# Null Hypothesis

• Sep 29th 2009, 11:20 AM
Null Hypothesis
Hey folks I've got a Q about what I should make my null hypothesis...

The actual question i'm fine with and know what the outcome shows but I dont know if I have the null hypothesis right. (i.e. should it be the other way round?)

So Q is...

For the sample in data set 1 use the sign test to test whether the median transit time differs from 3.50 seconds.

Data set 10 is...
Appearance transit times (seconds) of a bloodmarker for 11 patients with significantly occluded right coronary arteries.

1.80 | 3.30 | 5.65 | 2.25 | 2.50 | 3.50 | 2.75 | 3.25 | 3.10 | 2.70 | 3.00

The median transit time ( $M_0$) was originally thought to be 3.50 seconds.

------

My null hypothesis was,
$H_0 : M = M_0$, where $M$ denotes the actual median

And my answer came out to be, using a two tailed binomial test, that there was evidence at the 2.5% level to suggest median was not 3.5 and so we reject the null hypothesis.

Is this the right way to construct a null hyp? Or should it be $H_0 : M \neq M_0$?
• Sep 29th 2009, 02:24 PM
noname
The null hypothesis is the one valid unless the data contradicts it.
The median transit time was originally thought to be 3.50 seconds, so your null hypothesis is correct.

But in some case you may consider also the statistical significance