# Same as posting 12892. Please delete from Library.

A study of tobacco quitting was done. The authors concluded that a significant preictor of maintaining cessation was forbidding smoking in the house, and gave the following supportive statistics: OR 1.95; C.I. 1.05-3.6; p<0.01

What is the exposure?

What is the outcome?

What does OR stand for?

What does CI stand for?

State what an OR of 1.95 means in the above context?

Is an OR of 1.95 significant? Why or Why not?

What does p<0.01 mean?

is an OR with a p<0.01 more or less believable than an OR with a p<0.05?

https://brainmass.com/statistics/hypothesis-testing/same-as-posting-12892-please-delete-from-library-12865

## SOLUTION This solution is **FREE** courtesy of BrainMass!

What does OR stand for?

<br>-------------------------------------

<br>OR is the "odds ratio" = (odds of maintaining cessation if smoking forbidden)/(odds of maintaining cessation if smoking allowed).

<br>

<br>where "odds" of maintaining cessation if smoking fobidden =

<br>(proportion of people maintaining when smoking forbidden)/(proportion of people not maintaining when smoking forbidden).

<br>

<br>and "odds" of maintaining cessation if smoking allowed =

<br>(proportion of people maintaining when smoking allowed)/(proportion of people not maintaining when smoking allowed).

<br>

<br>

<br>What is the exposure?

<br>-----------------------------------

<br>I am assuming that when you say "exposure", you intended to say "expectation"...

<br>In this problem the Null hypothesis is : "Maintaining cessation is independent of whether or not smoking is forbidden"

<br> Under the null hypothesis (assumption of independence), it is reasonalble to "expect" the odds ratio to be 1. In other words, if forbidding smoking has no effect on maintaining cessation, the "expectation" of the odds ratio is 1.

<br>

<br>

<br>What is the outcome?

<br>--------------------------------

<br>Outcome: In this sample, the odds ratio is 1.95. The numerator of the odds ratio is almost twice as large as the denominator. We tend to believe, therefore, that the odds of cessation when smoking is forbidden is greater than the odds of cessation when smoking is not forbidden.

<br>

<br>

<br>What does CI stand for?

<br>-----------------------------------------

<br>CI stands for "Confidence Interval.

<br>

<br>State what an OR of 1.95 means in the above context?

<br>Is an OR of 1.95 significant? Why or Why not?

<br>-----------------------------------------------------------------------

<br>Whether or not the sample odds ratio of 1.95 is something we might reasonably expect to see due to sampling variation is addressed by finding a CI or "Confidence interval" with p or "p-value". This experiment can be interpreted as "We can be 99% confident that the true value of the odds ratio for the population lies between 1.05 and 3.6."

<br>

<br>What does p<0.01 mean?

<br>-----------------------------------------------------------------------------

<br> This result is statistically significant because we could expect a sample of this size to produce an odds ratio this far from 1 less than 1% of the time (p <0.01) if the variables were independent.

<br>

<br>

<br>Is an OR with a p<0.01 more or less believable than an OR with a p<0.05?

<br>------------------------------------------------------

<br>p < .01 is more believable than p < .05 (Substitute 5% in the question above) . The p-value is the probability of a Type I error, the probability of falsely finding a relationship between the two variables if in fact they are independent

https://brainmass.com/statistics/hypothesis-testing/same-as-posting-12892-please-delete-from-library-12865