EDUCATION

The education market in the 21st century is one driven by a number of initiatives to raise the standards – as well as the benchmarks – for primary education. Directives like “no child left behind” have meant that standardized testing – previously something managed at a local level – is now broad, standardized, and increasingly important.

Educators know that standardized tests are not simply multiple-choice, “fill-in-the-blank” response sheets. Essay Scoring is an integral part of every testing routine from 6th Grade Performance tests to the SAT.

The problem is, in a market where cash is always tight, and where resources are always limited, how do you cope with a testing format that requires significant human intervention – i.e., scoring?

This is where the education market turns to Content Analyst. Since our product is designed to analyze text in context, our education partners have developed powerful products that actually use our patented software to automatically score essays based on content, conceptual grasp of the subject, ability to effectively communicate an idea, and overall writing skills.

Essay Scoring – How it works

Essay Scoring blends a number of Content Analyst features into a unique set of application tools. First, the Content Analyst engine is prepared to be “trained” on what is identified by educators as good essays – and bad ones. Then, Content Analyst is trained by allowing it to “read” a range of essays that correspond to a particular area of value. If this is, for instance, grammatical accuracy, Content Analyst will be trained with a set of essays – all on a particular question – that range from great to grotesque.

Once trained, Content Analyst easily reads new essays, compares their salient points to what it has already learned about the subject, how the question should be answered, and an assessment of overall writing skills – and provides the appropriate score, virtually instantaneously.

Essay Scoring – How it delivers Value

Adding the element of machine scoring to testing – everything from “low-stakes” textbook essays to “high-stakes” academic testing such as the GMAT – has an immediate cost benefit. In the traditional education world, two different individuals score every essay: if the scores match within a given range, then the score is considered “valid.” If the scores do not, then a third individual scores that essay and the score is generally based on the two closest scores.

By employing electronic scoring, one of those individuals can be eliminated. The machine and one individual score an essay, and if the scores match within that range, then the score is considered “valid.” If not, then a second individual is involved as above.

The immediate effect is a 50% reduction in the overall cost of 1st pass essay – and our experience has shown no statistical difference between scoring based on two individual scorers versus one scorer plus the machine.