# one tailed test or two test?

The question is "Some guy on the internet designed a weight-loss program. You are wondering if it is working. You are given a sample of some people who did the program"

according to the solution given, this is a one tailed test. but my question is this: why isn't it a two tailed test

we can have a null hypothesis as

```
H0 : The weight loss program IS working
h1: the weight loss program IS NOT working
```

wont that be correct? and if so, then this makes it a two tailed test

Hi Moffat,

Considering that we're interested only in the ** weight loss** outcome and not in the

*outcome, I'd say this is a one-tailed test. The weight loss program*

**weight gain****decreases**people's weight or it doesn't.

Best,

Ned

Mr Ned, I hope that you answer my question.

shouldn't the null hypothesis be less than zero ( <0) as we are seeking the difference between before and after which indicates a negative figure?