9 Flashcards
Standard deviation in a sample is called the?
Standard error
Central limit theorem?
The sum or mean of a large number of measurements randomly sampled from any population is approximately normally distributed.
If there is a non normally distributed data at the first sample, what happens if you increase the sample size?
The distribution starts becoming normally distributed.
Y bar is normally distributed when?
•Y is normally distributed
•n is large because of the central limit theorem.
How do you estimate sigma when finding Z in a standard normal distribution for samples?
By using s since we know it.
Student’s t distribution?
A normal like distribution with sigma as s since sigma is unknown, in which the t distribution is a little bit wider and shorter than the standard normal distribution.
One-sample t-test?
Compares the mean of a random sample from a normal population with the population mean proposed in a null hypothesis.
If you have a normal distribution, the distribution of the mean must be?
Normal as well
Two groups?
A numerical variable measured in two groups (like two groups of people in which in group a, an experiment, and b, control).
How do you compare means if you have one categorical and one numerical variable? 2•
•Compare the mean of the difference between a numerical variable between two different groups but being as different as possible to each other. This is a paired comparison. •You can also a sample comparison design.
Paired t test?
Compares the mean of the differences to a value given in H_o by calculating the difference for each pair. It is a one-sample t test.
Assumptions of paired t test? 2•
•pairs are chosen at random
•population of differences has normal distribution
Not assume that individual values are normally distributed, only the differences are
In a paired t test, the number of datapoints is?
The number of pairs
Degrees of freedom in a paired t test?
df = number of pairs - 1
Comparing the means of two groups if not paired?
Estimate the difference in population means and your uncertainty on that estimate (confidence interval). Hypothesis test: 2-sample t test
