Inkblot tests get AI makeover

July 21, 2020 //By Rich Pell
Inkblot tests get AI makeover
Consumer perception analytics firm Inkblot Analytics has released a new platform for AI-powered projective tests - personality tests designed to let users respond to ambiguous stimuli, such as inkblots, presumably revealing hidden emotions and internal conflicts projected by the user into the test.

The new platform, called Brand Blots , says the company, marries projective testing with artificial intelligence to innovate new projective tests and improve the way results are analyzed. It is designed to allow market researchers to administer, score, and interpret projective tests with the help of the company's proprietary algorithms.

Using the platform's scoring and profiling algorithms, says the company, Brand Blots can predict attitudinal and behavioral aspects about a respondent despite never having asked survey questions about these aspects. The core benefit for market researchers is to be able to "ask fewer questions while gaining more insights."

Historically, projective tests have been time intensive to administer - the average inkblot (or " Rorschach") test can take two to three hours according to some studies - making them difficult to use in market research. The new platform is built to remove the obstacles of time and effort by digitizing and automating the testing process, enabling researchers to easily administer over 36 types of online projective tests instantly. Using its algorithms, says the company, it has isolated what's important so that it could cut analysis down to five minutes while still retaining meaningful results.

In addition, says the company, the ability of a projective test to help marketers understand consumer behavior has been restricted by the limited analytical approaches previously used by researchers. In contrast, Brand Blots uses machine learning for a complex and data-driven analysis of projective test responses that can yield aspects about a respondent despite never having asked survey questions about those aspects.

The company says the platform has been tested and perfected using a number of large proprietary datasets, resulting in benchmarks for all its projective tests, allowing researchers to compare results of either an individual respondent or a sample of respondents with results of the general U.S. population. The ability to predict attitudinal and behavioral aspects about a respondent, and the ability to compare these


Vous êtes certain ?

Si vous désactivez les cookies, vous ne pouvez plus naviguer sur le site.

Vous allez être rediriger vers Google.