Hypothesis Builder
Start from your research question, objectives and variables, choose the kind of hypothesis you need, and get null and alternative drafts built only from your own details. You can edit every draft, and every check explains itself.
Research question
Paste the question from the Research Question Builder, or write it here.
What the project sets out to do.
Objectives
One objective per line.
Variables
List each variable exactly as you will name it. Enter one per line. Only the variables you list are used.
What may influence the outcome, or defines the groups you compare.
The outcomes you will measure.
Variables that may change the strength or direction of a relationship.
Variables through which one variable may affect another.
Variables you will hold constant or adjust for.
Hypothesis type
Every hypothesis has a role (null or alternative), a form and a direction. The builder always drafts a null and an alternative together. Read about each type, then choose the form and direction.
The seven hypothesis types
Null hypothesis (H₀)
Role
States that there is no effect, difference or relationship in the population. It is the statement a statistical test evaluates.
Learn more about the null hypothesis
When it is appropriate
In quantitative studies that use inferential statistical tests.
Strengths
- Gives a precise statement that a statistical test can assess.
- Guards against claiming an effect without evidence for it.
Limitations
- Failing to reject it doesn't show that there is no effect: the study may have been too small to detect one.
- On its own it says little; it matters as the counterpart of the alternative hypothesis.
Examples
- There is no difference in mean reading scores between pupils taught with and without phonics.
- There is no relationship between hours of sleep and exam performance among first-year students.
References
References and wording for this hypothesis type are awaiting academic review.
Alternative hypothesis (H₁)
Role
States that there is an effect, difference or relationship: the expectation the study sets against the null hypothesis.
Learn more about the alternative hypothesis
When it is appropriate
Whenever a null hypothesis is tested, to state what the study expects to find.
Strengths
- States the expectation clearly, linking the study to theory or previous findings.
- Together with the null hypothesis, defines what the analysis decides between.
Limitations
- Results can support it but never prove it.
- It must be stated before the data are analysed, or it loses its meaning.
Examples
- There is a difference in mean reading scores between pupils taught with and without phonics.
- Hours of sleep are related to exam performance among first-year students.
References
References and wording for this hypothesis type are awaiting academic review.
Directional hypothesis
Direction
An alternative hypothesis that states which way the effect goes, such as higher or lower, positive or negative.
Learn more about the directional hypothesis
When it is appropriate
When theory or previous evidence gives a clear reason to expect a particular direction.
Strengths
- Makes a more precise claim.
- Shows how the study builds on existing theory or evidence.
Limitations
- An effect in the opposite direction doesn't support it.
- The direction must be justified from the literature before data are collected, not chosen after seeing the results.
Examples
- Students who sleep at least seven hours a night achieve higher exam scores than students who sleep less.
- Screen time is negatively related to sleep quality among teenagers.
References
References and wording for this hypothesis type are awaiting academic review.
Non-directional hypothesis
Direction
An alternative hypothesis that states that an effect exists without saying which way it goes.
Learn more about the non-directional hypothesis
When it is appropriate
When there is no strong basis for predicting a direction, or previous findings disagree.
Strengths
- Doesn't commit to a direction the evidence can't support.
- Allows an effect in either direction to be detected.
Limitations
- Less precise than a directional hypothesis.
- If the literature does suggest a direction, a non-directional hypothesis may undersell the study's grounding.
Examples
- There is a difference in job satisfaction between remote and office-based staff.
- There is a relationship between class size and pupil attainment.
References
References and wording for this hypothesis type are awaiting academic review.
Difference hypothesis
Form
States that groups differ, or as a null hypothesis that they don't, on an outcome.
Learn more about the difference hypothesis
When it is appropriate
When the independent variable defines groups, such as treatment and control, or two schools.
Strengths
- Matches studies that compare groups, such as experiments.
- Makes the comparison explicit.
Limitations
- The groups must be clearly defined.
- A difference alone doesn't show what caused it, unless the design rules out other explanations.
Examples
- Patients who receive the new physiotherapy programme report less pain than patients who receive usual care.
- There is no difference in voter turnout between urban and rural wards.
References
References and wording for this hypothesis type are awaiting academic review.
Relationship hypothesis
Form
States that variables are related, or as a null hypothesis that they aren't.
Learn more about the relationship hypothesis
When it is appropriate
When the variables vary along a scale, such as hours of study and exam marks.
Strengths
- Suits variables measured on a scale.
- Can describe the strength and direction of a relationship.
Limitations
- A relationship doesn't show cause and effect.
- A third variable may explain why the two are related.
Examples
- Hours of paid work are negatively related to exam performance among full-time students.
- There is no relationship between age and use of online banking.
References
References and wording for this hypothesis type are awaiting academic review.
Prediction hypothesis
Form
States that one or more variables predict an outcome, or as a null hypothesis that they don't.
Learn more about the prediction hypothesis
When it is appropriate
When the aim is to forecast an outcome from measures taken earlier.
Strengths
- Suits studies that forecast outcomes.
- Makes clear which variables are predictors and which is the outcome.
Limitations
- The predictors must be measured before the outcome.
- Prediction doesn't by itself show cause.
Examples
- First-year grades predict final degree classification.
- Higher attendance in the first term predicts lower dropout by the end of the year.
References
References and wording for this hypothesis type are awaiting academic review.
The hypothesis states that an effect exists, without saying which way it goes.
Generated drafts
Edit any draft in your own words. The checks below use your edited wording.
These drafts use only the variables, population and context in your project, in the form and direction you chose. Words in square brackets are missing from your project; replace them before you use the hypotheses. They are drafts: edit them, and check them with your supervisor.
Pair 1: independent variable → dependent variable
Relationship
Evaluation
Checks describe how your hypotheses fit your project. They never score or rank.
Fit with your methodology
Methodology For you to review
Your project has no methodology yet. Hypotheses are most common in quantitative research; add your methodology to see how they fit.
Why: Whether a study needs hypotheses depends on its methodology.
Pair 1: independent variable → dependent variable
Research question alignment Missing
Add your research question, so the hypothesis can be compared with what the study asks.
Objective alignment Missing
Add your research objectives, so you can see which objective each hypothesis serves.
Variable consistency Missing
The wording still contains placeholders in square brackets. Replace them with information from your project.
Population consistency Missing
Your project doesn't name a population. A hypothesis is a claim about a population, so name who it applies to.
Context consistency Missing
Replace “[context]” with where or when the hypothesis applies, or remove it if the hypothesis applies generally.
Testability Worth checking
A testable hypothesis names at least two variables and states how they are related. Check that both are named and the relationship is stated.
Measurability For you to review
Whether each variable can be measured depends on your instruments and definitions, which ResearchKit can't see. Check that you can measure every variable named.
Logical wording Looks aligned
Each hypothesis states a single claim as a statement.
Null and alternative consistency Looks aligned
The null hypothesis states the absence of what the alternative states, about the same variables.
Direction Looks aligned
The wording matches the direction chosen (non-directional).
Your project draft
This tool writes only the Hypotheses section. Everything else is shown as you entered it.
- Hypotheses
- H₀: There is no relationship between [independent variable] and [dependent variable] among [population] [context]. H₁: There is a relationship between [independent variable] and [dependent variable] among [population] [context].
Next: Variables
Every variable with its conceptual and operational definitions, indicators and measurement level.
What should I do next?
Next in the research journey: keywords.
Literature Explorer
Search scholarly literature by topic, see which keywords and topics recur, and copy citations to start a reading list.
How to search academic literature
How to turn a topic into search terms, use phrases, synonyms and operators, widen or narrow a search, judge the results, find DOIs and free full text, and record your search so you can repeat it.
How the builder works
- The builder reads your project draft: research question, aim, objectives, variables, population, context and methodology. Enter them in the first steps; later versions will bring them across from the other research tools.
- Hypotheses are drafted only from what you entered. No variable, direction or relationship is inferred: you choose the form and direction, and every relationship is one you listed. Anything missing stays in square brackets.
- Each draft pair carries its relationship as structured data (independent, dependent, moderator, mediator and control variables), so later tools can use it directly.
- The builder writes only the hypotheses section of your project draft. Everything else stays exactly as you entered it.
What this tool can't do
- It can't tell whether a hypothesis is supported by theory or previous research. Only your literature review can show that.
- It can't choose a direction for you. A directional hypothesis needs a reason from theory or evidence.
- It can't check that your variables can be measured, or that your sample is large enough to test the hypotheses.
- It reads wording by recognising common patterns, so its checks can miss or misread things.
- It can't determine the quality, originality or publishability of your research, or give supervisor or ethics approval.
Awaiting academic review
- Academic references for each hypothesis type: none have been added yet.
- The definitions, strengths, limitations and examples of each hypothesis type.
- The guidance on when hypotheses suit quantitative, qualitative, mixed methods and multi-method studies.
- The standard wording used for difference, relationship, prediction, moderation and mediation hypotheses.
Privacy
Everything you enter stays in your browser. Nothing is sent anywhere or saved, and it is cleared when you leave the page. If you start a project in your research workspace, your work is also saved in this browser, and only this browser, until you delete the project.