NBRI Comment Reports consist of verbatim responses to open-ended survey questions.
By definition, comments and words are qualitative data, so it is considered “soft” data in that it has not been proven to be representative of the thinking of the population. Clients often mistake non-representative survey comments for hard data, and risk being persuaded by them. The danger lies in the Psychology of Repetition which states that the human brain registers a piece of information as fact when one has been exposed to it as little as three times. So, when management reads comments from 1,000 employees or customers, and the same subject matter is mentioned 3, 10, or 20 times, management will have the impression that the issue is a widely held fact within the organization, when it is actually the opinion of so few as to not even warrant discussion. NBRI recommends text analytics so that our clients are not misled by spurious information.
Text Analytics Reports written by an NBRI organizational psychologist are generated using artificial neural networking and managed machine learning to extract the most meaningful information from responses to open-ended questions. These responses undergo sentiment and word frequency analysis to identify the highest impact themes, and determine the ratios of positive, negative, and neutral/mixed comments for each theme. In combination with the Comment Reports, this analysis provides color and context to the hard data of the quantitative analyses presented in the Data Reports and Executive Summary.
One or more open-ended questions may also be included in the survey to identify any themes that may have been omitted so that future surveys can be modified before subsequent deployments. For instance, “Benefits” may be a common theme, and “Health Insurance” may be a salient aspect of the theme that was not included in the survey. NBRI’s text analytics are most helpful in this endeavor.
Upon request, proper names and profanity may be removed from the verbatim comment reports, and comments may be sorted by any demographic used to sort the quantitative survey data in the Data Reports, such as division, business unit, or department.