1348 Hall Avenue
Sharon, PA 16146
Classifying, organizing, and summarizing what people are saying can be a challenging task. We apply state-of-the-art methods to produce reliable, accurate, and usable information from free-form text. Our methods include:
- Application of logic and theory to derive sets of mutually exclusive and collectively exhaustive usable categories.
- Iterative testing to develop classification rules that produce reliable results.
- Real-time monitoring of classification outputs to identify and correct anomolies in classification rules.
- Multisource convergence to resolve classification ambiguities.
- Classification output validation against established criteria.
You can preview our text coding webinar course on the American Association for Public Opinion Research website here.
Options for automated text classification are becoming more widely available. Our investigation into these automated technologies found they are fairly good at reliably classifying short text into a few simple categories. If you want to know which of two categories any short, free-form answer best fits, then automated machine classification may be the solution for you.
If you are interested in more detail about what your customers, clients, or constituents are saying, then rigorous human coding using our state-of-the-art methods is probably what you need.
We've used our state-of-the-art methods to classify and summarize some of the more complex sets of free-form text. Our text coding methods have generated new insights into:
- What people are worried about.
- What people like and don't like.
- How people explain things.
You can see the coding work we did for the American National Election Studies here.
Preview the online webinar course we gave for the American Association of Public Opinion Research.
See the coding reports and data we prepared for the American National Election Studies.