Your Second Thoughts May Be Distorted If You Have Anxiety, Study Finds

Scientists Pinpoint Why Overthinking Makes Anxiety-Driven Doubt Worse

Taking a moment to reflect before judging your own performance sounds like solid advice. But new research suggests that for people experiencing anxiety symptoms, pausing to second-guess themselves might deepen their self-doubt.

Scientists at University College London and the University of Copenhagen have discovered that anxiety-driven underconfidence grows stronger the longer someone takes to judge how they did on a decision they just made. The study, published in Psychological Medicine, analyzed confidence ratings from 1,447 people across multiple decision-making tasks. For anxious individuals, extra time for introspection actually makes their self-doubt worse.

The researchers found that people with higher anxiety symptoms accumulate negatively biased thoughts about their performance over time, deepening underconfidence as they spend more time reflecting.

How Scientists Measured Confidence in Real-Time

Researchers pooled data from four experiments where participants completed visual decision-making tasks, such as identifying which color appeared more frequently on screen or remembering objects from a brief display. After each decision, participants rated their confidence on a scale, and scientists measured how long they took to provide that rating.

The study collected information about participants’ anxiety levels using standardized questionnaires and tracked gender differences across the sample. Sample sizes for individual experiments ranged from 300 to 433 participants, all recruited online. The researchers excluded trials with unusually fast or slow response times to ensure data quality.

Anxiety and the Snowball Effect

The key finding emerged when researchers plotted confidence levels against the time participants took to report them. For people with higher anxiety symptoms, confidence dropped more steeply as they spent additional seconds contemplating their performance. The gap between low-anxiety and high-anxiety individuals widened with every passing moment of reflection.

This pattern held across multiple experiments and different types of tasks, showing the effect is robust and not limited to one specific scenario.

What makes this particularly notable is that the effect worked in the opposite direction for gender-related underconfidence. On average, women reported lower confidence than men immediately after making decisions, but that gap narrowed when they took more time to reflect. By the longest response times, the gender difference in confidence essentially disappeared.

Computational Models Reveal Why Time Helps Some but Hurts Others

To understand what drives these opposite patterns, researchers built a computational model based on drift-diffusion theory, which treats decision-making as an evidence accumulation process. Information builds up over time until it crosses a threshold that triggers a choice.

The team extended this framework to model different sources of distorted confidence: two kinds of confidence drag. One involves starting with a lower baseline of confidence. The other involves a downward pull that grows the longer someone thinks about their performance.

When they fit this model to individual participants’ data, anxiety stood out because confidence got worse with more reflection time, suggesting negative self-doubt built up during that extra time. People with anxiety symptoms showed both a lower starting point for confidence and this ongoing accumulation of negative thoughts about their performance.

Gender-related underconfidence, by contrast, stemmed only from that lower starting point, without the ongoing buildup of negative thoughts.

What the Patterns Might Mean

One possibility is that extra reflection gives anxious people more room to replay mistakes or imagine they got it wrong, even when their performance hasn’t changed. For the gender gap, extra time may let people double-check the decision in their mind and override a snap feeling of being probably wrong.

The study also offers a potential window into why anxiety might fuel rumination: the tendency to get stuck in negative thought loops. If people with anxiety symptoms are accumulating biased negative evidence about their abilities over time, extended reflection literally makes them feel worse about themselves, creating a feedback loop that reinforces self-doubt.

It’s worth noting that this study looked at split-second confidence ratings in simple online tasks, not treatment approaches or real-world performance reviews. The study focused on immediate post-decision reflection, typically lasting only a few seconds. Whether these patterns extend to longer-term rumination or self-evaluation remains an open question. The tasks used were also relatively simple perceptual and memory challenges, so whether the same dynamics would apply to evaluating performance in complex real-world situations like job interviews or creative projects is unclear.

In high-pressure moments, the first wave of doubt may not be useful information. But bigger real-world decisions still deserve careful thought, so the challenge may be knowing when someone is genuinely reflecting versus spiraling.

---