Week 7 - Sampling Distributions

Question 1

Sample descriptive statistics provide an estimate of what?

Answer 1

The population parameter

Question 2

What is statistical inference?

Answer 2

Being able to infer and generalize from a sample to a larger population

Question 3

Fixed means…

Answer 3

If you were to sample everyone it would be the same mean and SD

Question 4

What is the difference between what sampling and population distributions show?

population: Spread
sampling: Variation

Answer 4

Population distributions: the spread of the variable of interest across individuals in the target population
Sampling distributions: show how the estimate of the population mean varies between samples if the experiment was repeated and the mean was recalculated each time

Question 5

As sample size increases, what happens to the data?

Answer 5

The data becomes more normally distributed

Question 6

How does the increase in the n value affect the distribution of the mean?

Answer 6

The larger n is, the closer the sample mean matches the population mean

Question 7

What are the properties of sampling distribution?

Answer 7

1. The mean of the sampling distribution is equal to the sampling and population mean
2. SD of the sampling distribution is equal to the SD of distribution mean and the standard error of the mean
3. The shape of he sampling distribution is a normal bell curve and as sample size increases, approximation to the normal distribution does too (even if the population distribution is not bell-shaped)

Question 8

What is the central limit theorem? What n value is considered normally distributed for sampling distributions based on the mean?

Answer 8

When n => 30 for sampling distributions based on the mean, the data is considered normally distributed

Question 8

As sample size increases, what occurs to the sampling distribution?

Answer 8

Variability in the mean measurements decreases; standard error (SE) decreases

Question 9

What is the 95% confidence interval?

Answer 9

If we were to conduct the experiment 100 times, 95 times out of the 100, the value would fall within the 95% confidence interval range.

Question 10

What are parametric vs non-parametric tests?

Answer 10

Parametric tests assume the data is normally distributed (t-tests, ANOVA) while non-parametric tests are data that do not follow a normal distribution.

Question 11

What is the calculation to determine the 95% CI for a single population when the population SD is known?

Answer 11

Lower limit: mean - 1.96 (SE)
Upper limit: mean + 1.96 (SE)

Question 12

What is the calculation to determine the 95% CI for a single population when the population SD is not known?

Answer 12

Xu = m + tu x SE
XL = m + tL x SE

Tlower and Tupper are standardized test scored found using the t-distribution table

Question 13

What does the t-distribution look like?

Answer 13

The shape varies depending on sample size yet is still symmetric, bell-shaped, and centered on mean = 0.
The larger the df, the more the sample SD approximates the population SD, and the shape approximates the normal distribution.

Question 14

How do you calculate the degrees of freedom for t-distribution?

Answer 14

df = n - 1

Question 15

What is the difference between 1- vs 2-tailed tests?

Answer 15

One-tailed: there is a specific direction (question example: is a value larger or smaller)
Two-tailed: no specific direction (question example: is the value different)

Question 16

When do you use the general formula for calculating the 95% CI?

Answer 16

If you have the population standard deviation
If the study population is equal to or greater than n=30

If not you use the t-distribution formula

Question 17

What does a p-value tell you?

Answer 17

Tells you if the effect or difference is significant where values closer to zero are not likely due to chance and values closer to 1 are likely due to chance (0.05 cutoff)