lecture 5

Question 1

Paired T-Test

Answer 1

compares two observations that are correlated with each other
example: heart rate before and after running a marathon
Note: take the difference between individuals observations which will from the a single distribution which will be used for the t-test.

Question 2

what are Paired T-tests good for

Answer 2

-comparing before and after effects
- controlling and taking into account for differences among the individuals (at the same time and different times)

Question 3

what are you trying to see when you use a paired t-test?

Answer 3

you are trying to see the mean of the difference is = to 0

Question 4

Null for paired T-test is

Answer 4

when the mean of the difference among individuals= 0

Question 5

ANOVA (analysis of variance)

Answer 5

compares two or more means
allows for us to see if there is any difference among samples/groups

Question 6

what does ANOVA calculate

Answer 6

1.)the measurement of there mean among all groups. and then calculates the difference between the mean among all groups and all individual measurements
2.) calculates each the mean of the each group separately, and then calculates the difference between the mean of the group and individual measurements of that group. (do this for all groups separately)
Note: (summation of variance/distribution will give you a new score which is an F statistic.

Question 7

F-statistic

Answer 7

among group variance/ within group variance
all of the distances to the overall mean/ the sum of all distances from every observation to its own groups mean

Question 8

which variance in the f-statistic is typically bigger?

Answer 8

the Numerator because the lines are longer in the among group variance compared to the within group varience

Question 9

What makes the F-stat value bigger

Answer 9

the bigger the difference between the among group and the within group will make the f-stat number higher.
(take a big number/ by a small number)
Note: opposite will make the f-stat smaller

Question 10

if our stat null hypothesis is that there is no difference among means what do we expect our ratio value to be? (anova/Fstat)

Answer 10

if the among group variance value is= to within group variance value
then the ratio would have to =1

Question 11

what are the 3 assumptions for parametric test

Answer 11

-sampling independence
- homogeneity of variance
- normally distributed variance

Question 12

Sampling independence

Answer 12

Observations in each group are randomly and independently from the population (aka no bias)
every individual within a defined population has an equal chance of being selected to be in a sample every time an observation is made.

Question 13

Normally distributed variance

Answer 13

observations are normally distributed around the groups/samples mean
Note: if the distribution is skewed then it inflates the within the group variance which will mess up f-statistic

Question 14

is a multimodal distribution meet the normal distributed variance? can u change it to met this assumption?

Answer 14

no because the multimodality is treated like single group/sample when in actuality it is more than one group.
yes you can split the data to make them all unimodal.

Question 15

Homogeneity of variance (homoscedasticity)

Answer 15

the different groups have similar dispersion
the variance between the different groups need to be similar too each other
(group variance need to have a similar wideness)

Question 16

Unpaired t-test (two sample t-test)

Answer 16

compares the means of two independent groups to determine if they are significantly different
example: If you measure the weight of 50 men and 50 women

Question 17

Unpaired t-test (two sample t-test)

Answer 17

here is no significant difference between the means of two independent