LESSON 54 - RESEARCH PROPOSAL || HYPOTHESES: MEANING & TYPES OF HYPOTHESES
Based on RESEARCH METHODS CLASS WITH PROF. LYDIAH WAMBUGU's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.
A hypothesis is a testable prediction about the relationship or difference between variables in a population.
Briefing
A hypothesis in social science research is a prediction about how variables relate or differ, and it functions as the backbone for hypothesis testing using inferential statistics. In this lesson, hypotheses are defined as statements that forecast either a relationship or a difference between variables—such as between independent and dependent variables, or involving moderating and intervening variables. The key practical point is that hypotheses are not just “ideas”; they are testable claims that guide what gets measured and how results are interpreted.
The lesson distinguishes between the two most commonly used hypotheses in social science: the null statistical hypothesis and the alternative hypothesis. The null hypothesis (H0, or Hnaught) asserts that there is no relationship and no difference between variables. It is typically written as “no significant relationship” or “no significant difference,” and the lesson links “relationship” to categorical variables and “difference” to continuous variables—reflecting how researchers frame hypotheses depending on the type of data.
The alternative hypothesis (H1 or Ha) claims that a relationship or difference does exist. It comes in two forms. A directional alternative hypothesis specifies the direction of the effect—meaning the relationship is expected to be positive or negative. Because direction is already stated, it aligns with a one-tail test, where evidence is checked on only one side of the distribution. By contrast, a non-directional alternative hypothesis does not specify whether the relationship will be positive or negative. That uncertainty corresponds to a two-tailed test, where evidence is evaluated on both sides.
A major workflow recommendation is how to write hypotheses in the proposal versus how to handle them during analysis. The lesson advises stating the alternative hypothesis in Chapter One in a non-directional form—so the data can reveal whether the relationship is positive or negative. Then, when analysis begins (Chapter Four), researchers typically shift to testing the null hypothesis, because hypothesis testing is structured around comparing the null and alternative claims.
Finally, the lesson explains why hypotheses matter: hypothesis testing evaluates two mutually exclusive statements about a population using sample data. Researchers use probability and inferential statistics to infer population parameters from sample statistics, determining which claim is supported by the evidence. Since the null and alternative cannot both be true at the same time, collected data becomes the basis for deciding which statement is more consistent with the observed results. The lesson closes by previewing the next sections of the research proposal (1.6 to 1.8), which will continue building the proposal structure around these testable claims.
Cornell Notes
A hypothesis is a testable prediction about how variables relate or differ in a population. The lesson focuses on two core types used in social science: the null statistical hypothesis (H0) stating no relationship/difference, and the alternative hypothesis (H1/Ha) stating that a relationship/difference exists. Alternative hypotheses can be directional (specifies positive/negative direction, leading to a one-tail test) or non-directional (no direction specified, leading to a two-tail test). In proposal writing (Chapter One), it is recommended to state the alternative hypothesis non-directionally so the data can determine the direction. During analysis (Chapter Four), testing typically centers on the null hypothesis using inferential statistics and probability to infer population parameters from sample evidence.
What does a hypothesis mean in social science research, and what does it predict?
How do the null statistical hypothesis and alternative hypothesis differ?
Why does a directional alternative hypothesis lead to a one-tail test?
Why does a non-directional alternative hypothesis lead to a two-tail test?
What is the recommended approach to writing hypotheses in Chapter One versus testing in Chapter Four?
What role do inferential statistics and probability play in hypothesis testing?
Review Questions
- In your own words, how would you distinguish a hypothesis about a relationship from one about a difference, and how does that relate to variable types?
- When would you choose a directional alternative hypothesis instead of a non-directional one, and what tail test does each require?
- Why does hypothesis testing treat the null and alternative hypotheses as mutually exclusive, and how does that affect the logic of inference?
Key Points
- 1
A hypothesis is a testable prediction about the relationship or difference between variables in a population.
- 2
The null statistical hypothesis (H0) claims no relationship or no difference between variables.
- 3
The alternative hypothesis (H1/Ha) claims that a relationship or difference exists and can be directional or non-directional.
- 4
Directional alternatives specify the direction of the effect and correspond to a one-tail test; non-directional alternatives do not specify direction and correspond to a two-tail test.
- 5
Proposal writing typically uses a non-directional alternative hypothesis in Chapter One so the data can determine the direction.
- 6
Hypothesis testing uses inferential statistics and probability to infer population parameters from sample statistics.
- 7
Null and alternative hypotheses are treated as mutually exclusive, so evidence from the sample supports one claim over the other.