Scribd
Upload
48 views8 pages

1 Prop Z Test

1. The document provides summaries of common statistical hypothesis tests including one and two sample z-tests, t-tests, tests of proportions, chi-square tests, linear regression t-tests, and how to set up the null and alternative hypotheses and determine whether assumptions are met for each. 2. It describes how to calculate each test statistic such as z-scores, t-statistics, and chi-square values and the conditions needed such as sample size, normality, and independence. 3. Formulas are provided for calculating standard errors and confidence intervals for estimating population parameters from sample statistics for each test.

Uploaded by

Robert Jenkins
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
48 views8 pages

1 Prop Z Test

1. The document provides summaries of common statistical hypothesis tests including one and two sample z-tests, t-tests, tests of proportions, chi-square tests, linear regression t-tests, and how to set up the null and alternative hypotheses and determine whether assumptions are met for each. 2. It describes how to calculate each test statistic such as z-scores, t-statistics, and chi-square values and the conditions needed such as sample size, normality, and independence. 3. Formulas are provided for calculating standard errors and confidence intervals for estimating population parameters from sample statistics for each test.

Uploaded by

Robert Jenkins
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 8

STATS FRQ

1 Prop Z Test
Given: one sample proportion or percent is testing against a claimed
population proportion/percent
Ho: P=?
Ha: P<? || P>? || p!=?
Conditions: random
normal [np>=10 && n(1-p)>=10]
10%
Calculations:
Z=(obs-mean)/sd
sd=pp = sqrt[ P * ( 1 - P ) / n ]
1 Prop Z Int
Given: 1 sample proportion or percent and might e given a claimed
proportion or percent
Ho: P=?
Ha: P<? || P>? || p!=?
Conditions:
Random
normal [np>=10 && n(1-p)>=10]
10%
Calculations: (stats wizard off)
CI = pp Z* SE(pp )
SE = sqrt[pp * ( 1 - pp ) / n ]
Z*=invNorm(area||CI)

T Test
Given: a sample mean, or one list of data. 2 sample means, or 2 lists
from PAIRED data and testing against a claimed population mean

Ho: P=?
Ha: P<? || P>? || p!=?
Conditions:
Random
Normal: said in the problem||if n>30 CLT||when data is given, do
histogram&box
10%
Calculations:
t=(pbs-mean)/sd
sd=SE(xx )=Sx/sqrt(n)
Sx one var stats

T Test
Given: a sample mean, or one list of data. 2 sample means, or 2 lists
from PAIRED data and testing against a claimed population mean
Ho: P=?
Ha: P<? || P>? || p!=?
Conditions:
Random
Normal: said in the problem||if n>30 CLT||when data is given, do
histogram&box
10%
Calculations:
t=(pbs-mean)/sd
sd=SE(xx )=Sx/sqrt(n)
Sx one var stats

T Int

Given: a sample mean, or one list of data. 2 sample means, or 2 lists


from PAIRED data and testing against a claimed population mean
Ho: P=?
Ha: P<? || P>? || p!=?
Conditions:
Random
Normal: said in the problem||if n>30 CLT||when data is given, do
histogram&box
10%
Calculations:
CI = xx t*n-1 SE(pp )
sd=SE(xx )=Sx/sqrt(n)
Sx one var stats
t*n-1 = invT(area||CI,df ??)
2 Prop Z Test
Given: 2 sample proportions or percents
Asked: one group is greater/less/different than the other
Ho: p1=p2 || p1-p2=0
Ha: p1>p2 || p1<p2 || p1!=p2
Conditions:
Random: state both
Normal:
n1 pp 1>=10 n1 (1-pp 1)>=10
n2 pp 2>=10 n2 (1-pp 2)>=10
10% when doing observational study or independence when doing
comparative randomized experiment
Calculations:
CI = pp 1-2 Z* SE(pp 1-2)
SE(pp 1-2)= sqrt[ pp 1(1- pp 1)/n1+ pp 2(1- pp 2)/n2]

2 Prop Z Int
Given: 2 sample proportions or percents
Asked: one group is greater/less/different than the other
Ho: p1=p2 || p1-p2=0
Ha: p1>p2 || p1<p2 || p1!=p2
Conditions:
Random: state both
Normal:
n1 pp 1>=10 n1 (1-pp 1)>=10
n2 pp 2>=10 n2 (1-pp 2)>=10
10% when doing observational study or independence when doing
comparative randomized experiment
Calculations:
Z=(obs-mean)/sd = pp 1-2/sd
pp 1-2=sqrt[pp c(1- pp c)/n1 + pp c(1- pp c)/n2]
pp c=(x1+x2)/(n1+n2)
2 Sample T Test
Given: 2 sample means or 2 lists of data
Asked: one group is greater/less/different than the other
Ho: 1=2 || 1-2=0
Ha: 1>2 || 1<2 || 1!= 2
Conditions:
Random: state both
Normal:
n1 pp 1>=10 n1 (1-pp 1)>=10
n2 pp 2>=10 n2 (1-pp 2)>=10
10% when doing observational study or independence when doing
comparative randomized experiment
Calculations:
CI = pp 1-2 t*n-2 SE(xx )

SE(xx )= sqrt(S1^2/n1+S2^2/n2)
t*n-2=invT(area||CI,df)
2 Sample T Int
Given: 2 sample means or 2 lists of data
Asked: one group is greater/less/different than the other
Ho: 1=2 || 1-2=0
Ha: 1>2 || 1<2 || 1!= 2
Conditions:
Random: state both
Normal:
n1 pp 1>=10 n1 (1-pp 1)>=10
n2 pp 2>=10 n2 (1-pp 2)>=10
10% when doing observational study or independence when doing
comparative randomized experiment
Calculations:
t=(obs-mean)/sd = xx 1-2/sd
sd= sqrt(S1^2/n1+S2^2/n2)
Chi-square Goodness of Fit
Given: one sample distributed over two or more categories
Ho: results are consistent with the theoretical distribution
Ha: results are differ from the theoretical distribution
df=#of categories -1
Conditions:
Random
Expected Cells: common sense
Independence: 10%
Calculations:
observes in L1, expected in L2

Chi-square Independence || association


Given: two categorical variables
Keywords: independence, association, relationship
Ho: the two variables are independent, or no association
Ha: the two variables are not independent, or there is association
df=(#of col -1)(#of row-1)
Conditions:
Random
Expected Cells: common sense
Independence: 10%
Calculations:
observes in L1, expected in L2
chi^2 = (obs-mean)^2/exp
Chi-square Homogeneity
Given: at least two samples or two groups
Asked: if the populations or groups are distributed similarly or different
Ho: groups or populations are distributed the same
Ha: groups or populations are distributed the differently
df=(#of col -1)(#of row-1)
Conditions:
Random
Expected Cells Frequency: row x col / total-total
Independent: 10%
Calculations:
observes in [A] then chi-square then auto filled [B]
chi^2 = (obs-mean)^2/exp

Linreg T Test
Parameter: =0

Statistic: b=correlation from the table


Ho: =0 (no linear association)
Ha: <0 || >0 (there is positive||negative association)
Conditions:
Linear: do a scatterplot and LSRL
Independent: say something smart ^^
Equal Variance: residual plot
Normal: boxplot and histogram of residuals
Random
Calculations:
t=(b- )/SEb
SEb=S/Sx sqrt/(n-1)
Sx from one var stats

Linreg T Int
Parameter: =0
Statistic: b=correlation from the table
Ho: =0 (no linear association)
Ha: <0 || >0 (there is positive||negative association)
Conditions:
Linear: do a scatterplot and LSRL
Independent: say something smart ^^
Equal Variance: residual plot
Normal: boxplot and histogram of residuals
Random
Calculations:
b t* (SEb)
SEb=S/Sx sqrt/(n-1)
Sx from one var stats
t* = invT(area||CI,df n-2)

You might also like