On-line or on paper, surveys can help teams make decisions.  The best part about surveys is that you can define opinions and responses based on demographic profiles like age, education, management level, etc. 

In interaction design, surveys play a small role, if any.  However, my career started in market research (before the web, before GUI) and so I understand how surveys can solve specific kinds of problems.  I include this methodology here as a measure of that good faith, but it is not a critical skill set for my area of focus.

The key challenge of survey design is to ask questions of the respondents without biasing them by the sequence of questions or the phrasing of any single question, and to give the truest range of possible answers for them to choose from.  Once data is collected, interpretation is inevitably bound by the priorities and perspective of its analyst. 

Sidenote: Presenting quantitative data in visual form is a valuable skill set -- and part of an interaction designer's toolkit.  To the extent that a user can "see" the meaning of data, without having to study a table of numbers and text, visual representations can accelerate knowledge production and sharing.  In the health care world, it is essential to include visual representations of status conditions (critical care, especially), alerts, updates, irregularities in lab results, or other key data affecting diagnosis and response. 

I regularly point clients to Smart Money's Map of the Market as a stellar example of how visual representations of data can give our right-brain access to meaning-making.  The next level challenge for interaction designers lies in how to help users to pose questions to a visual representation so that it can be a source of insight. 

I've gone to SIG-Graph to study how visual representations of data can be applied to productivity applications, and there are many interesting studies and innovations out there ready for consideration.  I look forward to the opportunity to expand the marketplace in this direction.