Scribd
Upload
4 views3 pages

Hypothesis Testing Guide

This document provides a step-by-step guide to hypothesis testing when the population standard deviation is unknown. It outlines the use of T and Z statistics based on sample size, details procedures for single sample tests, two independent samples, and paired samples, and explains how to determine the significance of results using P-values. The final decision-making process involves comparing the P-value to a significance level to either reject or fail to reject the null hypothesis.

Uploaded by

6alalush8
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
4 views3 pages

Hypothesis Testing Guide

This document provides a step-by-step guide to hypothesis testing when the population standard deviation is unknown. It outlines the use of T and Z statistics based on sample size, details procedures for single sample tests, two independent samples, and paired samples, and explains how to determine the significance of results using P-values. The final decision-making process involves comparing the P-value to a significance level to either reject or fail to reject the null hypothesis.

Uploaded by

6alalush8
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

Step-by-Step Guide to Hypothesis Testing

Process for Unknown Population Standard Deviation (σ)

Part 1: The Golden Rule (Z vs T)


When the population standard deviation (σ) is **unknown**, you must use the
sample standard deviation (s) as an estimate. This forces a choice:
• Use the T-Statistic: If Sample Size (n) is Small (< 30).
• Use the Z-Statistic (Approximation): If Sample Size (n) is Large (≥ 30).
Note: In modern statistics, using T is always correct when σ is unknown, but
many textbooks use Z for large samples because the T-distribution looks exactly
like the Z-distribution when n is large.

Part 2: The Recipes


Case A: Single Sample Test
Scenario: You have one group of data (e.g., ”30 students”) and you want to com-
pare their average to a specific number (e.g., ”The national average is 75”).
Step 1: Gather Data
• Sample Mean (x̄)
• Target Mean (µ0 )
• Sample Standard Deviation (s)
• Sample Size (n)
Step 2: Choose Formula
x̄−µ
If n < 30 (T-Test): t= √0
s/ n
x̄−µ
If n ≥ 30 (Z-Test): z= √0
s/ n

Step 3: Get the P-Value


• For Z: Look up your calculated Z-score in a Z-Table. The number inside the
table is the probability (area) to the left.
• For T: Look up your calculated T-score in a T-Table. You need the ”Degrees
of Freedom” (df = n − 1). Find the row for your df and see where your
T-score falls.

1
Case B: Two Independent Samples
Scenario: You have two totally separate groups (e.g., ”Men vs. Women” or ”Treat-
ment vs. Placebo”). You want to know if their averages differ.
Step 1: Gather Data

• Group 1: Mean (x̄1 ), SD (s1 ), Size (n1 )

• Group 2: Mean (x̄2 ), SD (s2 ), Size (n2 )

Step 2: Choose Formula The logic is the same: if both samples are large
(> 30), use Z. If either is small, use T.
(x̄ −x̄ )
The Statistic (z or t) = r1 2
s2 2
1 + s2
n1 n2

Step 3: Get the P-Value

• For Z: Look up the score in the Z-Table.

• For T: Use a T-Table. For degrees of freedom, the simple estimate is to use
the smaller sample size minus 1 (df = smaller n − 1).

Case C: Paired (Dependent) Samples


Scenario: You measure the same people twice (e.g., ”Before training vs. After
training”). The data points are linked.
Step 1: Calculate Differences Do not use the two group means separately!
You must calculate the difference for every single person.

• Person A: After - Before = Difference A

• Person B: After - Before = Difference B

• ...and so on.

Step 2: Gather Data on Differences Now you have a new list of numbers
(the differences).
¯ The average of those differences.
• d:

• sd : The standard deviation of those differences.

• n: The number of pairs.

Step 3: Choose Formula This is now exactly the same as a Single Sample test
(Case A), but comparing against 0 (zero difference).
¯
d−0
If n < 30 (T-Test): t= √
sd / n
¯
d−0
If n ≥ 30 (Z-Test): z= √
sd / n

2
Part 3: The Final Decision
Once you have your P-value from the table:

1. Check your Significance Level (α). Usually, this is 0.05 (5%).

2. Compare:

• Is P-value < 0.05? → Reject Null Hypothesis. (There is a real differ-


ence).
• Is P-value > 0.05? → Fail to Reject. (It could just be luck).

You might also like