Exam 1 Flashcards
What does Matrix do in r and how would we create this?
Matrix(c(1,2,3,4,5,6))
Matrix combines the 2 columns
Why is sample different from Population parameters?
Because of sampling error.
On a scale of measurement, define Nominal scale and use starbucks as an example.
Nominal = unordered set of categories identified only by name. We can only tell if two individuals are the same or different.
-Tall, Grande, Venti (we don’t know the sizing yet).
On a scale of measurement, define Ordinal scale and use starbucks as an example.
Ordinal = ordered set of categories. It tells us the direction of difference between two individuals.
Tall -> Small
Grande -> Medium
Venti -> Large
On a scale of measurement, define Interval scale and use starbucks as an example.
Interval is like temperature… 0 means something.
4 steps to hypothesis testing
1) hypothesis: null or alternate
2) criteria for rejection: alpha value is >.05
3) statistic
4) decision: reject or retain null
Which test violates the assumption of independence of observation? Why?
Dependent. Because we are measuring the same person both times… so we must statistically differentiate them.
Nominal
Ordinal
Interval
Ratio
Nominal – Unordered categories (same or different: tall, grande, venti)
Ordinal – Ordered categories (direction of difference: tall -> small, grande ->med. venti ->large)
Interval – Zero is arbitrary (temperature, zero is a degree)
Ratio – Zero is absolute (weight, weight zero and you don’t exist)
Provide a written example of a t-test hypothesis testing from step 1 to 4.
1) hypothesis:
H0: No difference between means or mu1 - mu2 = 0 OR mu1 = mu2
H1: Difference between means
Mu1 - Mu2
2) criteria for rejection: alpha value is >.05
3) statistic:
4) decision:
reject or retain null
What do we need to know about computing an independent t-test?
numerator
df
variance
Numerator:
There are 2 separate samples so the numerator will be a difference of my 2 samples (x bar 1 minus x bar 2) subtracted by (mu 1 minus mu 2, which equals 0).
Denominator:
df: will be N-2, since there are 2 sample means.
Variance needs to be weighted due to unequal sample sizes, so we use POOLED VARIANCE: How much variance there is within the WHOLE GROUP.
- Take sum of squares (most raw)… per group. SS tells me HOW FAR AWAY IS MY MEANS FROM MY SCORES? Need SS to figure out variance.