T-DISTRIBUTION

WEEK 13
(MARCH 23-27, 2020)
This module is made to help you be mathematically equipped
and enhance your skills in Statistics and Probability despite of
the current situation of our country. Our main objective here is
for you to still continue learning, even with the absence of
classroom environment. Each lesson is introduced and discussed
with numerous examples for mastery. Aside from that, the
following components are included to serve as your guide in
your journey to learning.

LEARNING This contains a list of desired results or

behavioral goals that you are expected to
OBJECTIVES do at the end of the lesson.

Presenting prior knowledge before the CHALLENGE

concepts allows you to connect it to the
lessons that are to be discussed. YOUR MIND

DEFINITION &
CONCEPTS IN The definition and important concepts are
highlighted in solid boxes for emphasis
SOLID BOXES

PRACTICE EXERCISES
At the end of each discussion, a set of questions and activities is
presented. It allows you to master the and apply the concepts and process
in solving problems.

SEATWORK
This part contains a set of problems (usually lifted from your book)
covering mathematical concepts discussed in each lesson.
LEARNING OBJECTIVES
At the end of the lesson, the you are expected to….
o illustrates the t-distribution;
o constructs a t-distribution;
o identifies regions under the t-distribution
corresponding to different t-values;
o identifies percentiles using the t-table; and
o solves problems involving t-distribution.
CHALLENGE YOUR MIND
In the past lesson, you were acquainted to z-distribution in
dealing with problems involving normal distributions. This present
lesson introduces another distribution called the t-distribution.
What do you know about the t-distribution? Assess the
correctness of each statement below.

1. The range of t-distribution is -3 to 3.

2. The t-distribution has the same shape as the standard
normal distribution.
3. The t-distribution has a mesokurtic shape.
4. The t-distribution is more dispersed than the standard
normal distribution and when the sample size 𝑛
increases, it assumes the normal distribution.
5. The shape of the t-distribution depends on the
degrees of freedom.

Read the discussion part so you can assess whether

your assumptions about t-distributions are correct.

Barcelona, A. B., & Reynoso, L. C. (2019). Statistical Estimation. In Statistics and

Probability For Senior High School (p. 137). Golden Cronica Publishing, Inc.)
There are situations when sample values are not large enough
for the Central Limit Theorem to be applied.
What if it is impossible to include a sufficiently large sample in
drawing inferences about the population parameter? Can we
still obtain an interval estimate of the population mean and use
the z-distribution? The answer is yes, but there is a distribution
that is more appropriate to use when sample size is small.

The t-distribution which is also called

student’s t-distribution is similar to the z-
distribution as both of them have bell shaped
curve but it has a heavier tails.

Observe that the smaller 𝑛, the fatter its tails are. This happens
because the standard deviation of the t-distribution is
proportionally larger than z.
How is the t-distribution SIMILAR to the z-distribution?
1. Both distributions contain the transformation of
each individual score into standardized scores.
2. Both distributions have graphs which are bell-
shaped and symmetrical about the mean.
3. Both distributions can be used to calculate the
confidence interval of the difference of two
means.

How is the t-distribution DIFFERENT from the z-distribution?

1. The t-distribution is more variable since the t-

values depend on the variations of the mean
and the standard deviation while the z-values
depends on the fluctuations of the mean from
sample to sample.
2. While the graphs of both distributions are bell-
shaped, the graph of t-distribution is flatter at
the center and has heavy tails.
 The t-distribution is used when 𝑛 < 30 and the
sample standard deviation is known.
The value of t is given below:
𝑥−𝜇
𝑡= 𝑠
𝑛

The t-values and Region under the t-distribution

Unlike in a z-distribution, finding the region under the t-
distribution requires understanding of degree of freedom (df).

The degree of freedom is equal to the sample size minus 1.

𝑑𝑓 = 𝑛 − 1

The process of finding the region under the t-

distribution is similar to finding the area of a z-score
in the standard normal distribution. The column at
the right indicates the degree of freedom while the
upper rows shows the level of significance that
varies depending whether it for directional or non-
directional test.
o A ONE-TAILED TEST is a DIRECTIONAL TEST
with the region of rejection lying on either left or
right tail of the normal curve. For this test, the
level of significance of 𝛼 = 0.05 is allotted on
of the curve depending on the given situation. If
𝛼 = 0.05, then one end will still have a value of
𝛼 = 0.05.
o A TWO-TAILED TEST is a NON-DIRECTIONAL
TEST with the region of rejection lies on both tails
of the normal curve. For this test, half of 𝛼
(significance level) is allotted on both sides of the
curve. If 𝛼 = 0.05, then both ends will have a
𝛼 0.05
value of = = 0.025.
2 2
 t-distribution Table

Taken from: https://web.standford.edu/dept/radiology/cgi-bin/classes/stats_data_analysis/principles/t_dist.gif

Task 1:
Find the t-value with 10 degrees of freedom and
level of significance of 0.1. Consider the non-
directional test since no other details are given.

Solution:
To find 𝑡0.1,10′ , we have to consider the value that is located
in the intersection of the degrees of freedom in the column,
and the value of the α in the row. We can see the identified
value is 1.8120.
Task 2:
What is the critical value 𝑡 for constructing a
95% confidence interval for a mean of sample
size 𝑛 = 18 observation?

Solution:
Finding the degree of freedom, 𝑑𝑓 = 𝑛 − 1 = 18 − 1 = 17.
To find the significance level, simply subtract the 95% (0.95)
to 100% (1). Which gives 𝛼 = 0.05 for a non-directional
95% confidence interval, the critical value 𝒕 = 𝟐. 𝟏𝟏𝟎.

Note that we uses non-directional test in this task because

there is no specific given side of the curve.
PRACTICE EXERCISE #1

Find the critical values of t.

1. 𝑛 = 6 , 𝛼 = 0.10, one-tailed

2. 𝑛 = 7, 𝛼 = 0.10, two-tailed
n

3. 𝑛 = 12, 95% confidence

4. 𝑛 = 17, 95% confidence

5. 𝑛 = 24, 99% confidence

Calculating Confidence Interval for
Population Mean using t-distribution
The process and formula in calculating the confidence
interval for the population mean using 𝑡 − and 𝑧 −
distribution are comparably the same.

If the population standard deviation is unknown with a

confidence level of (1 − 𝛼) 100, then the confidence
interval can be calculated using the estimator below:

𝑠 𝑠
𝑥− 𝑡 𝛼
,𝑛−1
( ) <𝜇<𝑥+ 𝑡 𝛼
,𝑛−1
( )
2 𝑛 2 𝑛

where

𝒙 is the sample mean

𝒕𝜶,𝒏−𝟏 is the t-value with n-1 degrees

𝛼
of freedom
𝟐 that leaves an area of to the right
2

𝒔 is the sample standard deviation

𝒏 is the sample size

Task 3:
An admission officer of an educational institution
wants to know the mean age of all entering
mathematics majors. He computed a mean age of
18 years and a standard deviation of 1.2 years
on a random sample of 25 entering mathematics
majors purportedly coming from a normal
distributed population. With 99% confidence,
find the interval estimate of the population mean.

Solution:
Solving problems like this can be done by following some steps.

Step 1: Determine the confidence coefficient or the

critical value of t.

From the t-Table, for 𝑛 = 25, 𝑑𝑓 = 24. The 99% (or

0.99) confidence coefficient for this value from the table
is 𝑡 = 2.797.
 𝑠
Step 2: Find the maximum error (E) or the 𝑡𝛼,𝑛−1 ( ) .
2 𝑛
The maximum error is the value that you’ll add/
subtract to the sample mean to find the intervals.

𝑠 1.2
𝐸 = 𝑡𝛼,𝑛−1 ( ) = 2.797 ( )
2 𝑛 25

= 2.797 0.24
𝑬 = 𝟎. 𝟔𝟕𝟏𝟐𝟖 / 𝟎. 𝟔𝟕

Step 3: Find the lower and the upper

confidence limits.
𝑠 𝑠
𝑥 − 𝑡𝛼,𝑛−1 ( ) < 𝜇 < 𝑥 + 𝑡𝛼,𝑛−1 ( )
2 𝑛 2 𝑛

18 − 0.67 < 𝜇 < 18 + 0.67

17.33 < 𝜇 < 18.67

So, the lower and the upper confidence limits range

from 17.33 to 18.67.
Step 4: Describe the results.

Thus, we can say with 99% confidence that the

interval between 17.33 and 18.67 contains the
true mean age the population of entering
mathematics majors based on the sample of 25
entering mathematics majors.

Had any question or clarification

about this task? Feel free to
message your teacher, they’ll be
glad to answer you.
Task 4:
The mean time, in seconds, per commercial
advertisement on 8 local networks were recorded
and assumed to be approximately normally
distributed. If the mean time per commercial
advertisement was recorded to be 43.25 seconds
with a standard deviation of 5.1, find a 99%
interval for the mean time of commercial
advertisements by the local networks.

Solution:
Solving problems like this can be done by following some steps.

Step 1: Determine the confidence coefficient or the

critical value of t.

From the t-Table, for 𝑛 = 8, 𝑑𝑓 = 7. The 99% (or 0.99)

confidence coefficient for this value from the table is
𝑡 = 3.499.
 𝑠
Step 2: Find the maximum error (E) or the 𝑡𝛼,𝑛−1 ( ) .
2 𝑛
The maximum error is the value that you’ll add/
subtract to the sample mean to find the intervals.

𝑠 5.1
𝐸=𝑡 𝛼
,𝑛−1
( ) = 3.499 ( )
2 𝑛 8

𝑬 = 𝟔. 𝟑𝟎𝟗

Step 3: Find the lower and the upper

confidence limits.
𝑠 𝑠
𝑥 − 𝑡𝛼,𝑛−1 ( ) < 𝜇 < 𝑥 + 𝑡𝛼,𝑛−1 ( )
2 𝑛 2 𝑛

43. 25 − 6.309 < 𝜇 < 43.25 + 6.309

36.941 < 𝜇 < 49.559

So, the lower and the upper confidence limits range

from 36.941 to 49.559.
Step 4: Describe the results.

Therefore, we can be 99% confident that the

interval (36.941 , 49.559) contains the true value
of the population mean of commercial
advertisements played in all local networks.
PRACTICE EXERCISE #2

Solve the following problems.

1. The mean scores of a random sample of
17 students who took a special test is 83.5.
If the standard deviation of the scores is
4.1, and the sample comes from an
approximately normal population, find the
interval estimates of nthe population mean
adopting a confidence level of 95%.

2. The average weight of 25 chocolate bars

selected from a normally distributed
population is 200 g with a standard
deviation of 10 g. Find the point and the
interval estimates using 95% confidence
level.
 SEATWORK

Taken from the book, page 148.

 SEATWORK

Taken from the book, page 149.

References:

o (Barcelona, A. B., & Reynoso, L. C. (2019). Statistical

Estimation. In Statistics and Probability For Senior High
School (pp. 141–143). Golden Cronica Publishing, Inc.)

o (Belecina, R. R., Baccay, E. S., & Mateo, E. B. (2016).

Estimation of Parameters. In Statistics and
Probability (First, pp. 162–165). Rex Publishing House.)