experimental units

12

The explanatory (independent) variables in an experiment are often referred to as _____.

Factors

13

Any specific experimental condition applied to the subjects is known as ______.

A treatment

14

If an experiment has more than one _____, a _____ is a combination of specific values of each ______.

Treatment, treatment, factor

15

By assigning subjects to _____, we avoid confounding in experiments.

Treatments

16

Another advantage of experiments is that we can study the combined effects of several factors simultaneously, known as ______.

Interaction

17

Interaction

Can produce effects that we were previously unable to predict based at looking at each individual factor alone.

18

How can you tell what the individuals are?

At what level were participants randomized? Those are the individuals.

19

How many treatments are there, and what are they?


Q image thumb

4 treatments



1. Only physical activity intervention

2. Only nutrition intervention

3. Both

4. None


20

Does the control count as a treatment?


Yes


21

What is an uncontrolled experiment?


One that imposes


22

How many factors does this experiment have?


Q image thumb

2, normal/vegetarian diet, or unrestricted/restricted salt.


23

What is being manipulated in an experiment?


Factors/explanatory variables


24

Can it be an experiment if the individuals select their own treatment?


No - treatment must be deliberatedly imposed by the researchers.


25

A sampling survey is an ________ that relies on a ________ sample selected from the entire population at ________.

Observational study; random; a single point in time

26

Sample surveys are sometimes called ________.

Cross-sectional studies

27

What is undercoverage?

Occurs in a sample survey when some groups in the population are left out of the process of choosing the sample

28

Undercoverage in a sample survey is a possible cause of ________.

Bias

29

Two sources of bias in sample surveys:
1.
2.

1. Undercoverage
2. Nonresponse

30

Nonresponse

Occurs when an individual cannot be contacted or refuses to cooperate

31

How does nonresponse in a probability sample differ from opt-in polls that use a voluntary response sample?

Nonresponse: can be improved by increasing the weighting of particular groups

Voluntary response: cannot be improved because we never know what kind of population the respondents represent

32

Define response bias

Type of bias in a sample survey caused by the behavior of the respondent or the interviewer, or the types of questions being asked.

33

A woman understating what she weighs is a form of _______.

Response bias

34

Wording effects

Confusing or leading questions can induce strong bias, Sometimes this bias results from the order that questions are asked.

35

Sample survey

An observational study that relies on a random sample drawn from the entire population at one point in time.

36

A sample survey is sometimes referred to as a _______.

An observational study

37

Nonresponse occurs in this type of sample...

Probability sample

38

Case-control study

An observational study, in which case-subjects are selected based on a defined outcome, and a control group is selected separately to serve as baseline to which the control group is compared.

39

Is a case-control study observational or experimental?

Observational

40

Case-control studies tend to be _______ (retrospective, prospective).

Retrospective

41

Cohort studies tend to be _______ (retrospective, prospective).

Prospective

42

Cohort study

Subjects sharing a common demographic characteristic are enrolled and observed at regular intervals over an extended period of time.

43

Which is more prone to confounds: case-control or cohort studies?

Cohort studies - they start with a homogenous group.

44

Which supports incidence calculations: case-control or cohort studies?

Case-control

45

Placebo

A control treatment that is fake, but otherwise indistinguishable from the experimental group.

46

When is the placebo affect extra obvious?

When the response variable is pain

47

Why is randomization used in experiments?

To reduce bias

48

Most experiments are _______.

Randomized comparative experiments

49

3 principles of experiments:
1.
2.
3.

1. Control the effects of lurking variables on the response, most simply by comparing 2 or more treatments

2. Randomize: use impersonal chance to assign subjects to treatments

3. Use enough subjects in each group to reduce chance variation in the results

50

Completely randomized design

An experimental design where all the individuals are allocated at random among all the treatments. It is not necessary for the same number of individuals to be assigned to each treatment.

-Can compare any number of treatments

-Can have multiple factors

51

Simplest statistical design for experiments

Completely randomized design

52

Block

A group of individuals that are known to be similar in some way that is expected to affect the response to treatments.

53

Block design

The random assignment of individuals is carried out separately within each block.

54

Blocks are another form of ______.

Control.

55

If you separate people exposed to smoke, from people who are not exposed to smoke, you are using _______.

Block design

56

Matched pairs design

An experimental design, which compares exactly TWO treatments, either by using a series of individuals, or by using each individual twice. (i.e. giving a person two cuts, treating one, and assessing healing rate).

The assignment of the two treatments within each "pair" must be randomized (i.e. randomize the order).

57

Comparing treatments given to a set of identical twins is an example of _______.

A matched pairs experimental design

58

When the same individual is used twice in a matched pairs design, it is often referred to as "_______".

Repeated measures.

59

Is randomization in a matched-pairs design independent, or not?

NO

60

Is assignment to blocks random?

No

61

Double-blind experiment

Neither the subject, nor the people who interact with them know which treatment a subject is receiving.

62

The most serious potential weakness of experiments: _______.

Lack of realism

63

What is a block?

A group of individuals who are similar to each other in a meaningful way.

64

Are matched pairs or repeated measures a type of block study?

Yes

65

Can the same person in a repeated measure study be considered a block?

Yes