Research Methods

Which of the following describes the marketing research hypothesis?

A marketing research hypothesis is a testable, tentative statement about an expected relationship. Learn how to spot one and how it differs from a question.

On this page
  1. Short answer
  2. What a marketing research hypothesis is
  3. Null and alternative hypotheses
  4. How a hypothesis is tested
  5. Hypothesis, question, objective or “educated guess”?
  6. How to spot the correct description in a multiple-choice question
  7. A worked example
  8. When marketers use hypotheses

Short answer

A marketing research hypothesis is a testable, tentative statement about an expected relationship or difference between things the research measures, such as a price and purchase intent. It is a prediction the study will check, so it can turn out to be wrong. When a multiple-choice question asks which description fits, look for the option that describes a statement that can be tested and that predicts an outcome. Options that describe a question, a broad goal or a fact that is already known do not fit.

What a marketing research hypothesis is

A hypothesis has three features:

  • It is a statement, not a question. It says what the researcher expects to find.
  • It is tentative. It is a provisional idea, held until the evidence says otherwise.
  • It is testable. The researcher can collect data that supports or contradicts it.

Example: “Customers who receive a reminder email are more likely to reorder within a month than customers who do not.” This statement names two groups, predicts a difference, and can be checked against purchase records.

Null and alternative hypotheses

In quantitative research, a hypothesis is usually written as two statements:

  • The null hypothesis (H0) says there is no difference or relationship. Researchers usually set it up so that the data can reject it.
  • The alternative hypothesis (H1), sometimes called the research hypothesis, states the difference or relationship the researcher expects.

For the reminder-email example, H0 would be “reminder emails have no effect on the chance of reordering”, and H1 would be the statement above. If the data gives strong evidence against H0, the researcher has support for H1. Note that the data supports the alternative only indirectly, by rejecting the null.

How a hypothesis is tested

A hypothesis is tested in four steps:

  1. State the hypothesis before collecting data.
  2. Choose a design and a sample that can show whether the predicted difference or relationship exists.
  3. Collect the data and compare the results with what the hypothesis predicted.
  4. Accept or reject the hypothesis, and report how confident the researcher is in that conclusion.

Experiments and split tests are a common way to test causal hypotheses, because they change one thing while holding others steady. For an overview of causal, descriptive and exploratory designs, see types of market research.

Hypothesis, question, objective or “educated guess”?

These terms are related, but they are not the same:

  • A research question asks what the study should answer, such as “why do customers stop reordering?”
  • A research objective is the aim of the study, such as “to understand why reorders have fallen.”
  • A hypothesis is a testable prediction, such as “customers who do not join the loyalty scheme in their first month are less likely to reorder.”
  • An educated guess is a common everyday description of a hypothesis. It is accurate in spirit, but a hypothesis must also be testable, which the phrase alone does not say.

Some marketers avoid the word “hypothesis” and use “research objective” instead, especially in qualitative work, where a formal hypothesis plays a smaller role.

How to spot the correct description in a multiple-choice question

When you see several options, check each one against this list:

  • Does it describe a statement that predicts a relationship or difference? If not, it is probably describing a question or an objective.
  • Does it say, or imply, that the statement can be tested or shown to be wrong? A hypothesis must be testable.
  • Does it describe a specific expected outcome, rather than a general description of the market? “Sales are high” describes the market. “Sales will rise after the price change” is a hypothesis.
  • Does it describe something already proven, such as a fact or a confirmed result? A hypothesis is not yet confirmed.

If two options seem close, prefer the one that names what is expected and how it could be checked.

A worked example

A business wants to know whether a new packaging design affects sales.

  • Problem: Sales of the product have been flat since the packaging changed. (This describes the business situation, not a hypothesis.)
  • Research question: Does the new packaging change how likely shoppers are to buy the product?
  • Hypothesis: Shoppers who see the new packaging are more likely to pick up the product than shoppers who see the old packaging.
  • Null hypothesis: The packaging design has no effect on whether shoppers pick up the product.

The hypothesis is testable because the study can compare pick-up rates between the two designs. If the data shows no clear difference, the business can keep the null hypothesis and look at other explanations.

When marketers use hypotheses

Hypotheses are most important in quantitative and causal research, where the study is designed to test a specific relationship. Exploratory and qualitative research usually starts with questions and objectives, and forms hypotheses later, as themes emerge. For how questions and hypotheses fit together from the start of a project, see marketing research problems, questions and hypotheses.

The complete guide to research methods

Types of Market Research: Methods, Designs and When to Use Each
Read the guide

Keep exploring

Guides from other topics

All topics →