Nothing Special   »   [go: up one dir, main page]

How To Find Percentiles For A T-Distribution: Statistics For Dummies, 2nd Edition

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 5

HOW TO FIND PERCENTILES FOR A T-

DISTRIBUTION

RELATED BOOK

Statistics For Dummies, 2nd Edition

When you want to find percentiles for a t-distribution, you can use the t-table. A percentile is a number on a statistical
distribution whose less-than probability is the given percentage; for example, the 95th percentile of the t-distribution
with n– 1 degrees of freedom is that value of

whose left-tail (less-than) probability is 0.95 (and whose right-tail probability is 0.05).

The t-table shows right-tail probabilities for selected t-distributions. You can use it to solve the following problems.

Suppose you have a sample of size 10 and you want to find the 95th percentile of its corresponding t-distribution. You

have n – 1= 9 degrees of freedom, so, using the t-table, you look at the row for df = 9. The 95th percentile is the

number where 95% of the values lie below it and 5% lie above it, so you want the right-tail area to be 0.05. Move

across the row, find the column for 0.05, and you get

This is the 95th percentile of the t-distribution with 9 degrees of freedom.

Now, if you increase the sample size to n = 20, the value of the 95th percentile decreases; look at the row for 20 – 1 =

19 degrees of freedom, and in the column for 0.05 (a right-tail probability of 0.05) you find

degrees of freedom indicate a smaller standard deviation and thus, the t-values are more concentrated about the

mean, so you reach the 95th percentile with a value of t closer to 0.


T-distribution is the most famous theoretical probability distribution in continuous family of
distributions. T distribution is used in estimation where normal distribution cannot be used to
estimate population parameters.

1. 1. t distribution (Student’s t distribution)


2. 2. In the previous discussions, it was shown that when the population is normally
distributed, or when the sample size is large enough, the sampling distribution of
the mean is normally distributed. And of course, the bell curve is very handy to
use.
3. 3. However, in many cases where we can only obtain small sizes, the normal
distribution does not hold true. Instead, we use the t distribution which is the
distribution of t-scores.
4. 4. t = t score x = sample mean 𝝁 = population mean 𝑺 = sample standard
deviation 𝒏 = sample size
5. 5. EXAMPLE 1 Find the t-score for a sample size of 16 taken from a population
with mean 10 when the sample mean is 12 and the sample standard deviation is
1.5.
6. 6. Degrees of freedom - the number of observations in a data set that are free to
change without changing the mean. For a single group test 𝑑𝑓 = 𝑁 − 1 For 2-
group tests 𝑑𝑓 = 𝑁1 + 𝑁2 − 2
7. 7. PROPERTIES of a t distribution 1. The distribution has mean 0. 2. The
distribution is symmetric about the mean. 3. The variance is equal to 𝑑𝑓 𝑑𝑓−2 . 4.
The variance is always greater than 1, but approaches 1 when df gets bigger.
8. 8. What if n approaches infinity? The t distribution also approaches the standard
normal distribution.
9. 9. Increasing the sample size will….
10. 10. Increasing the degrees of freedom will…
11. 11. Increasing the degrees of freedom and sample size will make the t
distribution approach a normal distribution.
12. 12. The critical value is the thin line between rejection and acceptance.
13. 13. The confidence interval is actually the acceptance region.
14. 14. The t TABLE 1. The critical region appears at the top. 2. The degrees of
freedom are on the leftmost section. 3. Confidence interval are at the bottom.
15. 15. EXAMPLE 2 Find the t-score below which we can expect 99% of sample
means will fall if samples of size 16 are taken from a normally distributed
population.
16. 16. EXAMPLE 2 SOLUTION 1 − 𝛼 = 0.99 𝛼 = 0.01 𝑑𝑓 = 𝑛 − 1 𝑡0.99 = −𝑡0.01 𝑑𝑓
= 15 𝑡0.99 = −𝑡0.01 = −2.602
17. 17. EXAMPLE 3 If a random sample of size 25 drawn from a normal population
gives a mean of 60 and a standard deviation of 4, find the range of t- scores
where we can expect to find the middle 95% of all sample means.
18. 18. EXAMPLE 4 Compute the probability that 𝑃 (−𝑡0.05 < 𝑡 < 𝑡0.10).

You might also like