What Is an Alternative Hypothesis? (Definition and Examples)
By Indeed Editorial Team
Published May 31, 2022
The Indeed Editorial Team comprises a diverse and talented team of writers, researchers and subject matter experts equipped with Indeed's data and insights to deliver useful tips to help guide your career journey.
Forming a hypothesis is usually the first step of any research-based project or enquiry. Null and alternate hypotheses are two essential types that can help researchers establish whether to accept or reject certain claims. Learning about the concept of the alternative and null hypotheses can help you understand their importance and how to use them during research projects.
In this article, we define an alternative hypothesis, explain its differences and similarities with a null hypothesis, discuss different types of alternate hypotheses, and share some examples.
What is an alternative hypothesis?
To understand the meaning of an alternative hypothesis, it's essential to learn about the concept of the null hypothesis. A null hypothesis is an educated guess or working statement that you expect to be true. It proposes that there's no relationship between an independent variable and a dependent variable.
Researchers and statisticians usually consider the null hypothesis to be true unless there's sufficient proof to reject it. They test the hypothesis to accept or reject with a certain degree of accuracy and certainty.
The alternate hypothesis proposes an opposite working statement to the null hypothesis. If the null hypothesis proposes that a statement or relationship is true, the alternate hypothesis considers it to be false. When testing the null hypothesis, researchers and statisticians usually test the validity of the alternate hypothesis.
Once there's sufficient evidence to support the alternative hypothesis, it replaces the null hypothesis. Both alternative and null hypotheses are useful when conducting research in medicine, technology, science, psychology, mathematics, and statistics.
What are the differences between the alternative and null hypotheses?
Here are some ways in which both these hypotheses are different:
The method of describing the alternative and null hypotheses and their symbols differ. Ha or H1 is the symbol for the alternative hypothesis, whereas it's Ho for the null hypothesis. Researchers test Ho to determine whether they can accept or reject it. In case they find the Ho to be correct, instead of saying that the statement is true, they say that they cannot disprove it. An "equals to" sign follows the Ho whereas the "less than" or "greater than" sign follows the Ha. These symbols help explain their position and if the test data accepts or rejects the Ho.
The primary assumptions of the alternative and null hypotheses differ. A null hypothesis usually states that there's no relationship between two variables or sets of data. "Null," in this regard, signifies no change between two variables. Conversely, an alternate hypothesis proposes a working statement that can be opposite to or different from the null hypothesis. It claims that random factors and causes may influence the data.
An alternate hypothesis contradicts the proposal of the null hypothesis. It implies that the truth is either different from the null hypothesis or opposite to it. Researchers test to find errors or fallacies to determine whether they can find evidence to accept or reject the null hypothesis using the alternate hypothesis as a guide. As they're mutually exclusive, only one of them can be true. After testing, researchers may reject the null hypothesis and replace it with the alternate hypothesis or reject the alternate hypothesis and choose a new one for testing.
Researchers accept the null hypothesis when the p-value is greater than the level of significance. When the p-value is smaller than the level of significance, they accept the alternate hypothesis.
What are the similarities between alternative and null hypotheses?
There are some ways in which the alternate and null hypotheses are similar:
Both the hypotheses intend to propose a working statement and help researchers determine whether it's true or false using basic testing methods and assumptions. Researchers use both statements to help guide their tests and research methodology. The purpose of conducting tests is to find whether the data supports or rejects the claims of the hypotheses.
The null hypothesis helps verify or disprove a statistical assumption or advance a theory. It can also help verify the consistency of results between different experiments. Similarly, an alternate hypothesis helps researchers verify certain statements or assumptions that can help provide direction for the testing methodology. It allows the opportunity to discover new theories by disproving existing ones.
Even in cases where the research may not test the central problem or challenge, they can still test both the alternate and the null hypothesis. Both the statements are descriptive and propose a specific outcome based on certain information. The hypotheses have one and two-side testing capabilities, and their application can help test different research purposes.
Different types of alternative hypotheses
Here are two common types of alternate hypotheses that researchers use:
One-tailed directional hypothesis
This type of alternate hypothesis usually tests only one direction of the parameters and value. For instance, tests can only determine whether the difference is greater than, or less than, zero, but not simultaneously. There are two types of one-tailer directional alternate hypotheses:
Left-tailed: In this case, the sample proportion is less than the specified value.
Right-tailed: In this case, the sample proportion is greater than the specified value.
Two-tailed or non-directional hypothesis
This type of alternate hypothesis proposes that the parameters don't equal the null hypothesis. This means that the alternate hypothesis states that there are differences that are both greater than, and less than, the null value. This doesn't show the direction of difference between the two hypotheses, but simply that one exists and all that the researcher can gather is that the sample proposition isn't equal to the specified value.
Examples of alternative and null hypotheses
Here are two easy-to-understand examples of alternative and null hypotheses:
One-sided alternate hypothesis
Say that the executives of an organization want to test the efficacy of their hiring process. They believe that the company only shortlists candidates with relevant professional experience. Based on this assumption, they formulate this null hypothesis:
Job candidates who have at least five years of professional work experience are more likely to receive an invitation for an interview.
This null hypothesis is one-sided, which means it claims that the trend or relationship is evident only in one direction. The executives expect this number to be greater than zero. In contrast to the null hypothesis, they also create an alternate hypothesis. As the alternate hypothesis claims that the opposite of the null hypothesis is true, it says:
Job candidates without any years of professional work experience are as likely to receive an invitation for an interview as those with at least five years of experience.
In this example, the executives can create a test to calculate the years of work experience of all candidates who receive an invitation for the interview. They can collect relevant data, analyze it, and determine whether to accept or reject the initial claim. If the test is unable to prove the alternate hypothesis to be true, the executives can say that they cannot reject the null hypothesis. If they collect adequate evidence to establish that it isn't true, they conclude to reject the initial null hypothesis.
Some school students are competing in an advanced province-wide assessment. A researcher at the school claims that, due to the additional classes and training offered by the school, the average score among students would be higher than the provincial average. Based on this, the researcher creates the following null hypothesis:
The average test scores of students from the school will be higher than that of the province's average score by 100 points.
To test this, they also create an alternate hypothesis that claims the opposite is true:
The average test score of students from the school will be similar to that of the province's average score. The additional classes and training have little to no impact on the performance of the students and no correlation between the two different test scores exists.
In this example, the researcher uses a two-sided hypothesis. This is because the researcher wants to establish whether the average score of the school students is higher or lower than the provincial average. This test can help determine if the extra classes and training offered by the school have been beneficial or not. As per the findings, the researcher can accept or reject the initial null hypothesis and determine the direction of the relationship between the two variables, or here, the two sets of scores.
Explore more articles
- The Know Your Customer Process (KYC) and Its Importance
- What Is a Hiring Freeze? (With Effects, Benefits, and FAQs)
- 10 Good Employee Qualities and Their Importance To You
- How Employee Performance Is Measured and Why It's Important
- Functional Organization Structure: Pros and Cons
- How to Improve Your Skill Set and Keep Skills Current
- What Is a Logistics Company? (With Types and Benefits)
- What Is Relational Leadership? (With Principles and Tips)
- 7 Negotiation Courses to Explore (With Negotiation Styles)
- What Are Examples of Work Objectives? (And Their Categories)
- How to Deal with a Passive-Aggressive Boss (With Tips)
- How to Establish a High-Performance Culture as a Leader