In past decades, consumer surveys have become an established form of evidence in trademark and false advertising litigation, particularly in matters involving confusion, secondary meaning, or false or misleading advertising.[i] However, parties that provide surveys face the risk of a motion to exclude under Daubert v. Merrill Dow Pharmaceuticals[ii]—which, if successful, would mean that the party cannot use the survey evidence they commissioned, or the expert they commissioned it from. This makes it very important to present a survey without flaws significant enough to warrant exclusion.
Several widely cited authorities address the proper conduct of surveys, such as research design, the use of controls, the management of interviewers, or the selection of appropriate respondents. Another important issue that impacts the reliability of a survey is how to phrase trademark and false advertising survey questions, which is an issue that courts often examine. Below, we provide three tips that can aid in properly designing survey questions and help decision-makers determine whether the questions are likely to produce reliable data.
Tip 1: Survey questions must be written in a clear and understandable manner.
The first rule for writing litigation survey questions is fundamental: they must be written in a way that can be clearly understood by respondents taking the survey.[iii] Although this guideline is straightforward, it has several profound and complex implications, such as:
- Survey items must be written using language and phrasing that can be understood by respondents for that particular survey. For example, a survey written for a respondent database of medical professionals may phrase questions differently than one written for primary grocery shoppers. The phrasing of survey questions should also be designed in a manner that does not assume that respondents have a lawyer’s understanding of the law. [iv]
- Survey questions must ask about a single construct, so that when respondents answer the question, respondents and experts know what topic they were addressing in their answer.[v]
This second rule is also known as a prohibition on “double-barreled” questions. One authority provides an example: “Do you regularly take vitamins to avoid getting sick?” Although this seems simple, it really asks two questions: “Do you regularly take vitamins?” and “Do you take them to avoid getting sick?”[vi] The question’s writer assumes that the respondent takes vitamins to avoid getting sick. A respondent who has some other reason to take vitamins can answer no, which is incorrect as to taking vitamins, or yes, which is incorrect as to their reasons for taking vitamins.
Tip 2: Survey questions must be phrased to address the issues in dispute.
Another important guideline for survey questions is that they should be phrased to address the specific issues in dispute.[vii] There are a variety of different types of confusion, which might include:
- Confusion as to source: When a consumer mistakenly believes that the company that made one product also made another product
- Confusion as to connection: When a consumer mistakenly believes that the companies that make two products are connected, perhaps by means such as sponsorship, affiliation, permission, or endorsement[viii]
The questions for an Eveready survey begin in a manner similar to the original survey upon which this format is based, which asked, “Who do you think puts out the [product] shown here?” and followed up with, “What makes you think so?” Since then, Eveready-format surveys have asked about other types of confusion, such as authorization, permission, sponsorship, or approval.[ix] For example, Starbucks Coffee Company v. Marshall S. Ruben involved a survey focused on company names. The first question asked for the name of the company, and follow-up questions asked, “Do you think the company that owns this retail establishment is connected or affiliated with any other company? What other company?” A final set of questions asked about confusion as to authorization, permission, or approval.[x]
In conducting surveys for evidence in cases involving secondary meaning, surveys often conform to the anonymous source rule, which was described in GMC v. Lanard Toys as indicating that “knowledge that a product comes from a single source, even without naming that source, is sufficient to establish secondary meaning.” In other words, when following the anonymous source rule, respondents simply need to indicate that they associate a trademark with one source, even if they cannot identify the source.[xi]
In surveys conducted for false or deceptive advertising disputes, one important rule is to ask about what consumers believe the advertisement says about the product—not what consumers themselves believe about the product.[xii] This is important because advertisers are liable only for reasonable consumers’ interpretations of their ads.[xiii] Consumers may come to surveys with preexisting beliefs about products, but these are a form of bias that survey designers must control for; they are not useful evidence in false advertising cases.[xiv]
Tip 3: Survey questions must be phrased in a non-biased and non-leading manner.
The need for litigation survey questions to avoid biased and leading language can impact the survey’s design and implementation in several ways, including:
- Rotate responses to reduce bias on key questions
- Use neutral phrasing to balance positives and negatives and to prevent respondents from interpreting survey wording as subtle cues
- Include “Don’t know/Not sure” response options to discourage guessing
First, the responses to significant questions should be rotated. When the order of response options is randomized, it reduces the effect of biases that may occur if some respondents tend to pick responses in a certain position, such as the first or last option. With rotation, each response appears first or last an equal percentage of the time.[xv]
Secondly, the survey question should be phrased in a neutral manner. Typically, this means it must include both positive and negative phrasing. For example, in a false advertising study, researchers might want to investigate whether a commercial implies a certain message. The question to ask about that message would likely start with phrasing similar to “Does or does not the commercial communicate or imply that …” This phrasing, and others like it, are equally balanced between the positive and the negative.[xvi]
Similarly, a survey must not imply or suggest that the researchers are seeking a particular type of answer. It is well known that some survey respondents try to be helpful in answering questions. Just as in everyday conversation, respondents expect the communications they receive in survey questions to be informative and relevant; however, unlike in an everyday conversation, their ability to ask clarifying questions is severely restricted during an interview. As a result, respondents attempt to gather information during interviews, even when the inference they received was not actually intended by the interviewer. In other words, respondents may notice subtle cues from the way products are presented, the order of question phrasing, and even how responses are presented. Interviewers must be careful to minimize these cues and make sure that they do not unduly influence the data gathered.[xvii]
A third, perhaps underappreciated, implication of this “non-biased and non-leading” guideline is the importance of providing a “don’t know” response to discourage guessing for all key questions in a survey. In the absence of this option, respondents could feel forced to select a response they may not actually prefer over “don’t know.” Not only is providing this response option critical to reduce guessing, but it lets respondents know that “don’t know” is an acceptable response, which is also important.[xviii] As one source has written, “By signaling to the respondent that it is appropriate not to have an opinion, the question reduces the demand for an answer and, as a result, the inclination to hazard a guess just to comply.”[xix]
In Summary
Litigation surveys can offer critical insights in disputes involving consumer perception—whether the issue is confusion, secondary meaning, or misleading advertising. When thoughtfully designed with clear, relevant, and unbiased questions, these surveys can provide compelling evidence that supports key claims and withstands judicial scrutiny. To maximize reliability and relevance, survey questions must be clearly understandable to respondents, directly aligned with the legal issues at hand, and carefully phrased to avoid bias or leading language.
Given the complexity and high stakes of intellectual property litigation, partnering with experienced professionals is essential. The Litigation Surveys & Consumer Science team at IMS Legal Strategies brings deep knowledge of survey methodology, legal standards, and courtroom expectations. We ensure that surveys are not only methodologically sound but also tailored to the goals of each case. From designing robust survey instruments to preparing defensible expert testimony, working with our trusted specialists will help you present compelling, admissible evidence that can influence outcomes in trademark disputes.
Visit imslegal.co.uk to meet our experts and request a consultation.
