People often have mixed feelings about a topic and can simultaneously see both the positive and negative sides of things. But new research, published in PLOS One, suggests that professional pollsters are failing to account for this ambivalence in their assessment of political attitudes.
The study, which collected data between 2017 and 2019, found that approximately 4 in 10 college students displayed some level of ambivalence towards President Donald Trump.
“This work grew out of a collaboration that my wife and I have. I am an atomic physicist, working in the area of atomic clocks; my wife Lori is a developmental psychologist at Whittier College working in the area of child forensic interviewing,” explained co-author James C. Camparo, who is an adjunct professor of physics.
“In a collaboration for a study we were doing on prejudice (i.e., questionnaire data with Likert-scale responses), I realized that there was a one-to-one mathematical connection between the responses to Likert-scale questionnaires (i.e., ordinal data) and the ‘responses of atoms’ when ‘asked’ (i.e., experimentally probed) about their quantum-energy state.”
“For a 5-point Likert scale, the respondent can pick one out of five ordered categories,” Camparo said. “For an atom with five quantum states, the atom will find itself in one of five ordered energy states. Interestingly, when we probe atoms regarding their quantum-energy state, we find that they can exist simultaneously in two discrete quantum-energy states; we say that the atom is in a quantum superposition state.”
“Taking the analogy to psychology, if a respondent were allowed to mark more than one response on a Likert scale, we could say that the respondent was in an attitudinal superposition state. In other words, we could know that the respondent was ambivalent regarding that particular questionnaire item.”
When people mark two responses on a survey item, their response is typically considered “bad data” and tossed out. But the researchers believe that such responses might provide important information about a respondent’s attitude — namely, their ambivalent state of mind.
In a previous study regarding attitudes towards affirmative action, the Camparos found that many people double-marked item responses when given the opportunity. The researchers used a density matrix, a statistical tool used in physics to describe the quantum state of a physical system, to measure ambivalence.
“We then wanted to examine this tool for assessing ambivalence in a more real-world situation, which led to the study we published in PLOS One,” Camparo said. “As we started to assess the data, however, we realized that we were uncovering a much more interesting story: political opinion pollsters may be missing ambivalence, and that ambivalence might have played an important role in Trump’s 2016 election win.”
In their new study, the researchers created 28-item questionnaires regarding attitudes towards Trump’s presidency using statements from actual Quinnipiac and Brookings Institute surveys. About 250 undergraduates indicated how much they agreed or disagreed with each statement on a 7-point scale. To measure ambivalence, the participants were informed that they could mark one or two responses per item.
For example, participants could respond to the statement “I would say that Donald Trump is someone who shares my values” with both “strongly agree” and “neutral” or with both “agree” and “disagree.”
The researchers found that approximately 40% of the participants displayed some level of ambivalence, regardless of their political affiliation. The findings indicate that “political opinion polling is likely missing ambivalence in their surveys, and that ambivalence may be playing a larger role in election outcomes than is presently appreciated,” Camparo told PsyPost.
The study also provides evidence that ambivalence can be measured using density matrices.
“The mathematical tools developed by physicists may have important possibilities for the social sciences more broadly (e.g., quantum-field theory),” Camparo explained. “However, these tools must be carefully re-phrased in the language of the social sciences. For example, just because one uses tools developed by physicists in their study of quantum particles, one shouldn’t leap to the conclusion that our minds and social interactions are ‘quantum’ in nature.”
The study, “Are political-opinion pollsters missing ambivalence: “I love Trump”… “I hate Trump”” was published March 11, 2021.