Are people really blind to their own biases, or have we been overselling the power of the unconscious mind? For years, psychology classes and popular talks have claimed that hidden attitudes quietly steer our choices when we least expect it. A new 2025 review in the Annual Review of Psychology urges a rethink. It argues that much of the “unconscious bias” story rests on methods and assumptions that don’t always hold up.

The authors, Bertram Gawronski and Olivier Corneille, examine three big claims about unawareness: not knowing your attitude, not knowing what shaped it, and not knowing how it guides behavior. Across each claim, they find that people often have more insight than the field assumes. For example, many can predict their scores on indirect tests like the IAT, even without taking the test first.

This doesn’t mean awareness is perfect. It means the evidence for deep, pervasive unawareness is thinner than advertised. In this article, you’ll see where the classic arguments stumble, when unconscious processes might still matter, and how better methods can tell the difference between true unawareness and simpler explanations like measurement quirks, social pressure, or plain forgetting.

Why claims about unawareness matter in psychology

Many psychological theories suggest that people frequently lack awareness of their mental processes, particularly regarding their attitudes. This assumption has shaped decades of research, particularly in social psychology, where concepts like implicit bias and unconscious prejudice have been used to explain why people behave in ways that seem inconsistent with their stated beliefs.

But how accurate is this assumption? Are people truly blind to their own attitudes, or are researchers overestimating how hidden these mental states really are? A comprehensive 2025 review published in the Annual Review of Psychology critically examined the empirical evidence behind three core claims about unawareness in attitude research. The findings call into question just how widespread or meaningful this type of unawareness actually is.

The three kinds of “unawareness” under scrutiny

The review, authored by Bertram Gawronski and Olivier Corneille, identified three distinct forms of unawareness: (1) unawareness of the attitude itself, (2) unawareness of the environmental causes that shaped the attitude, and (3) unawareness of how that attitude influences behavior. Each of these claims has become almost axiomatic in psychology, frequently cited to explain biased decision-making, discriminatory behavior, and irrational preferences.

However, the authors argue that while these concepts are compelling, the actual evidence behind them is weaker than often assumed. By breaking down each form of unawareness and analyzing the methodological challenges involved, the review highlights the need for more rigorous standards before making claims about unconscious attitudes.

Are we unaware of our own attitudes?

Discrepancies between “implicit” and self-reported measures

In social psychology, a key belief is that individuals can have implicit attitudes that they are either unable or unwilling to disclose. This claim is often based on studies showing low correlations between self-reported attitudes and scores on indirect measures like the Implicit Association Test (IAT). These discrepancies are typically interpreted as evidence that people are unaware of their true attitudes.

But the review challenges this interpretation. In many cases, people can accurately predict how they will score on implicit measures like the IAT, even without prior experience with the test. This suggests they may be more aware of their biases than the discrepancies imply. Instead of uncovering concealed attitudes, these inconsistencies may stem from differences in how each method gathers information, such as differences in timing, task demands, or the influence of social desirability.

The limitations of physiological, indirect, and behavioral indicators

Researchers have attempted to avoid self-report entirely by relying on physiological measures, such as brain activity or skin conductance, and behavioral evidence, including consumer decisions or nonverbal behaviors, as supposedly more objective signs of attitude. When these diverge from what people say, unawareness is often blamed.

Yet this strategy has its own problems. Physiological measures can be noisy, imprecise, and influenced by factors unrelated to attitude, such as arousal or attention. Behavioral measures, on the other hand, are influenced not just by personal preferences but by various factors such as social norms, habits, or contextual cues. Treating these measures as a gold standard for “true” attitudes risks circular logic, especially when they are validated by the very self-reports they’re supposed to improve upon.

What people can tell about their own biases and where they fall short

One significant discovery from recent studies is that people are often quite accurate in predicting their own results on implicit measures, at least when judging themselves. Although they may be unaware of their place compared to others, they frequently recognize, for instance, that they hold stronger biases toward one group over another. This internal accuracy undermines the idea that such biases are entirely unconscious.

However, this doesn’t mean all self-insight is perfect. Some people show poor accuracy in predicting their scores, suggesting individual differences in self-awareness. And when people are surprised by their implicit feedback, it may reflect not a lack of awareness, but a mismatch in how different systems quantify and label bias. In short, the surprise may stem more from a flawed metric than from unconscious prejudice.

Environmental causes: Do we know why we feel the way we do?

Evaluative conditioning and mere exposure under the microscope

Two of the most widely examined processes through which attitudes develop, often outside of conscious awareness, are evaluative conditioning (EC) and the mere exposure effect (ME). EC refers to forming an attitude toward a neutral stimulus simply by pairing it with something positive or negative. ME refers to increased liking for things simply because we’ve seen them before. Both have been cited as examples of attitudes forming outside conscious awareness.

But the review finds that many of these claims are overstated. Strong evidence for EC occurring outside of awareness is surprisingly scarce. Early studies used subliminal stimuli or attention-diverting tasks, but more rigorous designs often fail to replicate these effects when awareness is properly measured. In many cases, EC seems to require attention, memory, or conscious recollection to produce reliable attitude change.

Why attention, memory, and stimulus strength complicate conclusions

Much of the confusion stems from the fact that unawareness is often confounded with other factors like stimulus strength or divided attention. For instance, studies claiming that people learned attitudes from subliminal images often used outdated tech or unreliable awareness checks. When newer studies improve these controls, the unconscious effects frequently disappear.

Likewise, memory plays a complicated role. Some research shows EC and ME effects even when participants fail to consciously recall the pairings or exposures, but this does not necessarily mean they lacked awareness during learning. Lack of later recollection isn’t the same as absence of awareness at the time of the stimulus. This distinction is often overlooked in experimental designs and interpretations.

Evidence for and against unconscious influence of environmental cues

The idea that environmental cues shape our attitudes without us realizing it is intuitively appealing. But the review urges caution. While people may not always be able to name the exact moment their attitudes shifted, this doesn’t mean they were unaware of the influencing event. For instance, individuals might draw on a sense of familiarity or ease, which are conscious experiences that influence their judgments even if the initial source is forgotten.

Interestingly, some studies show that when people are aware of their exposure, the effects of EC and ME are stronger and more consistent. This suggests that awareness might play a more active role in attitude formation than previously thought. It may not be that we’re blind to what shapes our preferences, but that we underestimate how much these everyday cues matter.

Also Read

Behavioral effects: Do our attitudes shape behavior without us knowing?

Awareness of actions vs. awareness of influence

A key question in attitude research is whether our behavior can be guided by attitudes we are unaware of, or if we can act on known attitudes without recognizing their influence. These two types of unawareness are often lumped together but are conceptually distinct. In the first case, the individual may be unaware they’re even performing a behavior. In the second, they may be aware of their behavior but unaware of what caused it.

The review indicates that the first type, being unaware of the behavior itself, is likely to occur mainly in low-level, automatic reactions such as brief facial expressions or subtle eye movements. But for more deliberate actions, like choosing a job applicant or voting in an election, people are typically aware of what they’re doing. The issue then becomes whether they understand the psychological forces behind their decisions.

When evaluations bias information weighting and interpretation

Even in deliberate decision making, attitudes can influence how people evaluate or interpret information, often without  them realizing it. For example, if someone holds a subtle gender bias, they might unconsciously prioritize certain qualities in male job candidates and different qualities in female ones, justifying their decision with “objective” criteria that conveniently match their underlying preference.

This process, known as biased information weighting, is particularly insidious because it doesn’t require overt prejudice or conscious intent. A similar mechanism operates when people interpret ambiguous behavior. Studies suggest that racial attitudes can influence how people judge the same behavior differently, depending on the race of the person performing it. While the behavior is visible and the judgment deliberate, the influence of attitudes on perception may fly under the radar.

Why proving unawareness is methodologically difficult

Proving that a person is unaware of a behavior or the attitude behind it is not only challenging but also inherently complex. Awareness is a slippery concept to measure, and asking people whether they were influenced by an attitude often raises their awareness in the process. This makes it hard to design studies that can truly distinguish between unawareness and unwillingness to acknowledge a bias.

Adding to the complexity, some studies rely on indirect methods and interpret discrepancies as evidence of unconscious influence. However, unless we could rule out alternative explanations such as social desirability, measurement error, or task framing, these conclusions remain uncertain. The review emphasizes that compelling evidence for unawareness of behavioral effects is still lacking, despite the theoretical appeal of the idea.

The bigger issue: Misinterpreting awareness gaps

Why failing to report a bias doesn’t always mean unconsciousness

Psychological researches assume that if someone is unable or would prefer not to report a bias, it must be unconscious. But this leap is often unwarranted. There are many reasons someone might fail to acknowledge a bias, including embarrassment, confusion about the question, or disagreement with the experimenter’s framing.

Just because people don’t offer the “correct” explanation for their behavior doesn’t mean they’re unaware of it. They may be aware but unwilling to say so. Or they may conceptualize their behavior differently than researchers do. Interpreting these gaps as unawareness risks overstating the unconscious nature of everyday decisions and downplaying the complexity of self-insight.

The danger of relying on flawed methods or assumptions

Many early studies on unconscious attitudes used methods that lacked rigorous controls or relied on questionable assumptions. For example, the assumption that physiological responses reflect “true” attitudes has been undermined by findings showing that these responses can vary due to many non-attitudinal factors. Similarly, indirect measures like the IAT have been interpreted as uncovering hidden biases, but they may simply reflect associations people are already aware of.

Even more sophisticated paradigms can fall into the same trap if they don’t properly distinguish between lack of awareness and other explanations, like low reliability or conflicting motivations. The review warns that researchers often underestimate how much of their findings can be explained by methodological artifacts or alternative psychological processes.

Differentiating attitudes from stereotypes, beliefs, and social norms

Part of the confusion stems from a lack of clarity about what’s actually being measured. Attitudes are often conflated with other psychological constructs like stereotypes, semantic beliefs, or cultural norms. But these are not interchangeable. For example, someone might believe that men are more common in leadership roles (a belief), without actually preferring male leaders (an attitude).

When these constructs are not clearly distinguished, it becomes challenging to identify whether attitudes are truly shaping behavior and if that influence is conscious. If a hiring bias reflects cultural expectations rather than personal evaluation, attributing it to an unconscious attitude may miss the point. Untangling these influences is crucial for understanding the real source of biased actions.

Methodological roadblocks and recommendations

Problems with self-report, indirect tests, and between-subject designs

Much of the confusion in this field stems from the tools used to measure attitudes and awareness. Self-report measures are easy to administer but vulnerable to social desirability and self-presentation concerns. Indirect tests such as the IAT aim to avoid those limitations but with their own issues such as low reliability, unclear validity, and sensitivity to task design.

Between-subjects designs, where researchers compare different individuals’ responses, introduce another layer of complexity. These studies operate on the assumption that individuals are aware not only of their own feelings but also of how they measure up against others, which sets a high level of awareness. Low correlations between subjective and objective measures in such designs may reflect ignorance about others, not about the self.

How to test for awareness rigorously without introducing awareness

One of the central paradoxes in awareness research is that trying to measure awareness can change it. Simply asking participants to reflect on a behavior or feeling can make them more aware of it. This presents a serious challenge for designing experiments that aim to detect true unawareness.

Some researchers have tried to solve this by measuring awareness indirectly, through prediction tasks or concurrent judgments, rather than retrospective debriefs. Others use within-subject designs that compare a person’s evaluations across multiple stimuli, allowing researchers to isolate discrepancies without requiring self-other comparisons. These methods offer promising alternatives but are still underused in the field.

Proposed best practices for future studies

The review offers a roadmap for better awareness research. First, researchers should avoid treating indirect measures as automatically diagnostic of unawareness. Second, studies should keep procedural elements consistent between awareness and evaluative measures, including factors like timing, reliability, and task difficulty, to prevent confounding variables.

Third, before claiming unawareness, researchers must rule out more plausible explanations, like unwillingness to report, lack of insight into others’ views, or poor measurement design. And finally, when possible, studies should manipulate attention or memory directly, rather than relying solely on correlational data. These methodological improvements could clarify when, how, and to what extent attitudes operate outside of awareness.

Looking ahead: What do we actually know?

When attitudes might truly evade awareness

Despite the widespread skepticism expressed in the review, the authors do not entirely dismiss the possibility that some attitudes may genuinely operate outside of awareness. One area that holds promise is the discrepancy between people’s evaluations of abstract categories (types) and specific examples (tokens). For instance, someone might claim to dislike a category like “red wine,” but consistently rate individual red wines highly in blind tastings. These inconsistencies may reflect flawed generalizations rather than deliberate self-deception.

Unconscious attitudes may also arise from quick, automatic reactions that occur before conscious reflection can step in, such as reflexively flinching at a noise or exhibiting micro-expressions in response to emotionally charged events. However, even in these cases, it remains difficult to disentangle true unawareness from rapid, but conscious, processing. Researchers still face significant hurdles in proving these effects are truly unconscious rather than merely uncontrolled or habitual.

The role of meta-cognition and self-perception in gaining insight

Becoming aware of one’s own attitudes usually relies on meta-cognition which is the ability to evaluate one’s own thinking. People can form beliefs about their preferences by observing their own behaviors over time, as proposed in self-perception theory. For instance, if a person realizes they regularly avoid a certain type of individual or object, they might think that they possess a negative attitude toward it, even if they had never consciously identified it before.

Importantly, awareness may not emerge through introspection alone. External feedback, patterns of decision-making, and exposure to social norms can all shape how individuals come to understand their attitudes. Awareness then, as this study implies, is not a fixed trait but rather an ongoing process that develops across various contexts and relies greatly on reflection, memory, and motivation.

Individual differences in self-awareness

Not everyone has the same level of insight into their own attitudes. Some people show remarkable accuracy in predicting their own scores on implicit measures, while others perform poorly. These individual differences could stem from personality traits, cognitive ability, or even cultural norms about self-reflection. This variability is often overlooked in studies that emphasize group-level effects.

The review highlights that understanding these differences may be key to advancing the field. If some people are more attuned to their internal states than others, researchers should not assume a uniform level of unawareness across participants. Future studies that directly include measures of meta-cognitive ability or self-reflectiveness could provide a more detailed insight of when attitudes operate outside of our awareness and for whom this occurs.

Bottom line: The case for unconscious attitudes is weaker than we thought

Most evidence for unawareness doesn’t hold up under scrutiny

After decades of bold claims about hidden biases and unconscious influences, this review offers a sobering reassessment. Many of the key findings that have been used to support the idea of unconscious attitudes, such as gaps between indirect and direct measures, lose strength when studied through a more rigorous methodology. Apparent evidence for unawareness often dissolves when researchers account for confounds like measurement reliability, task structure, and participant expectations.

The conclusion isn’t that attitudes never operate without awareness, but that the empirical foundation for such claims is much shakier than it appears in textbooks or TED Talks. Most studies that assert evidence of unconscious influence do not eliminate more mundane explanations, such as forgetting, social discomfort, or a lack of motivation to reflect.

Awareness may be more prevalent and more complex than assumed

Rather than treating unawareness as the default, the review suggests a shift in focus: toward understanding the conditions under which people do become aware of their attitudes, and how that awareness emerges over time. People may not always have perfect insight, but they often know more than researchers assume. In fact, when measured properly, individuals show surprising accuracy in predicting their own evaluative tendencies, even on tasks designed to bypass introspection.

Ultimately, the concept of unconscious attitudes is too significant to dismiss, yet too uncertain to accept uncritically. Future research will need to raise the bar for evidence and refine its definitions. What’s needed now isn’t more sweeping claims about hidden minds, but careful, well-designed studies that recognize the complexity of human self-knowledge.

Sources: