Statistics Flashcards
What’s a census?
Measurement of every member of a population
What are the benefits of a census?
Accurate results
What are the drawbacks of a census?
They are expensive
May destroy the products being tested ( may have to break the thing being tested)
What are sampling units?
Individuals of the population
What is the sampling frame?
A list of all the sampling units
What is random sampling?
Sampling with each unit having the same chance of being selected. (use of a random number generator or lottery sampling)
Advantages of random sampling?
Bias free
-ves if random sampling??
Requires a sampling frame
What is systematic sampling?
Taking a kth unit, 1st –> 11th –> 21st ect
+ves of systematic sampling?
Quick to use
-ves of systematic sampling?
Requires a sampling frame
What’s stratified sampling?
The sample represents the groups (strata) of the population
(sample/population) *strata
then pick randomly
+ves of stratified sampling?
Reflects the population
-ves of stratified sampling?
Data must be classified into strata
Quota sampling?
The same a stratified sampling but with interviewers
+ves of quota sampling?
no sampling frame
-ves of quota sampling?
not random/ bias
Opportunity sampling?
Quota filled with people that are available at the time
+ves of opportunity sampling?
Cheap and easy
-ves of opportunity sampling?
unlikely to be representative
mean (x̄)?
Σx / n or Σf(x) / Σf
Quartiles of Listed Data?
Q1 = n/4 Q2 = n/2 Q3 =3n/4
Quartiles of listed data when decimal and whole?
If you get a decimal, round up
If you get a whole number find the midpoint
How to interpolate?
Find the true class limits e.g
10-12 –> 9.5=< x <12.5
lower bound of true class limit + class width*(number into class/( total number in the class))
What is the benefit of using the IQR?
Ignores extremes
Variance?
(Σx^2)/n -(x̄)^2
mean of the squares - square of the mean
Coding?
y= ax + b the ȳ =ax̄=b
σy = aσx
What is frequency density?
Frequency / class width
Formula for the area of the histogram?
area = frequency x k
What is PMCC? (r)
only takes values between -1 and 1
Measures the strength of correlation
What is interpolation?
Estimation within the date range
What is extrapolation?
estimation outside the data range
P(B|A)
P(BnA)/P(A)
P(A u B)
P(A) + P(B) - P(AnB)
What is a discrete uniform distribution?
Probabilites of all outcomes are equal
What is a binomial distribution?
X ~ B(n, p)
When to use a binomial distribution?
Fixed number of trials
Fixed p of success
Independent trials
Either success or failure
Standard normal dist?
Z ~ N(0,1^2)
Z =y-μ/σ
Conditions of represent a Binonial as normal?
n is large
p is about 0.5
μ = np
σ^2 = np(1-p)
What must be done after approximating the normal?
Continuity correction
What is the actual significance level of a test?
When you reject Ho when it is true
Sxx to variance?
Sxx / n