Lecture 2 - Missing Data

Question 1

What is missing data?

What is the criteria for an inadequately sampled participant or variable?

Answer 1

The absence of data points in a dataset, which can pose significant challenges in data analysis.

Over 50% of their data is missing.

Question 2

What is selection bias?

Answer 2

Selection bias occurs when the sample of participants included in a study is not representative of the population intended to be analyzed, leading to spurious effects.

Question 3

What are the types of missing data?

Answer 3

Missing Completely at Random (MCAR), Missing at Random (MAR), and Missing Not at Random (MNAR).

Question 4

What is MCAR?

Answer 4

MCAR means the probability of a data point being missing is independent of any observed or unobserved data.

Question 5

What is MAR?

Answer 5

MAR means the probability of a data point being missing is related to some observed data but not to the missing data itself.

Question 6

What is MNAR?

Answer 6

MNAR means the probability of a data point being missing is related to the value of the missing data itself.

Question 7

How can missing data reduce power?

Answer 7

Missing data can reduce statistical power due to a smaller effective sample size.

Question 8

How can missing data cause biased estimates?

Answer 8

Systematic missing data can lead to biased parameter estimates.

Question 9

What is Little’s MCAR test?

Answer 9

A statistical test used to determine if data are missing completely at random.

Question 10

What are separate variance t-tests used for?

Answer 10

Used to follow up significant results from Little’s MCAR test to further investigate the nature of missing data.

Question 11

What is estimation maximization (EM)?

Answer 11

A method for imputing missing data based on maximum likelihood estimation.

Question 12

What is multiple imputation (MI)?

Answer 12

A method that involves creating multiple datasets with imputed values to account for the uncertainty in the imputations.

Question 13

How should missing data be handled after a study?

Answer 13

Be transparent about the extent and nature of missing data.

Question 14

What are SPSS checks for missing data?

Answer 14

Checking participant sampling via the number and proportion of missing data, and assessing the adequacy of variable assessment.

Question 15

How can inadequately sampled participants be dealt with?

Answer 15

Inadequately sampled participants can be handled by:
1. Deletion: Removing them from the analysis if their missing data is not systematic and does not introduce bias.
2. Imputation: Use imputation methods to estimate missing data, provided the assumptions for imputation are met.
3. Sensitivity Analysis: Conduct sensitivity analysis to assess the impact of including or excluding these participants on the study’s conclusions.

Question 16

What is the difference between selection bias and missing data?

Answer 16

Selection bias concerns who participates in the study, while missing data concerns the data provided by participants.

Question 17

How can inadequately assessed variables be dealt with?

Answer 17

Inadequately assessed variables can be handled by:
Deletion: Remove the variable from the analysis if it is not crucial to the research question and if its removal does not undermine the study’s validity.
Imputation: If the variable is crucial, use imputation methods to estimate the missing data, ensuring that the assumptions for imputation are met and the imputation method is appropriate for the type of data.
Sensitivity Analysis: Conduct a sensitivity analysis to understand the impact of including or excluding the variable on the study’s conclusions.

Question 18

What is the impact of systematic missing data?

Answer 18

It can lead to biased estimates and erroneous conclusions.

Question 19

What is the goal of missing data analysis?

Answer 19

To understand whether missing data will cause problems for analyses and inferences, and to estimate missing data if possible.