STATS

Question 1

How does SEM and SD compare?

Answer 1

SEM (standard error of the mean) is an inferential statistic describing the likelihood that a sample mean is the actual population mean. SEM = SD/SQRT(n).

SEM is thus always smaller than the SD. It can be misused in place of SD to make the data look less variable.

Question 2

When to use a student’s t test?

Answer 2

Comparing Unpaired groups with con(t)inuous data

Question 3

When to use a paired t test?

Answer 3

Paired data that is continuous

Question 4

When to use ANOVA?

Answer 4

Continuous data, more than two groups. A significant results says that at least two of the groups are different. Use post-hoc analysis to figure out which ones are actually different.

Question 5

When to use Wilcoxon Rank Sum (Mann Whitney U) test?

Answer 5

When you have 2 independent groups and ordinal data (non-parametric data)

Question 6

When to use Chi-Squared test?

Answer 6

Nominal data (2 or more independent groups)