Skills Lab 11- Mediation

Question 1

The direct and indirect effect will never be bigger than the

Answer 1

Total effect

Question 2

Within the model definition code what does ~ mean

Answer 2

Is predicted from

Question 3

Within the model definition code what does = mean

Answer 3

Is created from

Question 4

What is the code for defining the model

Answer 4

my_mod

Question 5

What is the code to fit the model

Answer 5

my_fit

Question 6

What function is used to fit the model

Answer 6

sem() from the lavaan package

Question 7

What is listwise deletion

Answer 7

Where an entire participant is excluded from analysis if any single variable is missing

Question 8

What is a better method to use than “listwise deletion” from the sem() function

Answer 8

• full information maximum likelihood (FIML)

- multiple imputation

Question 9

What does FIML produce

Answer 9

Unbiased parameter estimates and standard errors where there is missing data

Question 10

FIML can only be used when what is used

Answer 10

maximum likelihood estimators

Question 11

Maximum estimators work by

Answer 11

Choosing the parameters that make the data most likely to have happened

Question 12

What needs to be added to get maximum likelihood estimators

Answer 12

estimator = “ML”

Question 13

What code adds maximum likelihood estimators with robust standard errors

Answer 13

estimator = “MLR”

Question 14

What code can be used to look at how good the model is (the model fit statistics)

Answer 14

broom::glance()

Question 15

What code can be used to look at the actual effects (the model parameters)

Answer 15

broom::tidy(my_fit, conf.int = TRUE)

Question 16

What code can be used to create a diagram of the pathways

Answer 16

semPlot::semPaths()