One-tailed and two-tailed hypothesis tests

Watch this thread
jduxie4414
Badges: 17
Rep:
? You'll earn badges for being active around the site. Rep gems come when your posts are rated by other community members.
#1
Report Thread starter 2 years ago
#1
I have a random question that I haven't really had time to think about, so I thought I'd ask it here!

There isn't a specific question in mind, so I won't be able to use numbers but I can probably make one up!

So for a one tailed test, you are finding the probability that the population parameter is larger/ smaller (you pick one dependent on question) than first thought, and if this probability is less than alpha, the significance level, you reject H0 and accept H1. This means you have a 5% chance of saying the value is larger. (Or smaller if that's what you choose!)

With a two tailed test, you are testing the probability that the population parameter is different to what you originally thought (so it can be higher or lower, you don't care which). This means you split alpha into two at each end of the distribution to calculate the critical values, therefore there's an alpha/2 chance of the value being rejected.

So say that alpha is 5% , and you get a probability population parameter being more extreme of 3.2% (for example). If a one tailed test, you'd reject H0 (as 3.2%<5%) and assume that the true value is greater. However, if this was a two tailed test, 3.2%>2.5% (alpha /2) you'd accept H0 and assume that the value hasn't changed, even though with a one tailed you've just said it's larger!!

I was wondering if anyone has any explanation for this? I think it came up in one of my FM lessons and no one could answer it!
0
reply
sachinihimara
Badges: 16
Rep:
? You'll earn badges for being active around the site. Rep gems come when your posts are rated by other community members.
#2
Report 2 years ago
#2
It's because I'm the one tailed it's within 5% significance that the result not supporting H0 whereas in the two tailed test it isn't within 5% significance that it is not supporting H0 because its accepting 2.5% on the other end as well
I'm not sure if it makes sense but feel free to pm me if you need me to explain it better
Last edited by sachinihimara; 2 years ago
0
reply
jduxie4414
Badges: 17
Rep:
? You'll earn badges for being active around the site. Rep gems come when your posts are rated by other community members.
#3
Report Thread starter 1 year ago
#3
(Original post by sachinihimara)
It's because I'm the one tailed it's within 5% significance that the result not supporting H0 whereas in the two tailed test it isn't within 5% significance that it is not supporting H0 because its accepting 2.5% on the other end as well
I'm not sure if it makes sense but feel free to pm me if you need me to explain it better
Yeah, I understand that it just doesn’t make sense that you’re saying yeah the value is larger, but no the value isn’t different!
0
reply
Gregorius
Badges: 14
Rep:
? You'll earn badges for being active around the site. Rep gems come when your posts are rated by other community members.
#4
Report 1 year ago
#4
(Original post by jduxie4414)
I have a random question that I haven't really had time to think about, so I thought I'd ask it here!

There isn't a specific question in mind, so I won't be able to use numbers but I can probably make one up!
OK, so you're on the right lines with your thinking, but there's an important misapprehension lurking in here, so let's correct that!

You're concerned with a statistical model where you're assuming that the observations you're collecting are drawn from a probability distribution with unknown, but fixed, parameters, and you're interested in drawing some conclusion about those population parameters from those observations. So, important point: the underlying population parameters are fixed: they are not random variables, so you can't make probability statements about them. What is random are the observations that you have drawn, and any statistics that you calculate from them.

Let's think about a single population parameter, and let's call it \mu, and let's assume it represents the population mean. From your sample of observations, if you want to make inferences about the population mean, a good guess is that we should look at the sample mean as a statistic, and see how it compares with our idea of what the underlying population mean might be. For the range of probability distributions you're used to, it's possible to work out theoretically what the sampling distribution of the sample mean is, given a particular value of the population mean. That is, the distribution of sample mean values you'd get if you repeatedly draw samples from the underlying population, and calculated their means.

So suppose we hypothesize that the population parameter  \mu = \mu_{0}, and we want to use the sample mean to decide whether this is sensible. We know the sampling distribution of the sample mean, given that  \mu = \mu_{0}, so we can work out probability statements about the sample mean, on the assumption that the population mean is  \mu_{0}.

The basic strategy is that you work out the probability that the value of the sample mean is "as extreme, or more extreme" as its observed value relative to  \mu_{0}. In other words, if the observed value of the sample mean is "far away" from  \mu_{0}, you infer that this is unlikely, and therefore reject the hypothesis that  \mu = \mu_{0}.

Now that's the "simple" case, where you assume that the underlying population parameter takes a particular value. More complicated is where you assume that the underlying population parameter takes a range of possible values:  \mu \le \mu_{0}, for example. But the idea is the same: if you assume this, what's the probability of seeing the observed value of the sample mean?

So for a one tailed test, you are finding the probability that the population parameter is larger/ smaller (you pick one dependent on question) than first thought, and if this probability is less than alpha, the significance level, you reject H0 and accept H1. This means you have a 5% chance of saying the value is larger. (Or smaller if that's what you choose!)
Nearly: you are finding the probability that the sample mean takes the value it does (or more extreme one way), under the assumption that the population mean is  \le \mu_{0}. You reject that assumption if the probability is too small. In this case, "too small" is often taken to be 5%.

With a two tailed test, you are testing the probability that the population parameter is different to what you originally thought (so it can be higher or lower, you don't care which). This means you split alpha into two at each end of the distribution to calculate the critical values, therefore there's an alpha/2 chance of the value being rejected.
Again, you are finding the probability that the sample mean takes the value it does (or more extreme two ways), under the assumption that the population mean is  \mu_{0}. In this case, you're interested in whether the sample mean is much more than  \mu_{0}, or much less, so you need to take account of two tails of the distribution.

So say that alpha is 5% , and you get a probability population parameter being more extreme of 3.2% (for example). If a one tailed test, you'd reject H0 (as 3.2%<5%) and assume that the true value is greater. However, if this was a two tailed test, 3.2%>2.5% (alpha /2) you'd accept H0 and assume that the value hasn't changed, even though with a one tailed you've just said it's larger!!
Yes, this is what you'd do (only you never "accept H0", you "do not reject H0" - remember that it's quite possible that you don't have enough evidence to reject H0 even when it is wrong - absence of evidence is not evidence of absence).

I was wondering if anyone has any explanation for this? I think it came up in one of my FM lessons and no one could answer it!
There are a couple of points here:

In a two-tailed test, you are implicitly doing two statistical tests; you are testing whether the test statistic is "too large", and you are testing if it is "too small". If you remember what the \alpha level of a test is, it's the probability of rejecting H0 when it is in fact true. If you do two tests, then you've got two chances to reject H0, so you have to add up the probabilities of rejecting a true H0 for both tests. For a one-sided test, you only have one chance, so you can "spend" all of your \alpha in one go.

The other point is that the assumptions underlying a two-sided test are different from those underlying a one-sided test. For a one-sided test you are allowing your underlying population parameter a much wider range of possibilities, compared to a two sided test. This affects the whole probability structure of your problem. So when you say that with that value of 3.2% you'd reject H0 in one case, and you would not reject it in the other, the point is that they are different H0's, with different assumptions. So you're not accepting a particular H0 in one case and rejecting it in the other, you're accepting or rejecting different statements
1
reply
X

Quick Reply

Attached files
Write a reply...
Reply
new posts
Back
to top
Latest
My Feed

See more of what you like on
The Student Room

You can personalise what you see on TSR. Tell us a little about yourself to get started.

Personalise

Has advance information helped during your exams?

Yes (62)
65.96%
No (25)
26.6%
I didn't use it to prepare (7)
7.45%

Watched Threads

View All