Hypothesis tests (do this for each hypothesis) / and or revealing results about your expectation

- Briefly repeat hypothesis (“In H1, we hypothesized that […]”)
- Report the result in brackets - both the coefficient, and the significance level
- e.g., “Indeed, the effect of A on B is statistically significant (beta = xx,
*p*= .012)” - e.g., “A increases B by x% (beta = xx,
*p*= .025)“

- e.g., “Indeed, the effect of A on B is statistically significant (beta = xx,
- Explain the result (the “why?”)
- For a confirmed hypothesis, increase the intuition of your hypothesis (“apparently, as hypothesized, A leads to B
**because**[repeat, etc.]…”) - For an unconfirmed hypothesis, explain why you don’t find the effect (e.g., a conceptual reason (effect doesn’t exist, and yo u h ave an
**argument why it may not exist), or a measurement issue (e.g., data problems), etc.)**

- For a confirmed hypothesis, increase the intuition of your hypothesis (“apparently, as hypothesized, A leads to B

Remaining variables

- Explain the effects of control variables here (“The control variables age and gender have face valid effects. For example, all else equal, age increases the intention to purchase (beta = xx,
*p*= .12). Education turns out to be not a significant predictor of intention to purchase (beta = xx,*p*= .63). This may be the case because

- Explain the effects of control variables here (“The control variables age and gender have face valid effects. For example, all else equal, age increases the intention to purchase (beta = xx,

**Note:** While hypothesis tests are usually carried out against a *p* value of .05 (or sometimes .10), you need to report the exact *p* value in the text and tables with, e.g., three digits after the decimal point (e.g., *p* = .049, instead of *p* < .05). If you cannot give an exact *p* value because of rounding (e.g., the true *p* value is .00001, but it would round to *p* = .000), you need to write it down as *p* < .001. The *p* needs to be printed in italic always.