The Three Most Common Types of Hypotheses

The Three Most Common Types of Hypotheses

Simple main effects (i.e., X leads to Y) are usually not going to get you published. Main effects can be exciting in the early stages of research to show the existence of a new effect, but as a field matures the types of questions that scientists are trying to answer tend to become more nuanced and specific.  In this post, I’ll briefly describe the three most common kinds of hypotheses that expand upon simple main effects – at least, the most common ones I’ve seen in my research career in psychology – as well as providing some resources to help you learn about how to test these hypotheses using statistics.

Incremental Validity

“Can X predict Y over and above other important predictors?”


This is probably the simplest of the three hypotheses I propose. Basically, you attempt to rule out potential confounding variables by controlling for them in your analysis.  We do this because (in many cases) our predictor variables are correlated with each other. This is undesirable from a statistical perspective, but is common with real data. The idea is that we want to see if X can predict unique variance in Y over and above the other variables you include.

In terms of analysis, you are probably going to use some variation of multiple regression or partial correlations.  For example, in my own work I’ve shown in the past that friendship intimacy as coded from autobiographical narratives can predict concern for the next generation over and above numerous other variables, such as optimism, depression, and relationship status (Mackinnon et al., 2011).


“Under what conditions does X lead to Y?”

Of the three techniques I describe, moderation is probably the most tricky to understand.  Essentially, it proposes that the size of a relationship between two variables changes depending upon the value of a third variable, known as a “moderator.”  For example, in the diagram below you might find a simple main effect that is moderated by sex. That is, the relationship is stronger for men than for women:


With moderation, it is important to note that the moderating variable can be a category (e.g., sex) or it can be a continuous variable (e.g., scores on a personality questionnaire).  When a moderator is continuous, usually you’re making statements like: “As the value of the moderator increases, the relationship between X and Y also increases.”


“Does X predict M, which in turn predicts Y?”

We might know that X leads to Y, but a mediation hypothesis proposes a mediating, or intervening variable. That is, X leads to M, which in turn leads to Y.  In the diagram below I use a different way of visually representing things consistent with how people typically report things when using path analysis.


I use mediation a lot in my own research. For example, I’ve published data suggesting the relationship between perfectionism and depression is mediated by relationship conflict (Mackinnon et al., 2012). That is, perfectionism leads to increased conflict, which in turn leads to heightened depression. Another way of saying this is that perfectionism has an indirect effect on depression through conflict.

Helpful links to get you started testing these hypotheses

Depending on the nature of your data, there are multiple ways to address each of these hypotheses using statistics. They can also be combined together (e.g., mediated moderation). Nonetheless, a core understanding of these three hypotheses and how to analyze them using statistics is essential for any researcher in the social or health sciences.  Below are a few links that might help you get started:

Are you a little rusty with multiple regression? The basics of this technique are required for most common tests of these hypotheses. You might check out this guide as a helpful resource:

David Kenny’s Mediation Website provides an excellent overview of mediation and moderation for the beginner.

Preacher and Haye’s INDIRECT Macro is a great, easy way to implement mediation in SPSS software, and their MODPROBE macro is a useful tool for testing moderation.

If you want to graph the results of your moderation analyses, the excel calculators provided on Jeremy Dawson’s webpage are fantastic, easy-to-use tools:


29 Responses to “The Three Most Common Types of Hypotheses”

  1. dipak sarkar says:

    I want to see clearly the three types of hypothesis

  2. paul says:

    Thanks for your information. I really like this

  3. Karen Deller says:

    Thank you so much, writing up my masters project now and wasn’t sure whether one of my variables was mediating or moderating….Much clearer now.

  4. Adem, Abdulaziz says:

    Thank you for simplified presentation. It is clearer to me now than ever before.

  5. Ahmed Ibrahim says:

    Thank you. Concise and clear

  6. Osama says:

    hello there

    I would like to ask about mediation relationship:
    If I have three variables( X-M-Y)how many hypotheses should I write down? Should I have 2 or 3?
    In other words, should I have hypotheses for the mediating relationship?
    What about questions and objectives? Should be 3?
    Thank you.

    • sean says:

      Hi Osama. It’s really a stylistic thing. You could write it out as 3 separate hypotheses (X -> Y; X -> M; M -> Y) or you could just write out one mediation hypotheses “X will have an indirect effect on Y through M.” Usually, I’d write just the 1 because it conserves space, but either would be appropriate.

      • Najm says:

        Hi Sean, according to the three steps model (Dudley, Benuzillo and Carrico, 2004; Pardo and Román, 2013)., we can test hypothesis of mediator variable in three steps: (X -> Y; X -> M; X and M -> Y). Then, we must use the Sobel test to make sure that the effect is significant after using the mediator variable.

        • sean says:

          Yes, but this is older advice. Best practice now is to calculate an indirect effect and use bootstrapping, rather than the causal steps approach and the more out-dated Sobel test. I’d recommend reading Hayes (2018) book for more info:

          Hayes, A. F. (2018). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach (2nd ed). Guilford Publications.

  7. Marisol says:

    Hi! It’s been really helpful but I still don’t know how to formulate the hypothesis with my mediating variable.

    I have one dependent variable DV which is formed by DV1 and DV2, then I have MV (mediating variable), and then 2 independent variables IV1, and IV2.

    How many hypothesis should I write? I hope you can help me 🙂

    Thank you so much!!

  8. Helly says:


    Could you help me formulate my research question? English is not my mother language and I have trouble choosing the right words.
    My x = psychopathy
    y = aggression
    m = deficis in emotion recognition

    thank you in advance

  9. Kabiru says:

    I have mediator and moderator how should I make my hypothesis

  10. Ashley says:

    Can you have a negative partial effect? IV – M – DV. That is my M will have negative effect on the DV – e.g Social media usage (M) will partial negative mediate the relationship between father status (IV) and social connectedness (DV)?

    Thanks in advance

    • sean says:

      Hi Ashley. Yes, this is possible, but often it means you have a condition known as “inconsistent mediation” which isn’t usually desirable. See this entry on David Kenny’s page:

      Or look up “inconsistent mediation” in this reference:

      MacKinnon, D. P., Fairchild, A. J., & Fritz, M. S. (2007). Mediation analysis. Annual Review of Psychology, 58, 593-614.

  11. Mr. Garba Dahiru says:

    This is very interesting presentation. i love it.

  12. Mr. Garba Dahiru says:

    This is very interesting and educative. I love it.

  13. George says:

    Hello, you mentioned that for the moderator, it changes the relationship between iv and dv depending on its strength.
    How would one describe a situation where if the iv is high iv and dv relationship is opposite from when iv is low. And then a 3rd variable maybe the moderator increases dv when iv is low and decreases dv when iv is high.

    • sean says:

      This isn’t problematic for moderation. Moderation just proposes that the magnitude of the relationship changes as levels of the moderator changes. If the sign flips, probably the original relationship was small. Sometimes people call this a “cross-over” effect, but really, it’s nothing special and can happen in any moderation analysis.

  14. Mehak says:

    i want to use an independent variable as moderator after this i will have 3 independent variable and 1 dependent variable…. my confusion is do i need to have some past evidence of the X variable moderate the relationship of Y independent variable and Z dependent variable.

  15. Andrey Andoko says:

    Dear Sean
    It is really helpful as my research model will use mediation. Because I still face difficulty in developing hyphothesis, can you give examples ?
    Thank you

  16. ella bella says:

    Hi! is it possible to have all three pathways negative? My regression analysis showed significant negative relationships between x to y, x to m and m to y.

  17. Pavithra Ratenom says:

    I have 1 independent variable, 1 dependent variable and 4 mediating variable
    May I know how many hypothesis should I develop?

  18. Tina says:

    I have 4 IV , 1 mediating Variable and 1 DV

    My model says that 4 IVs when mediated by 1MV leads to 1 Dv

    Pls tell me how to set the hypothesis for mediation

  19. Tina says:

    I have 4 IVs ,2 Mediating Variables , 1DV and 3 Outcomes (criterion variables).

    Pls can u tell me how many hypotheses to set.

    Thankyou in advance

  20. how to says:

    I am in fact happy to read this webpage posts which carries tons of useful information, thanks for providing such data.

  21. I see you don’t monetize,
    don’t waste your traffic, you can earn additional bucks every month with new monetization method.
    This is the best adsense alternative for any type of website (they approve all websites), for
    more info simply search in gooogle: murgrabia’s tools

  22. belay mengstie says:

    what if the hypothesis and moderator significant in regrestion and insgificant in moderation?

Leave a Reply

Your email address will not be published. Required fields are marked *