Date: Wed, 22 Jun 2005 13:03:59 -0600
Reply-To: Alan Churchill <SASL001@SAVIAN.NET>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Alan Churchill <SASL001@SAVIAN.NET>
Subject: Re: Interview questions re data management
In-Reply-To: <s2b978dd.026@MAIL.NDRI.ORG>
Content-Type: text/plain; charset="us-ascii"
Peter,
Let me provide some of my suggestions on code management and hopefully that
will provide some questions for you to ask. I am responding privately on
this one since I am not sure of applicability to your question. If it seems
valid, you can reply to the group.
I would focus on metadata management skills. How do they organize
information about the programs. An example would be a SAS dataset that
describes the 96 cities and the variables, and then have code generators
create the SAS programs. That way cities and variables can change but you
merely have to update a central repository and the code adjusts itself.
Another organizational place I have used is Excel or XML. Regardless it
makes life easier.
I would also ask how they approach code management and code naming. If you
pour all of your code into a directory called "code", probably not good.
There needs to be an organizational structure. For example:
clinical_information\metadata
clinical_information\formats
clinical_information\informats
clinical_information\macros
etc.
Then it is good to have the code automatically created from 1 program that
controls all of the others. I have also named programs according to order:
Step1_ProcessAllWeblogs.sas
Program header comments are critical and should be formatted properly:
/*--------------------------------------------------------------------+
| Copyright (c) 2005 by Savian, LLC |
| 999 whereever street, anywhere, colorado |
| ---all rights reserved |
+-------------+-------------------------------------------------------+
| product: | NewProducte |
| subsystem: | DataReader |
| program: | samplecode.sas |
| support: | sasl001@savian.net |
| version: | sas 9.13 |
| language: | base |
+-------------+-------------------------------------------------------+
| purpose: | this is the main program that drives data reading |
| | data. |
+-------------+-------------------------------------------------------+
| usage: | fill in the appropriate macro parameters listed below |
| | |
+-------------+-------------------------------------------------------+
| parameters: | see below |
+-------------+-------------------------------------------------------+
| comments: | some nifty comments go here. |
| | |
+-------------+-----------+-----------+-------------------------------+
| history |alchur | 01jun2005 | Initial coding. |
| | | | |
+-------------+-----------+-----------+------------------------------*/
A person needs to also document changes made to the code so that code
management can happen.
I would also ask them their experience at parsing information and reading
data from various data sources. ODBC and OleDb questions are good as is
stuff such as XML (imo).
Thanks,
Alan
Savian
"Bridging SAS and Microsoft Technologies"
http://www.savian.net
-----Original Message-----
From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of Peter
Flom
Sent: Wednesday, June 22, 2005 11:42 AM
To: SAS-L@LISTSERV.UGA.EDU
Subject: Interview questions re data management
I have to test an applicant for a job, and not sure what questions would
be best for the data management part of the job
The project involves analysis of data from 96 large cities in the USA.
There will be about 400 variables for each city, and the will be
multiple years of data (about 8 years).
Key things:
Cleaning and keeping track of data as it comes in
Development of varaible names that make sense.
Developing of formats, attributes and so on that make sense.
Thorough documentation of same
This is all stuff that's new to me, so I am not sure what to ask. Any
suggestions welcome
Peter
Peter L. Flom, PhD
Assistant Director, Statistics and Data Analysis Core
Center for Drug Use and HIV Research
National Development and Research Institutes
71 W. 23rd St
www.peterflom.com
New York, NY 10010
(212) 845-4485 (voice)
(917) 438-0894 (fax)