5 Basic Steps in Any Hypothesis Test

Step 1: Determine hypotheses (H0 and Ha).

H0: μd = 0 (μ1 –μ2=0)

Ha: μd > 0 (μ1 –μ2>0) Ha: : μd ≠ 0 (μ1 –μ2≠0) Ha: μd <0 (μ1 –μ2<0)
upper-sided two-sided lower-sided

Step 2: Verify necessary conditions, compute an appropriate

test statistic.
Step 3: Assuming H0 is true, find Decision Rule
Step 4: Decide whether or not Reject H0.
Step 5: Report the conclusion in the context of the problem.
Hypothesis Test for Population
Mean Difference (H0: μd =μd0)
 Required assumptions:
 Data are 2 dependent random samples (on the
same set of objects)
 Large n(≥30) or Normal Distribution of
differences X1 - X2
 Test Statistics:
 Population Difference in Means
Test Data Population Sample Response Explanatory
Scenario Parameter Statistics Variable

Population 2 Numerical Group

Difference Independent
in Means Samples (GPA, (Gender)
Weight)

Do students who live on campus tend to spend more time studying

in the library than students who live off-campus, on average?
Do women spend more money than men, on average?
Do women have more shoes than men, on average?
Name Scenario
 A researcher is studying the effect of a new teaching
technique for middle school students.
 A class of 30 students is taught using the new technique
and their mean score on a standardized test is compared to
the mean score of a class of 27 students who were taught
using the old technique.
 One-sample test
 Two independent sample test
 Paired test
Name Scenario
 A company claims that the economy size version of their
product contains 32 ounces.
 A consumer group decides to test the claim by examining a
random sample of 100 economy size boxes of the product,
since they have received reports that the boxes contain less
than the 32 ounces claimed.
 One-sample test
 Two independent sample test
 Paired test
Name Scenario
 At some universities, an athletic director decides to test
whether or not athletes do in fact have lower GPAs. If
there is evidence of lower GPAs, additional academic
support for athletes will be instituted.
 A random sample of 200 student athletes and a random
sample of 500 non-athlete students is taken and their
GPAs are recorded.
 One-sample test
 Two independent sample test
 Paired test
Name Scenario
 As part of a biology project, some high school students
compare heart rates before and after running a mile for 40
of their classmates.
 They want to see if the after heart rate of student’s their
age is faster than the before heart rate, on average.
 One-sample test
 Two independent sample test
 Paired test
Name Scenario
 A hospital is studying patient costs and they decide to
follow 500 surgery patients’ hospital and medical bills for
a year after surgery and compare them to the estimated
costs provided to the patients before surgery.
 They want to see if the estimated and actual costs are
comparable on average.
 One-sample test
 Two independent sample test
 Paired test
Name Scenario
 A group of students is interested if BU students in West
campus exercise more regularly than those in East, on
average.
 One-sample test
 Two independent sample test
 Paired test
Name Scenario
 A group of students is interested if BU female students
spend more time studying than BU male students, on
average.
 One-sample test
 Two independent sample test
 Paired test
Name Scenario
 Can we conclude that students tend to sleep more at the
beginning of the semester than at the end of the semester,
on average?
 One-sample test
 Two independent sample test
 Paired test
Hypothesis Test for Population
Difference in Means (H0: μ1 – μ2 =0)
 Required assumptions:
 Data are 2 independent random samples
 Large samples (n1≥30 and n2 ≥30) or Normal
Distribution of X1 and X2

 Test Statistics: Z or t
Hypothesis Test for Population
Difference in Means (H0: μ1 – μ2 =0)
Hypothesis Test for Population
Difference in Means (H0: μ1 – μ2 =0)
 Case 1: Two independent Populations –
Population Variances Known
Attributes Test Statistic Confidence Interval
 Example
 Estimating Difference in Mean Walking Distance
between Two Physical Therapy Programs.
Physical Therapy 1
4 Sessions
(1 hour/day)
Patients Undergoing
Total Knee Randomize
Replacement
Physical Therapy 2
2 Sessions
(2 hour/day)

 Response: the number of feet patients can walk

independently
 Example
 Estimating Difference in Mean Walking Distance
between Two Physical Therapy Programs.
 Response: the number of feet patients can walk
independently
 Summary:
4-day program: n1=18
2-day program: n2=16
 Question: How can we compare physical therapy
programs?
 Example
 Question: How can we compare physical therapy
programs?
4-day program: n1=18
2-day program: n2=16
 Idea 1: Compute CIs for mean # feet for both
therapies:
 Example
 Question: How can we compare physical therapy
programs?
4-day program: n1=18
2-day program: n2=16
 Idea 1: Compute CIs for mean # feet for both
therapies:

 4-day program: 95% CI (46.03 feet, 48.37 feet)

 2 –day program: 95%CI (23.6 feet, 25.6 feet)
 Example
 Question: How can we compare physical therapy
programs?
4-day program: n1=18
2-day program: n2=16
 Idea 2: Compute 95% CI for difference in population
means:
 Example
 Question: How can we compare physical therapy
programs?
4-day program: n1=18
2-day program: n2=16
 Idea 2: 95% CI for difference in population means:
(21.06 feet, 24.14 feet)
 Does this 95% CI contain 0?
 What does it mean?
 Example
 Question: How can we compare physical therapy
programs?
4-day program: n1=18
2-day program: n2=16
 Idea 3: Hypotheses testing(HT): Is there a
significant difference between two physical therapy
programs (at 5% sign.)?
 Example
 Idea 3: Hypotheses testing(HT): Is there a
significant difference between two physical therapy
programs (at 5% sign.)?
4-day program: n1=18
2-day program: n2=16
 Step1:
 Parameter:
 H0: Ha:
 Significance level α =_______
 Example
 Idea 3: Hypotheses testing(HT): Is there a
significant difference between two physical therapy
programs (at 5% sign.)?
4-day program: n1=18
2-day program: n2=16
 Step2:
 Assumptions:
 Test Statistic:
Hypothesis Test for Population
Difference in Means (H0: μ1 – μ2 =0)
 Case 1: Two independent Populations –
Population Variances Known
Attributes Test Statistic Confidence Interval
 Example
 Idea 3: Hypotheses testing(HT): Is there a
significant difference between two physical therapy
programs (at 5% sign.)?
4-day program: n1=18
2-day program: n2=16
 Step3: Decision Rule:

Reject H0 if __________________
Rejection Rule

H0: μ=μ0(p=p0)

Ha: μ>μ0(p>p0)
Ha: μ≠μ0(p≠p0) Ha: μ<μ0(p<p0)
upper-sided
two-sided lower-sided
Reject H0
Reject H0 Reject H0
if Z≥ Z1-α
if Z≥ Z1-α/2 (t≥ t1-α/2,df) if Z≤- Z1-α
(t≥ t1-α, df)
or Z≤- Z1-α/2 (t≤- t1-α/2,df ) (t≤- t1-α,df)
 Example
 Idea 3: Hypotheses testing(HT): Is there a
significant difference between two physical therapy
programs (at 5% sign.)?
4-day program: n1=18
2-day program: n2=16
 Step4: Decision:

Reject H0 Fail to Reject H0

 Example
 Idea 3: Hypotheses testing(HT): Is there a
significant difference between two physical therapy
programs (at 5% sign.)?
4-day program: n1=18
2-day program: n2=16
 Step5: Conclusion:
Based on two independent samples of n1=___ and n2=___, there
____significant evidence to conclude, at α = _____, that Physical
therapy 4-day program is different than 2-day program, on
average.
 Hypothesis Test for Population
Difference in Means (H0: μ1 – μ2 =0)
 Case 2:

Attributes Test Statistic Confidence Interval

n1≥30, n2 ≥30

n1<30 or n2 <30

df=n1+n2-2

Common Standard
Deviation Estimate
Example
 To improve medical students' analytic skills a
new curriculum implemented across medical
school.
 New curriculum may be differentially effective
among male and female students.
 To evaluate, random sample of male and female
students who completed the new curriculum are
selected and given a test to assess their analytic
skills.
Example
 Question: Is there a significant Difference in Mean
Number of problems Solved Between Males and
Females?
 Male: n1=15
 Female: n2=12
 Step1:
 Parameter:
 H0: Ha:
 Significance level α =_______
Example
 Question: Is there a significant Difference in Mean
Number of problems Solved Between Males and
Females?
 Male: n1=15
 Female: n2=12
 Step 2:
 Assumptions:
 Test Statistics:
 Hypothesis Test for Population
Difference in Means (H0: μ1 – μ2 =0)
 Case 2:

Attributes Test Statistic Confidence Interval

n1≥30, n2 ≥30

n1<30 or n2 <30

df=n1+n2-2

Common Standard
Deviation Estimate
Example
 Question: Is there a significant Difference in Mean
Number of problems Solved Between Males and
Females?
 Male: n1=15
 Female: n2=12
 Test Statistics: t = 4.43
 Step 4: Decision Rule:
Reject H0 if _______________
Example
Question: Is there a significant Difference in Mean
Number of problems Solved Between Males and
Females?
 Male: n1=15
 Female: n2=12
 Step 3: Decision:
Reject H0 Fail to Reject
Example
Question: Is there a significant Difference in Mean
Number of problems Solved Between Males and
Females?
 Male: n1=15
 Female: n2=12
 Step 5: Conclusion:
Based on two independent samples of n1=___ and
n2=___, there ____significant evidence , at α = _____,
to conclude that ______________________________