DATE: October 16, 2026 at 10:00AM
SOURCE: PSYPOST.ORG
** Research quality varies widely from fantastic to small exploratory studies. Please check research methods when conclusions are very important to you. **
-------------------------------------------------
TITLE: Many people judge candidates and policies by a single deciding feature, new research suggests
When people make complex political decisions, they tend to rely on simple mental shortcuts rather than exhaustively weighing every piece of information, according to a study published in PNAS. The research indicates that people frequently switch between different decision strategies depending on the issue at hand and the complexity of the choice environment. These findings suggest that standard survey methods in political science, which often assume people act as fully thorough evaluators, may obscure how individuals actually reach conclusions.
Social scientists frequently rely on choice-based conjoint experiments to measure public opinion on complex topics. In this type of survey, participants view side-by-side profiles of two options, such as hypothetical political candidates or policy proposals, with features that vary randomly, such as age, profession, or cost. Participants are then asked to pick the option they prefer, allowing researchers to infer how much specific features sway voter preferences on average across the entire population.
A key open question is whether people answering these multi-feature surveys actually weigh every detail, as many standard analysis methods implicitly assume, or rely on simplifying mental shortcuts known as heuristics. Some evidence points toward shortcuts. A 2020 eye-tracking study found that when survey tables included more features and options, people looked at a smaller fraction of the information on screen. In the consumer realm, a 2009 study found that although standard models assume people balance every feature against the others, a sizable minority of participants were better described by rules that rank features and decide based on the most important one.
Similar rankings appear in politics: a 2024 study of candidate choice found that voters tend to focus on party affiliation first, weighing traits like gender or race mainly when two candidates share the same party. Separately, research covered by PsyPost in 2026 found that voters often use left and right political labels as shortcuts to guess where candidates stand.
To understand the cognitive rules guiding these choices, study authors Nidhi V. Banavar, a postdoctoral scholar in the Travers Department of Political Science at the University of California, Berkeley, and Kirk Bansak, also of the University of California, Berkeley, created a mathematical framework to classify individual decision strategies. Their model analyzes a person’s survey selections to infer what attributes matter to them, which values they favor, and which cognitive rule best accounts for their behavior.
“We were interested in the question of how people go about making political choices,” Banavar told PsyPost. “When we estimate complex constructs like political attitudes, a lot of analysis methods implicitly assume that people weigh everything in front of them. We wanted to test whether that was true, and if not, to find how they were deciding instead.”
“This connects to a broader issue, which is that most political attitude research is done at the aggregate level,” Banavar explained. “The aggregate is perfect for giving us the high level snapshot and, for better or for worse, collapses a lot of the variation we see in people: both in terms of their actual preferences and in terms of how they make choices based on those preferences. We were interested in uncollapsing the aggregate and so built a model and designed experiments to do so.”
The model examines four distinct decision strategies that differ in cognitive demand. The most thorough strategy, known as Weighted Additive, involves a rational calculation where an individual weighs every single attribute and sums them up to pick the highest-scoring option. A simpler shortcut strategy, known as Tally, involves counting how many preferred traits each option possesses and choosing the profile with the higher count, without worrying about which traits matter more.
The simplest shortcut rule, called Take The Best, directs a person to choose based solely on whichever single feature matters most to them; they only examine a second feature if the options are tied on their favorite trait. The fourth classification, Contaminant, captures random guessing equivalent to a coin flip.
The scientists evaluated this model across two experiments involving 3,762 American adults recruited through Prolific, with quotas matching national demographics for age, sex, and political affiliation. Participants completed 90 choices divided into three blocks of 30 decisions across three common political topics: selecting between two hypothetical immigrants seeking admission to the United States, choosing between two Senate primary candidates, and choosing between two national climate policy packages. Each option varied across six attributes that each had two possible values.
In the first experiment, which included 1,964 participants, the researchers altered the complexity of the profiles by testing whether allowing ties changed how people decided. Half of the participants saw profiles where traits were never identical between the two options, while the other half saw profiles that contained one to three matching features, such as both candidates being in their 50s. This design was intended to test whether forcing people to look beyond their top attribute would push them toward more thorough evaluation strategies.
The results showed that the simplest shortcut rule, Take The Best, was the most common strategy across all three political domains. In the condition without ties, about 54% of participants were classified as using this single-attribute rule, while roughly 30% were classified as using the rational Weighted Additive strategy, and about 14% used the Tally strategy. When ties were introduced, forcing people to scan deeper into the profiles, use of the Take The Best rule dropped to roughly 45%, while rational strategy use increased to about 40%. Across both conditions, shortcut strategies accounted for the majority of participant behavior.
The authors also discovered that an individual’s strategy was not fixed. Depending on the condition, 67% to 76% of participants were classified as using different strategies for different issues, with about 10% using a completely different strategy across all three domains. The presentation order of the topics was randomized, and participants did not display signs of survey fatigue or increased guessing in later rounds. Instead, people appeared to adapt their decision rules to the specific context of each topic.
“On one hand, the big takeaway is the prevalence of heuristic use and how the majority of individuals use informational shortcuts,” Banavar noted. “This is an important finding people are not exhaustively weighting and evaluating all the information presented to them.”
“However, I think a bigger takeaway is that people tend to use different strategies depending on the political domain,” Banavar added. “How someone processes information is not a trait: just because I may choose a candidate based on one factor doesn’t mean that I’ll choose a climate policy based on one factor. This means we shouldn’t put people into buckets based on how they make political choices.”
In the second experiment, which included 1,798 participants, the researchers tested another common variation in survey design: asking participants to rate each profile individually on a 1-to-7 scale in addition to making their comparative choice. Among those asked to give ratings, fewer participants were classified as using Take The Best in all three policy domains, though it remained the most common strategy. That group also had more people classified as using the more thorough strategies, along with a small increase in people classified as random guessers, which the authors suggest may reflect some respondents losing interest. The authors caution that these shifts were not consistently statistically reliable across domains.
To check whether these classifications reflected something about how people were actually thinking, rather than just statistical patterns in their choices, the researchers analyzed response times. Participants classified as using the rational Weighted Additive strategy took the longest to answer, followed by those classified as using Tally, while those classified as using Take The Best responded much faster.
When ties were present, only the Take The Best group answered more slowly; response times for the other groups were essentially unchanged. Because this shortcut requires checking additional traits only when top traits tie, the authors interpret this selective slowdown as evidence that the classifications capture real differences in mental processing, though they stop short of claiming the model directly observes thought.
“While we note that our findings were obtained (necessarily) in the context of a limited set of experiments, we do believe their theoretical and practical value has generalizability,” Banavar pointed out. “For instance, while the precise pervasiveness of heuristic usage is certainly context dependent, we still found that the simplest heuristic strategy was our most commonly classified strategy across all of our experiments, regardless of complexity manipulations. Our finding that strategy classifications are not fixed for individual people was also robust, and we have no reason to believe there would be within-person strategy variation only for our experiments and not for others.”
The researchers also investigated whether demographic characteristics predicted which strategy a person chose. Across variables such as race, gender, education, and political affiliation, associations were generally weak, and most were not statistically distinguishable from chance after adjusting for multiple tests. Older participants showed a slight tendency toward Take The Best, but the effect was small and held up after statistical correction in only 2 of 12 tests. Participants who rated a topic as less complex, or who said they knew less about it, also showed slight tendencies toward the shortcut, but these patterns appeared in only some analyses and largely did not survive correction.
“I was surprised by how difficult it was to relate strategy classification to respondent characteristics,” Banavar observed. “There’s a lot of evidence in cognitive science research showing that this is a difficult problem. There is also, however, substantial work in political science that would generate some predictions about relating strategy classification and respondent characteristics.”
“In the end we found very little, even though we considered demographics and various other measures of how meaningful a given political topic was to each person,” she said. “I think there’s a lot more to explore on this front.”
When the authors broke aggregate preferences down into individual weightings, they found that population-level averages concealed large differences. For every feature, many participants appeared to give it little or no weight, and for several of the most heavily weighted features, the pattern split into two sharp groups: people who effectively ignored the feature and a distinct group who weighted it very heavily. For example, conjoint studies commonly find that female candidates are slightly preferred over male candidates on average; the new results suggest this average blends people who prefer women, people who prefer men, and many who put little weight on gender at all.
As with all research, there are some caveats to consider. The study used simplified experimental profiles where each feature only had two possible choices, such as a candidate being either in their 50s or 70s. In real-world political decisions, options often have many more layers and continuous values, which might prompt respondents to rely even more heavily on mental shortcuts than observed here.
There are also important nuances in how mental shortcuts should be understood, especially regarding whether they reflect poorly on respondents. “First, the fact that we classify the majority of individuals as using heuristics or not using the rational strategy does not mean that they are lazy,” Banavar stressed. “We have finite resources and it is a good thing to selectively deploy them.”
The classification approach has its own limits. Because strategies were inferred from choices, someone who weighs every feature but cares overwhelmingly about one would make the same choices as a Take The Best user, and the model would classify them the same way; the response time data suggest such mix-ups were not widespread. Simulations also showed that the model tends to classify random guessers as Tally users, so guessing may be somewhat more common, and Tally somewhat less common, than the estimates suggest.
“Secondly, while we have built a cognitive model, the model doesn’t actually tell us what’s happening inside someone’s head,” Banavar stated. “It’s a model that classifies, based on the choices that people make, what decision strategies best describe their choices. It is a technical clarification but an important one.”
Future research could expand on this framework by testing features with more than two values, exploring why people pick the strategies they do, and modeling response times alongside choices.
The study, “Classifying strategy use in multiattribute subjective choice: Application to conjoint experiments in political science,” was authored by Nidhi V. Banavar and Kirk Bansak.
-------------------------------------------------
Private, vetted email list for mental health professionals: https://www.clinicians-exchange.org
Unofficial Psychology Today Xitter to toot feed at Psych Today Unofficial Bot @PTUnofficialBot
-------------------------------------------------
#psychology #counseling #socialwork #psychotherapy @psychotherapist @psychotherapists @psychology @socialpsych @socialwork @psychiatry #mentalhealth #psychiatry #healthcare #depression #psychotherapist #PoliticalDecisionMaking #HeuristicsInPolitics #ConjointAnalysis #TakeTheBest #WeightedAdditive #TallyStrategy #PolicyDecisionCognition #SurveyDesign #VoterBehavior #CognitiveStrategies