Scribd
Upload
7 views

10_Comparing_two_population_means_example--ANSWER_KEY

A study comparing the average annual salaries of college graduates and high school graduates found that college graduates earn an average of $36,000, while high school graduates earn $29,000, indicating a difference of $7,000. A two-sample t-test showed a significant result with a p-value of 0.0025, leading to the rejection of the null hypothesis and suggesting that college graduates have a higher average salary. A 95% confidence interval for the difference in mean salaries was calculated to be between $2,122 and $11,878, providing further evidence of the salary disparity.

Uploaded by

chat.services2
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
7 views

10_Comparing_two_population_means_example--ANSWER_KEY

A study comparing the average annual salaries of college graduates and high school graduates found that college graduates earn an average of $36,000, while high school graduates earn $29,000, indicating a difference of $7,000. A two-sample t-test showed a significant result with a p-value of 0.0025, leading to the rejection of the null hypothesis and suggesting that college graduates have a higher average salary. A 95% confidence interval for the difference in mean salaries was calculated to be between $2,122 and $11,878, providing further evidence of the salary disparity.

Uploaded by

chat.services2
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

Comparing two population means

The formula for the standard error of the sample means:

For two independent random variables X and Y, if T = X ± Y,


then σ2T = σ2X + σ2Y.

sT =!𝜎!" + 𝜎#"
A study was done to determine whether people who go to college have a higher
average annual salary than those who do not. The study involved 202 randomly
selected 32-year-old college graduates and 190 randomly selected 32-year-olds
who were graduated from high school only. Here are the results of the study:

College High School


Mean
$36,000 $29,000
Income
St. dev.
$27,000 $22,000
Income
n 202 190

𝑥̅! − 𝑥̅"# = $36,000 - $29,000 = $7,000

Does this give us any evidence that the college graduates have a higher average
annual salary than that of high school graduates? *
Yes. This is some evidence that college graduates have a higher average annual
salary.
Could the difference in outcome have happened by chance alone? Or is there
convincing evidence that people who are graduated from college have a higher
average annual salary than those who are graduated from high school only?
Name of test Two-sample t test for µ1 − µ2 .
Define µ$ = the mean annual salary among all 32-year-old college
parameters: graduates.

µ%& = the mean annual salary among all 32-year people who
were graduated from high school only.

Null and 𝐻' : 𝜇! = 𝜇"# OR 𝐻' : 𝜇! − 𝜇"# = 0


alternative 𝐻( : 𝜇! > 𝜇"# 𝐻( : 𝜇! − 𝜇"# > 0
hypothesis
and α level: * We choose α = 0.05 (following convention)
Check Random: The data come from two independent random
conditions samples.
(difference in Independent (10% condition): When sampling without
population replacement, check that n1 ≤ N1/10 and n2 ≤ N2/10.
means): *
Normal/Large Samples:
• Both population distributions (or the true distributions of
responses to the two treatments) are Normal
• Both sample sizes are large (n1 ≥ 30 and n2 ≥ 30)
• If a population (treatment group) has unknown shape
and the sample size is small (n < 30), we can use the t-
procedure if the graph of the sample data does not show
strong skewness or outliers.

Here, nc = 202 ≥ 30 and nHS = 190 ≥ 30. So, we can safely use a t
procedure.
Check Random: The data came from two groups in a randomized
conditions experiment. The two treatments were randomly assigned.
(experiment): Independence/10% condition: Not necessary to check 10%
condition in case of experiments. The random assignment
allows us to view these two groups as independent. We assume
the subjects acted independently.
Normal/Large Samples: Same as above.
Test statistics, (x1 − x 2 ) − ( µ1 − µ2 )
P-value t=
s12 s2 2
+
n1 n 2
x C = 36,000 x HS = 29,000
sC = 27,000 sHS = 22,000
nC = 202 nHS = 190
€
* = __

Using technology, df = __ ßGet this from calculator

P-value = P( x C ̶ x HS ≥ __) = P(t ≥ __) = 0._


Casio Go to module 2 (Statistics) à TEST (F3) à t (F2) à 2-
calculator SAMPLE (F2). Calculator input: µ1 > µ2
syntax x 1 = 36,000 x 2 = 29,000
s1 = 27,000 s2 = 22,000
n1 = 202 n2 = 190 Pooled: Off
Press EXE, which returns
t=
p=
df =
tcdf(lower: 2.82, upper: ∞, df: 382.32)
Decisions and Because the P-value (0.0025) is less than α (0.05), we
Conclusion: * reject H . There is convincing evidence that college
0
graduates have a higher average annual salary than that
of high school graduates.
Interpretatio Assuming that the average annual salaries are the same
n of P-value for college graduates and high school graduates, there
is approximately 0.0025 probability of getting a sample
difference of ($36,000 - $29,000 =) $7,000 or more by
random chance alone.
Question Construct and interpret a 95% confidence interval for the
difference in mean incomes between those who were
graduated from college and those who were graduated from
high school only.
Name of Two-sample t-interval for µC − µ HS .
procedure
Confidence 95%
level
Define Same as test
parameters
Check Same as test
conditions
Calculations Option 1: Using technology: df = 382.32
invT(area in left tail: 0.025, df: 382.32) reveals t* = 1.966
Option 2: df = 190 – 1 = 189 (conservative) à t* = 1.973
(𝑥̅! − 𝑥̅"# ) ± 𝑡 ∗ (𝑆𝐸)

x C = 36,000 x HS = 29,000
sC = 27,000 sHS = 22,000
nC = 202 nHS = 190

*
Note: Reporting the t* value is optional; students can just put “t*” in the
formula.
7000 ± 1.966 ∗ (2484) = ($2122, $11,878)
Interpretation We are 95% confident that the interval from __ to __
captures the true difference in the mean salary (college
graduates - high school graduates) at age 32.
Observation The entire confidence interval is greater than 0.
Therefore, there is convincing evidence that college
graduates have a higher average annual salary than that
of high school graduates.

Study: Zimmerman, Seth. “The Returns to College Admission for Academically Marginal
Students.” Journal of Labor Economics, 2014, vol. 32 (4).
https://www.jstor.org/stable/10.1086/676661?seq=1

Material adapted from the Skew The Script curriculum (skewthescript.org)


Scratch:

(%&,((()"*,((())(
𝑡= ! !
=__
,!",$$$ -!!,$$$
!$! %&$

!",$$$! !!,$$$!
(36,000 − 29,000) ± 1.966, +
!$! %&$

You might also like