Summary Table for Statistical Techniques

Type of Minitab
Inference Parameter Statistic Examples Analysis Conditions
Data Command
What is the average
weight of adults? 1-sample t-interval Data approximately normal
One
s Stat > Basic statistics OR
1
Estimating a population Sample mean
Numerical What is the average
x ± tα / 2 > 1-sample t Have a large sample
mean mean µ x n
cholesterol level of size (n ≥ 30)
adult females?

Is the average GPA

of students at a school Ho : µ = µo
higher than 8.0? Ha : µ ≠ µo or Ha : µ > µo Data approximately normal
One
. or Ha : µ < µo OR
Test about a population Sample mean Stat > Basic statistics
2 mean mean µ x
Numerical The one sample t test:
> 1-sample t
Have a large sample
Is the average x − µ0 size (n ≥ 30)
Winter temperature t=
s
in State n
less than 10ْ C?
What is the
One proportion of males 1-proportion Z-interval
population Sample in the world?
Estimating a proportion Categorical ˆ ˆ
p (1 − p ) Stat > Basic statistics Have at least 5
3 proportion
proportion
(binary) pˆ ± zα / 2 > 1-sample proportion in each category
p p̂ What is the n
proportion of
students that smoke?
Ho : p = po
Is the proportion
Ha : p ≠ po or Ha : p > po
of females different
One or Ha : p < po
Sample from 0.5?
population The one proportion Z-test:
Test about a proportion Categorical Stat > Basic statistics
4 proportion
proportion
(binary) Is the proportion of > 1-sample proportion
n po ≥ 5 and n (1-po ) ≥ 5
p p̂ pˆ − p0
students who fail in z=
Stats less than p0 (1 − p0 )
0.1?
n
 Type of Minitab
Inference Parameter Statistic Examples Analysis Conditions
Data Command
How different are
the mean GPAs of
Independent samples from
Difference in males and females?
Difference in the two populations
two
Estimating the population 2-sample t-interval Stat > Basic statistics
two sample How many fewer
5 difference of means means Numerical
colds do vitamin C
> 2-sample t Data in each sample are
two means ( x1 − x 2 ) ± t * × s.e.( x1 − x 2 ) about normal or large
µ1-µ2 x1 − x 2 takers get, on
samples
average, than non
vitamin C takers?

Do the mean pulse

rates of exercisers Ho : µ1 = µ2
Difference in and non-exercisers Ha : µ1 ≠ µ2 or Ha : µ1 > µ2 Independent samples from
two Difference in differ? or Ha : µ1 < µ2 the two populations
Test to population two sample The two sample t test: Stat > Basic statistics
6 compare two means means Numerical
Is the mean EDS ( x − x2 ) − 0 > 2-sample t Data in each sample are
means µ1-µ2 x1 − x 2 score for dropouts
t= 1 about normal or large
s.e.( x1 − x2 )
greater than the samples
See text, page 445, for
mean EDS score for
the s.e. of the difference
graduates?

What is the Differences approximately

Mean of Sample mean paired t-interval
difference normal
Estimating a paired of in pulse rates, on the sd Stat > Basic statistics OR
7 mean with difference difference Numerical average, before and d ± tα / 2 Have a large number
µD > Paired t
paired data after exercise?
n of pairs (n ≥ 30)
d

Is the difference in Ho : µD = 0 Differences approximately

Mean of Sample mean IQ of pairs of twins Ha : µD ≠ 0 or Ha : µD > 0 normal
Test about a paired of zero? or Ha : µD < 0 Stat > Basic statistics OR
8 mean with difference difference Numerical
d −0 > Paired t Have a large number
paired data µD Are the pulse rates t=
sd of pairs (n ≥ 30)
d of people higher
after exercise? n
 Type of Minitab
Inference Parameter Statistic Examples Analysis Conditions
Data Command
How different are
the percentages of
Difference in male and female
two two-proportions Z-interval Independent samples
Difference in smokers?
Estimating the population from the two populations.
two sample Stat > Basic
difference of Categorical
9 two
proportions proportions (binary)
How different are ( pˆ 1 − pˆ 2 ) ± statistics
Have at least 5 in each
the percentages of > 2 proportions
proportions p -p pˆ 1 − pˆ 2 zα / 2 × s.e.( pˆ 1 − pˆ 2 ) category for both
1 2 upper- and lower-
populations.
class binge
drinkers?

Ho : p1 = p2
Is the percentage of
males with lung
Ha : p1 ≠ p2 or Ha : p1 > p2
Difference in cancer higher or Ha : p1 < p2
Difference in The two proportion z test: Independent samples
two than the percentage
two sample pˆ 1 − pˆ 2 from the two populations.
Test to population of females with lung Stat > Basic
10 compare two proportions proportions Categorical
cancer?
z= statistics
proportions
(binary) 1 1  > 2 proportions
Have at least 5 in each
pˆ (1 − pˆ ) +  category for both
p1- p2 pˆ 1 − pˆ 2 Are the percentages  n1 n2  populations.
of upper- and lower-
class binge drinkers x + x2
different? pˆ = 1
n1 + n2
Relationship Is there a Ho : The two variables are
between two relationship not related
categorical between smoking Ha : The two variables are All expected counts should
variables and lung cancer? related Stat >Tables be greater than 1
OR The observed
Relationship The chi-square statistic: > CrossTabu-
11 in a 2-way Difference in counts in a Categorical Do the proportions lation > Chi- At least 80% of the cells
two-way table c2 = should have an expected
table two or more of students in each Square analysis
population class who smoke (Observed − Expected ) 2 count greater than 5
proportions differ? ∑
all Expected
cells
 Type of Minitab
Inference Parameter Statistic Examples Analysis Conditions
Data Command
The form of the equation
that links the two variables
Ho : β1 = 0 must be correct
Slope of the
population Sample
Ha : β1 ≠ 0 or Ha : β1 > 0
Is there a linear or Ha : β1 < 0 The error terms
regression estimate of Stat >
Test about a relationship between The t test with n-2 degrees of are normally distributed
12 slope
line the slope Numerical
height and weight of
Regression
freedom: > Regression
a person? b −0 The errors terms have equal
β1 b1
t= 1 Variances
s.e.(b1 )
The error terms are
independent of each other
Each population is normally
Ho : µ1 = µ2=˙˙˙ = µt Distributed
Population Ha : not all the means are
Test to Sample mean Is there a difference
means of the equal
compare of the t between the mean Independent samples
t populations Stat >ANOVA
13 several populations Numerical GPA of Freshman, The F test for one-way
> Oneway
from the t populations
means Sophomore, Junior ANOVA:
µ1,µ2,….., ,µt
x1,x2,….., ,xt and Senior classes? MSTR Equal population standard
F= deviations
MSE
Test of 2 variables are continuous
Strength
& Population Ho: ρ = 0 Related pairs
Direction Correlation Sample
correlation Is there a linear H a: ρ ≠ 0 Stat > Basic
of Linear No significant outliers
14 Relationsh ρ
Numerical relationship between Statistics >
height and weight? Correlation
ip of 2 r 𝑟𝑟√𝑛𝑛 − 2 Normality of both variables
Quantitati "rho" 𝑡𝑡 =
ve √1 − 𝑟𝑟 2 Linear relationship between the
Variables variables

Sample Are the variances of

Test to length of lumber Ho: σ12 = σ22 Each population is normally
Two population variances
Compare variances of 2 produced by Stat > Basic Distributed
15 Two Numerical Company A Ha: σ12 ≠ σ22 statistics >
populations
Population different from 2 variances Independent samples from
Variances σ12, σ22 those produced by the 2 populations
2
s1 , s2 2 F = s1 / s2
2 2

Company B