Final Review Flashcards
Random Sample
Every member of the pop. has an equal chance of being selected.
Why are simple random samples preferred?
To prevent bias; so the sample represents the population.
Why not voluntary response sample?
May not represent the entire population because only those with a strong opinion respond.
Why do organizations do voluntary response samples?
Cheaper, easier, faster.
Correlation and causality
Does not show one causes the other, or that one prevents the other.
Loaded questoins
The answer is neither one or the other; evokes a passionate or politically incorrect response.
Frequency distribution
Title (measurement) , classes (name), frequency
*Make sure sum of fx=# in sample
Histogram
Title, frequency, name of classes (can use class widths or midpoints)
Mean of frequency distribution
Sum of class midpoints * frequency/sample size x̄=Σ(f*x)/Σf
Standard deviation
s= ✓Σ(x-x̄)^2/(n-1)
Variance
s^2
Or case
ADD
And case
MULTIPLY
And case with replacement
Denominator doesn’t change
And case without replacement
Denominator changes
Combination rule (P of getting x answers correct)
Px= nCrp^xq^n-x
Mean of probability distribution
μ=Σ(x*Px)
Standard deviation of probability distribution
σ=✓Σ(x^2*Px)-μ^2
Binomial dist. (M&Ms) mean
μ=np
Binomial dist. (M&Ms) standard deviation
σ=✓npq
Binomial dist. (M&Ms) range
μ+/- 2σ (no partial pieces!)
Find the indicated birth weight/pregnancy length
find z score (A2) z=(x-μ)/σ -> x=z*σ+μ
Critical Value for zα/2
α/2=1-confidence level/2 1-α A2=Zα/2
Critical Value for tα/2
df=n-1 1-confidence level=% in 2 tails (A3)=tα/2
Confidence level using margin of error
E=zα/2✓(p̂q̂/n)
Confidence interval using margin of error
p̂-E<p><p̂+E
8 step method
Step 1, Symbolic form
Step 2, If step 1 is false then (opposite step 1)
Step 3 If there is no equal sign, use I as H1.
Step 4 Select a significance level).05 or 0.01.
Step 5 Determine what formula and chart to use
Step 6 Find test stat and (P-value or critical region)
Step 7 Fail to or reject the null
Step 8 Restate with “significant evidence”
Find P-value
Find z for p. Look at area in tail(s) compared to α. Make sure area relates to tails.
P-value≤α=no go P-value>α=null fly
Traditional method
Find z of p. Find z of α. Compare z’s. If test stat falls in critical region, reject Ho
2 samples: P value method
Ho:P1=P2, step 5 p̅= q(bar)=, step 6 z=(p̂1-p̂2)-0/✓p̅qbar/n1+p̅qbar/n2
2 samples: Confidence interval method
Use (A3), step 6 tα/2= E=tα/2✓s1^2/n1+s2^2/n2 (x̄1-x̄2)-E<(x̄1-x̄2)+E , step 7 μ1-μ2=0 if CI contains zero fail to reject, step 8 not sufficient to support claim.