Using software to analyze social media comments and sentiments, researchers have found that 51 percent of tweets by private users of Twitter accounts contained stigma, when making reference to about Alzheimer's disease and the people who deal with this condition.

The new software developed at Oregon State University (OSU) may be applicable to a range of other social science research questions, the researchers said, and already shows that many people may not adequately appreciate the power of social media to greatly transcend the type of interpersonal, face-to-face communication humans are most accustomed to.

"As a society it's like we're learning a new skill of text communication, and we don't fully understand or reflect on its power to affect so many people in ways that we may not have intended," said Nels Oscar, an OSU graduate student in the College of Engineering. "Social media is instant, in some cases can reach millions of people at once, and can even instigate behaviors. We often don't even know who might eventually read it and how it will affect them."

When it comes to Alzheimer's disease, thoughtless or demeaning comments on a broad level via social media can take an already-serious problem and make it worse.

The topic studied by Oscar and his colleagues, who published a paper in the Journals of Gerontology: Psychological Sciences, is of growing importance.

A global tripling of individuals with some form of dementia has been projected in coming decades, from 43 million now to 131 million by 2050.

In the research, the software was designed to recognize and interpret the use of various keywords associated with Alzheimer's disease, such as dementia, memory loss or senile.

The system was improved by comparing results to the same comment evaluated by human researchers, and ultimately achieved an accuracy of about 90 percent in determining whether a comment was meant to be informative, a joke, a metaphor, ridicule, or fit other dimensions.