How to extract one value from a variable, and link it with a whole other variable

[aczech2001]

Dear All,

I am a complete newbie to SAS, so apologies for lacks of jargon, and probably posting a very simple and not very stimulating question. My deep appreciation for your help.

I have the data-set with the following 2 categories: 

GENDER: Male (1), Female (2)

CAUSE FOR LIMB AMPUTATION: Trauma (1), Vascular with Diabetes (2), Vascular without Diabetes (3), Oncologic (4)

My task is to see if there is a higher percentage of female patients whose reason for amputation was 'trauma'.

My question is how to extract only count for patients who answered 'trauma' and produce a crosstab, only featuring 'Trauma' x 'Gender'?

How can one achieve it using 'Tasks and Utilities' instead of the code?

I attach data in case someone wants to fiddle with it.

Thank you

Agnieszka

[Reeza · Accepted solution]

1. Use an Utilities>Import task to import your data

2. Use a Data>Filter task to get only the Trauma records

3. Use a Statistics>Table Analysis to perform a Chi Square test

Documentation for each task specifically can be found here:

https://documentation.sas.com/?activeCdc=webeditorcdc&cdcId=sasstudiocdc&cdcVersion=3.71&docsetId=we...

---

@aczech2001 wrote:

Dear All,

I am a complete newbie to SAS, so apologies for lacks of jargon, and probably posting a very simple and not very stimulating question. My deep appreciation for your help.

I have the data-set with the following 2 categories: 

GENDER: Male (1), Female (2)

CAUSE FOR LIMB AMPUTATION: Trauma (1), Vascular with Diabetes (2), Vascular without Diabetes (3), Oncologic (4)

My task is to see if there is a higher percentage of female patients whose reason for amputation was 'trauma'.

My question is how to extract only count for patients who answered 'trauma' and produce a crosstab, only featuring 'Trauma' x 'Gender'?

How can one achieve it using 'Tasks and Utilities' instead of the code?

I attach data in case someone wants to fiddle with it.

Thank you

Agnieszka

---

[aczech2001]

Thank you so much! This seemed to work!