Presenting Comments
Comments present unique challenges for presentation.
( First page > Data Presentation > This page)

1. Consider the following possibilities:
   1. A comment makes a particularly pejorative comment about a key executive. He or she might respond defensively and lose their focus on the overall survey results.
   2. The CEO is singled out for a significant number of negative comments, so everyone starts worrying about how to take care of the CEO's reaction rather than attending to the implications of the data. Even if the CEO has enough Teflon in their blood to manage their defensiveness, others may not know that.
   3. A comment references an illegal act (one of my surveys had several comments about an unreported rape of an employee by her supervisor), which puts the company in the awkward spot of having to take some action, but having too little information to to take any intelligent action. It also means any concern for the overall survey results will suddenly be pushed to the background. When the CEO blurted out that he did not believe the allegation, the corporate counsel interrupted to say they were negotiating a six-figure settlement with the victim.
   4. A comment references something so specific that the author of the comment is obvious to many or, worse yet, to just one. For example, a respondent might share an anecdote from a performance evaluation that uniquely identifies them to the very same manager about which they are complaining.
   5. Respondents hear that the comments were handed out to managers and executives, and suddenly everyone is worried about whether their comment might identify them somehow.
   6. A manager decides that they know the source of a particularly damning comment. Whether they are accurate or not doesn't matter; they could take some recriminating action that will most likely get shared quickly through the rumor mill and sour the entire effort.
2. Read the comments for anything that might compromise the confidentiality of the respondent.
   1. Remember that the 50 people in a particular regional office may have generated only half a dozen comments. If those comments are coded for the office, a reader may be able (or think they are able) to pick out the source. Once you compromise the confidentiality of one respondent, you make the survey feel more dangerous for all respondents.
   2. A phrase or linguistic style can be as revealing as a signature. Be wary of language that is likely to point at a particular person.
   3. Some employees are quite comfortable revealing their identity. I've seen comments that were signed, included a phone number, or even asked that the CEO contact them for more detail. While it's refreshing that employees feel comfortable being so forthcoming, the rule of confidentiality still needs to be respected. If you reveal the identity of anyone (even with their permission), you raise the spectre of whether others might have had their identity revealed as well.
3. Comb through the comments, first for content, and then for making them more palatable.
   1. Consider replacing any names with that person's position. For example, mention of the woman running the Customer Service department could be replaced with "[Dept Head]". Be up front with the editorial practice; simply announce that names of individuals have been replaced with their position to make sure that we aren't distracted by single cases. In the rare case where a particular individual is mentioned dozens of times, handle that piece of data offline, not in the general data presentation.
      1. Keep in mind that comments about individuals in survey comments are seldom actionable. And if you have a manager that has provoked such voluminous response, it is highly likely that he or she is already known to many as one of the rotten apples in the basket. There is little value in adding public humiliation or accusation to the pile. There IS value is asking how a manager can be so widely perceived in such a poor light and still be allowed to retain their position. But that's another matter. If the survey is to be repeated, consider adding questions about managers who don't adhere to the company values on treatment of employees, or about managers showing respect and consideration for employees at all times. Let the data tell you if it's a company-wide problem or just an isolated case or disgruntled employee.
   2. Consider deleting comments that are extreme, vicious, or obscene. In any survey of a few hundred people, there will always be a few people hungry for an anonymous chance to vent their anger and rage against their (real or imagined) attackers. While it's fair to include a comment about the "few people who were quite angry about the merger", it will not serve you to put inflammatory language in front of your audience. In my early years I once let such a comment slip through, and the minute it was discovered the presentation was derailed and it took more than an hour to get people back on track. Within a day, the rumor mill had spread the message that the senior executives were "going to find out who had made such nasty comments".
4. Read the comments for evidence of topics missed in the survey.
   If employees list a pot pourri of topics in their comments, then the survey apparently covered the right topics. If, on the other hand, 40% of the comments mention child care, reduced medical benefits, incompetent IT staff, or poor treatment of women, then it is more likely that the survey has a hole in it.
5. Read the comments for emotional tone.
   There is a difference between apathy, annoyance, cynicism, and bitterness. They could all generate the same numerical data, but the comments will often reveal the more subtle tone behind the numbers. Conveying that information is often more valuable than just listing topics mentioned.
6. Consider a summary of your processing of the comments rather than distributing them in their entirety.
7. An excellent use of comments is to use them as examples of the sentiment behind the numbers. Find comments that reflect the dominant numeric results.

Back to Top