What is an alternative hypothesis and a null hypothesis?

The null hypothesis refers to the supposition that the variation you see within a set of data is that way due to random chance.





The alternative hypothesis supposes that there's a reason that some stuff turned up differently from others.





For example. One city has two sports teams. Each team has been around for about the same amount of time, but one team has 10 championships while the other has 2. (You can use pretty much any numbers you want for the example. It could be 6 and 5 if you wanted. Whatever). The null hypothesis states that each team has the same chance of winning the championship in any given year and that one team just got lucky a few more times than the other. The alternative hypothesis would give a set of reasons for why the team with more championships has a better chance of winning than the team with fewer championships.





One more example. Blue smurfs and green aliens take an intelligence test. The mean and median scores for one group is a bit higher for one group than the other. The null hypothesis states that they really have about the same level of intelligence and that a few lucky guesses one way or the other caused the slight variance in score; it's not something that should make you assume that one group is actually smarter than the other. The alternative hypothesis would basically say one group is smarter than the other.





In order to "reject the null hypothesis" (or conclude that something other than random chance caused a variance in the data), there's various ways of measuring how big that difference is and whether or not you should draw conclusions from that. If you "fail to reject the null hypothesis," that means the differences within the set of data were not statistically significant (stats word), and while it doesn't guarantee that there isn't a reason for the differences in the scores, it does mean you can't rule out the possibility of coincidence or random chance.|||The null is that there will be no effect or there is no relationship between your variables (there is no relationship between temperature and sun burns).





The alternative hypothesis states that there is a relationship between your variables, or there will be an effect (temperature will have an effect on the number of people reporting sunburn). The alternative hypothesis can be one or two tailed, a two tailed also predicts a direction (the hotter it is, the more people will report sunburn).|||if I remember correctly . . .


null hypothesis is that treatment (whatever you are doing in experimenrt) will have no effect or that there will be no difference between exp group and control group|||null means none, zip, zero