Hypothesis Test Writing

[mengland00]

I am trying to remember the correct way to write hypotheses, as I am a bit out of practice.

I am conducting bivariate analysis on variables associated with bullying and delinquency, specifically looking at interactions between one variable, whether or not the student has been bullied (responses to the survey question "How many times have you been bullied in school?"), and a series of other variables related to student delinquency. An example of one of the delinquency variables would be responses to the question "During the past 12 months, how many times were you in a physical fight?"

My question is how would I write the hypothesis for this if my thinking is that students who responded that they are bullied are likely to report having been in fights?

Is it as simple as:

H0: There is no relationship between being a bullying victim and having engaged in a physical fight.

Incidentally, the Pearson's correlation was found to be .846 (2-tailed) using SPSS. In this case I would reject the null, right (n = 941)?

Any help would be much appreciated.

[JohnM]

Your null hypothesis seems OK to me, and yes, you probably would reject it with a correlation that strong.

You can run a test to see if that value of r is significantly different from 0. Check our online resources post in the Examples section for more details.

[mengland00]

Many thanks for your speedy response.