When do you use z vs t?
z: proportions
t: means or slope
Recite the conditions for describing the distribution of quantitative data.
What is shape, outliers, center, and variability?
Our a=0.05 and our p-value is 0.00167
Reject the Ho
What is one thing that experiments have that observational studies don't?
What is random sampling/assignment?
How can you tell in a distribution if mean>median?
When distribution is skewed right
A student wants to know if the proportion of blue MnMs in a bag is greater than 0.2. What test should the student use?
What is a 1 sample z test for proportions?
Conditions for constructing a 1 sample t interval for mean
Random sampling, 10% rule, large counts (n>=10)
What is alpha=0.10
A majority of people responding that they work to care for the environment in a survey may lead to what sampling bias?
What is social desirability bias?
What is the IQR rule?
Q1-1.5(IQR)=x
Q3+1.5(IQR)=y
Anything <x is an outlier, anything y> is an outlier
What test would you use to find if the true mean amount of shoes in a shop are sold in a week?
What is a 1 sample t test for mean?
What is DUFS used for?
What describes the relationship between two quantitative variables in a scatterplot?
What significance level(s) would make a p-value of 0.09 statistically significant?
0.01 and 0.05
What does both random sampling and random assignment allow for?
What is generalizability and causation?
A distribution that measures how many tries it takes for a singular "success"
What is a geometric distribution?
A company claims 90% of shoe customers prefer Blue, 9% Red, and 1% Green. A survey of 100 customers (observed) finds 80 Blue, 15 Red, and 5 Green. What test would you use to see if these proportions are true?
What is a chi-square test of goodness fit?
What must be true for a distribution to be binomial?
Binary (success/fail), independent (each trial doesn't affect another), number of trials constant, success probability is equal throughout experiment.
Name the set of hypotheses you would use for a chi-square test of independence
Null (Ho): The two variables are independent, no association
Alternative (Ha): The two variables are dependent, association
Blocking in an experiment helps prevent what?
What are confounding variables?
Regardless of the original population's distribution, the distribution of sample means approaches a normal distribution (a bell curve) as the sample size gets larger, typically when n>=30.
What is the Central Limit Theorem (CLT)?
What is a t-test for slope
List all the conditions that need to be met to perform a slope (t)-test.
Linear relationship, independent, normal, equal SD, random
Interpret a p-value of 0.41.
If the null is true, there is a 0.41 chance of the data being as large or larger
Response vs nonresponse bias?
Response bias: Participants provide inaccurate or false answers (leading questions, social desirability) Nonresponse bias: Selected individuals fail to participate, leading to unrepresentative data
Give me your best definition of "statistics!"
The science of collecting, analyzing, interpreting, and presenting data