Lecture 10

Question 1

What is causality/causal relation?

Answer 1

A relationship between cause and effect

e.g. ‘a’ causes ‘b’

Question 2

In general, a process can have many causes. Those causes are called the ____ for it in its past.

Answer 2

causal factors

Question 3

An effect can be a cause of many others effects. This is called a ____.

Answer 3

causal chain

Question 4

Fill in the blanks and name the following condition (cause):
If ‘x’ is a ____ condition (cause) of ‘y’, then the presence of ‘y’ ____ indicates the presence of ‘x’. The presence of ‘x’, however, does not imply that ‘y’ will occur.

Answer 4

Necessary conditions: necessary, necessarily

Question 5

Fill in the blanks and name the following condition:

If ‘x’ is a ___ condition (cause) of ‘y’, the presence of ‘x’ indicates the presence of ‘y’.

Answer 5

Sufficient: sufficient

Question 6

What are the specific criteria for inferring causality? (3)

Answer 6

1. The cause precedes the effect in time (‘y’ changes after ‘x’: time sequence)
2. The two variables are empirically correlated with one another (if ‘x’ changes, ‘y’ changes statistically: correlation)
3. The logical relation (is ‘x’ is removed, then ‘y’ disappears: “the observed correlation between the two variables cannot be explained away as the result of the influence of some third variable that causes the two under consideration”)

Question 7

What is correlation?

Answer 7

The simultaneous change in values of two variables statistically.
e.g. the positive correlation between cigarette smoking and the incidence of lung cancer; the negative correlation between age and normal vision

Question 8

Many people are confused with the differences between ___ ___ and ___ and make incorrect conclusions.

Answer 8

causal relations; correlations

Question 9

True or false: a causality must be a correlation.

Answer 9

True

Question 10

True or false: A correlation must be a causality.

Answer 10

False

Question 11

True or false: A correlation is a necessary condition, but not a sufficient condition, for causality.

Answer 11

True

Question 12

True or false: A causality is a sufficient condition for a correlation.

Answer 12

True

Question 13

Which one is true?
A) a researcher is more likely to find evidence that an effect exists
B) a researcher is more likely to find evidence that an effect does not exist

Answer 13

A
“In many cases, a researcher is more likely to falsely find evidence that an effect exists that to correctly find evidence that it does not”

Question 14

What term do most studies use to claim correlations in order to proceed cautiously?

Answer 14

Statistically

- ‘a’ is STATISTICALLY associated with ‘b’, rather than ‘a’ CAUSES ‘b’

Question 15

Why study correlation?

Answer 15

Since correlation is a necessary condition of causality, understanding correlation helps us understand causality in the end.

Question 16

True or false: To study causal relations, you need to find a lagged process (e.g. ‘y’ lags ‘n’ intervals for time series analysis) and you have to find a strong logical relation between ‘x’ and ‘y’

Answer 16

True
(lag = fixed amount of passing time; one set of observations in a time series is plotted (lagged) against a second, later set of data)
e.g. Change of industrial structure, family size and divorce rate
https://www.khanacademy.org/math/probability/scatterplots-a1/creating-interpreting-scatterplots/v/correlation-and-causality

Question 17

True or false: It is easy to confirm causality in the real world.

Answer 17

False

Question 18

How can you control extraneous variables?

Answer 18

Study design

Question 19

What is internal validity?

Answer 19

The extent to which we can make clear-cut inferences about cause-consequence relations (in the sample being studied)

Question 20

Name and describe threats to internal validity. (7)

Answer 20

1. History; during the course of research, extraneous events may occur that will confound the results
2. Maturation or the passage of time; people continuously grow and change
3. Testing; the process of testing itself affects the results
4. Instrumentation changes; if we use different measures of the dependant variable at posttest than pretest, how can we be sure they are comparable to each other?
5. Statistical regression; because we are most likely to begin interventions for human problems that are inherently variable when those problems are are their most severe, we can expect some amelioration of the problem to occur solely because of the natural peaks and valleys in the problem and not necessarily because of the interventions
6. Sampling and selection bias; comparisons don’t have any real meaning unless the groups being compared are really comparable
7. Ambiguity about the direction of causal influence; there is a possibility of ambiguity concerning the time order of the independent and dependent variables

Question 21

Name this type of test, and describe its limitations:

X O

Answer 21

One-shot case study; it doesn’t even establish correlation, and it fails to control for any threats to internal validity

Question 22

Name this type of test and its limitations:

O1 X O2

Answer 22

One-group pretest-posttest; it does not account for factors other than the independent variable that might have caused the change between pretest and posttest results–factors usually associated with the following threats to internal validity: history, maturation, testing, and statistical regression.

Question 23

Name this type of test and its limitations:
X O

       O

Answer 23

Posttest only with non-equivalent groups; fails to control for the threat of selection bias–without pretests, we have no way of knowing whether the score of the two groups would have differed as much to begin with.

Question 24

Some studies using pre-experimental designs can be valuable despite their extremely limited degree of internal validity. What tends to make them valuable as pilot studies? (4)

Answer 24

1. To generate tentative exploratory or descriptive information regarding a new intervention about which little is known
2. To learn whether it is feasible to provide the new intervention as intended
3. To identify obstacles in carrying out methodological aspects of a more internally valid design that is planned for the future
4. To see if the hypothesis for a more rigorous study remains plausible based on the pilot study results

Question 25

What do experimental designs do?

Answer 25

They provide maximum control for threats to internal validity by randomly assigning research participants to ‘experimental’ and ‘control’ groups.

Question 26

True or false: An experimental design is a poor vehicle for the controlled testing of causal relations.

Answer 26

False; it is an excellent vehicle

Question 27

What is an experimental group?

Answer 27

A group of subjects exposed to the independent variable (intervention) of an experiment.

Question 28

What is an control group?

Answer 28

A group for which the effect of the independent variable (intervention) is absent or held constant, in order to provide a comparison.

Question 29

What is randomization, and why is it the generally preferred method for achieving comparability?

Answer 29

Randomization is the random assignment of experimental and control groups in studies; random means no observable pattern, and the researcher cannot manipulate the study.

Question 30

What is this an example of?
R O1 X O2
R O1 —- O2

Answer 30

The classic experimental design, also called the pretest-posttest control group design. 
R = random assignment
O1 = pretest
X = tested interventions
O2 = posttest

Question 31

True or false: In experimental design, because participants are assigned on a random basis, it can be assumed that the experimental group is statistically the same as the control group at the beginning.

Answer 31

True

Question 32

True or false: In experimental design, if the improvement were caused by history or maturation (not by intervention), there would be no way the experimental group could improve any more than the control group.

Answer 32

True

Question 33

What is this an example of, and what does it assume?
R X O
R O

Answer 33

Post-only control group design; this design assumes that any difference between the two groups at the posttest reflects the causal impact of the intervention.

Question 34

What does the pretest-posttest control group design assume?

Answer 34

That there is no effect from pretest.

Question 35

If we think that a pretest might bias posttest response, what type of experimental design might we opt for?

Answer 35

Post-only control group design

Question 36

What is this an example of?
R    O1    X    O2
R    O1           O2
R            X     O2
R                    O2

Answer 36

Solomon four-group design

Question 37

Why would we use a Solomon four-group design?

Answer 37

If we would like to know the amount of pretest-posttest change, but are worried about testing effects.
This design compares the results of the four groups to understand the effects of the intervention as well as the effects of pretests.

Question 38

What does randomization involve?

Answer 38

Randomly allocating the experimental participants across the treatment groups.
“The principal technique of randomization simply entails using procedures based on probability theory to assign research participants to experimental and control groups.”
e.g. the researchers might flip a coin, or ask the participants to select a piece of paper with a number 1 or 2 on it from a box

Question 39

What is a random sample?

Answer 39

A sample where every individual has the same probability of being sampled.

Question 40

What is matching?

Answer 40

Matching is where pairs of participants are matched on the basis of their similarities on one or more variables, and one member of the pair is then randomly assigned to the experimental group and the other to the control group.

Question 41

What is one way to further improve the chances of obtaining comparable groups?

Answer 41

Combine randomization with matching.

Question 42

What is a blind experiment?

Answer 42

An experiment in which information about the test is masked from the participant in order to reduce or eliminate bias.

Question 43

What is a double-blind experiment?

Answer 43

An experiment where both the tester and participant are blinded, to reduce or eliminate bias.

Question 44

What is external validity?

Answer 44

The extent to which the experimental results can be applied to real situations.
“In statistics, external validity refers to the extent to which the results of an investigation may be generalized to the population as a whole, and to other populations, settings, measurement devices, etc.

Question 45

What is one of the things external validity depends on?

Answer 45

The adequacy of the sampling.
In some cases, the internal validity is strong, but the external validity is weak. The problem is that the sample used is different from the general population.

Question 46

What are threats to external validity? (2)

Answer 46

1. Sampling

2. All of the threats to internal validity