Reporting Results

Question 1

Answer 1

X = Exposure to virus
Y = Catching a cold
Z = Stress levels

Question 2

Criteria for hypotheses

Answer 2

*weed out if not …

Question 3

What is a hypothesis?

Expectation

Answer 3

1) A reasoned expectation

Question 4

What is a hypothesis?

Relations

Answer 4

Question 5

What is a hypothesis?

Falsification

Answer 5

Question 6

What is a hypothesis?

Answer

Answer 6

Question 7

Formulate hypothesis:
1) X –> Y
2) X –> A –> Y
3) X –> B –> Y, conditional on A

Answer 7

Question 8

Are the hypothesis supported?

Answer 8

H1 = yes
H2 = yes
H3 = uncertainty, no statistical significant measure of mediation -> comes down to decision making

Question 9

Which effect is strongest?

Answer 9

Hard to say without standardizing all the effects

Another problem: highly controlled estimates

Question 10

Can we say which mediator is stronger?

Answer 10

Sequence doesn’t allow for answering the question, no chance for choice mediator

also the nature of the hypothesis does not really allow it

Question 11

Interpret coefficient of father’s education in model 3

Answer 11

Question 12

How to interpret a coefficient?

Answer 12

be clear about
1) units
2) what is modelled, i.e. the mean
3) controls

Question 13

Interpret p-value of father’s education in model 3

Answer 13

If the 0 hypothesis is true, ergo no effect in the population, a repeated 7.8% of samples you could draw at rondom from the population would produce a test statistic as large as ours.

Question 14

Reporting steps:

Answer 14