# Help needed. Which test to use?

#### wenwong

##### New Member
Hey guys, I am a junior doctor doing some research and I should have paid more attention to statistics during med school.

I have a data with 5 "treatments" and looking at 3 different outcomes.
for example: Treatment 1 (n=100) - Outcome 1 "30 successful", outcome 2 " 40 successful", outcome 3 "1- successful"

Questions:
1. What test should I use to calculate the P value for the treatment
2. What test can I use to compare Treatment 1 - 5 against their outcomes.

#### Karabiner

##### TS Contributor
Which research questions do you want to answer, exactely?

With kind regards

Karabiner

#### wenwong

##### New Member
2 main questions,
1. How effective is treatment 1-5 in achieving outcome 1-3. For example i know treatment 1 has 30% success rate. but I dont know how to calculate the p value
2. Is treatment 1-5 better in achieving outcome 1-3.

Thanks again.

Wen

#### Karabiner

##### TS Contributor
1. How effective is treatment 1-5 in achieving outcome 1-3. For example i know treatment 1 has 30% success rate. but I dont know how to calculate the p value
Which p value? You can calculate 5*3=15 percentages in order to see effectiveness.
Maybe you can add 95% confidence intrevals, in order to demontrate that the sample
percentages are not necessarily exact estimates of the population percentages.

2. Is treatment 1-5 better in achieving outcome 1-3.
You could start by constructing three 5x2 crosstables, each with with 5 groups and outcome achieved/not achieved.
The global test would be Chi².

With kind regards

Karabiner