How to Improve Critical Thinking

You can improve critical thinking, though the gains are usually modest. What works is learning specific reasoning strategies on purpose and then practicing them on real problems, ideally with discussion and feedback. A 2015 meta-analysis pooled 341 effect sizes from experimental and quasi-experimental studies and found an average effect of g = 0.30 for critical thinking instruction. Dialogue, authentic problems and mentoring all had positive effects (Abrami et al., 2015). Knowing the subject you’re thinking about matters too (Willingham, 2019).

If you’re not sure what counts as critical thinking in the first place, start with our guide to what critical thinking is.

What research says works

Strategies that help, from the largest meta-analysis

Abrami and colleagues reviewed studies that measured critical thinking with standardized tests. They concluded that effective strategies exist for teaching critical thinking skills, both general and subject-specific, and critical thinking dispositions, at every educational level and in every subject area (Abrami et al., 2015). The results varied a lot from study to study (same source).

Researchers at the Technical University of Munich summarized the paper’s more detailed findings (Clearing House Unterricht, TUM):

What was studied	Effect size (g)	Number of studies
Combining authentic problems, dialogue and mentoring	0.57	19
Whole-class discussion led by a teacher	0.42	16
Small-group discussion led by a teacher	0.41	14
Teacher asking questions	0.38	19
Applying ideas to problem solving	0.35	31
Critical thinking dispositions as the outcome	0.23	25

The same summary says it made no significant difference whether critical thinking was taught as abstract principles in dedicated lessons or folded into regular subject content. The approach worked best when general principles were taught explicitly and then applied to concrete topics (same source).

Explicit instruction

Willingham’s 2019 review for the NSW Department of Education concludes that critical thinking skills should be explicitly taught and practiced. He cites evidence that simply exposing students to this kind of work, without explicit instruction, is less effective (Willingham, 2019). In one study he describes, psychology students who were taught principles for judging evidence, such as the difference between correlational research and true experiments, did better at evaluating psychology evidence than a comparison group (same source).

The 1990 APA Delphi panel described teaching the same way. Make the procedures explicit, model them, give feedback, and move from simple practice cases to realistically complex ones (Facione, 1990).

Arguing with other people

In a multiyear study, young adolescents built argument skills through electronic dialogues with peers on social issues. They then wrote better individual essays on new topics than a comparison group that had done extensive essay-writing practice plus whole-class discussion. They also paid more attention to whether evidence was relevant (Kuhn & Crowell, 2011).

Targeted debiasing

Some biases respond to direct training. In two experiments, one on how people weigh new evidence on social issues and one on how they test first impressions of others, telling participants to “consider the opposite” reduced bias more than telling them to be as fair and unbiased as possible (Lord et al., 1984). A one-shot debiasing training made graduate students less likely to pick an inferior, hypothesis-confirming solution in a surprise business case modeled on the Space Shuttle Challenger launch decision (Sellier et al., 2019). A 2025 study found that a single training session also reduced confirmation bias among national risk analysts and students (Heerma van Voss et al., 2025).

Practical habits that follow from the evidence

The studies above mostly involve classrooms and training sessions. These habits apply their findings to learning on your own:

1. Learn named strategies, then use them. Pick a concept, such as correlation vs. causation or a common fallacy, and apply it to articles you actually read. Explicit principles applied to real topics did best in the meta-analysis (Clearing House Unterricht, TUM). Our list of logical fallacies is a good place to start.
2. Build knowledge in the areas you care about. A principle like “look for hidden assumptions” won’t help much with an op-ed on a topic you know little about. Without background knowledge, you may not even understand the author’s claims (Willingham, 2019).
3. Compare examples that look different but work the same way. People tend to fixate on a problem’s surface details and miss that it has the same underlying structure as one they’ve already solved. Comparing two such problems side by side helps people spot that structure later (same source).
4. Discuss and disagree in good faith. Dialogue had positive effects in the meta-analysis (Abrami et al., 2015), and structured argument with peers built reasoning skills that carried over to writing (Kuhn & Crowell, 2011).
5. Ask what would be true if you were wrong. Considering the opposite is a tested way to counter confirmation bias (Lord et al., 1984).
6. Get feedback from someone more skilled. Mentoring had positive effects (Abrami et al., 2015), and feedback on performance is part of how the Delphi experts described teaching these skills (Facione, 1990).
7. Check your own work. Halpern’s model for teaching critical thinking includes a metacognitive step: checking for accuracy and monitoring your progress toward the goal (Halpern, 1998).
8. Come back to skills over the years. For skills meant to last, Willingham suggests planning on at least three to five years of practice (Willingham, 2019). Our guide on how to learn something new covers ways to make learning stick.

What the evidence doesn’t support

• All-purpose brain workouts. Classical music, learning an instrument and learning chess have all been tried as general boosters of intelligence, and none has worked as hoped. School programs that teach general thinking skills as add-ons have had limited success. When transfer to new kinds of problems was tested, the gains were absent or modest and faded quickly (Willingham, 2019).
• Expecting curriculum-wide programs to add much. College students’ critical thinking skills and dispositions improve substantially over a normal college experience. Programs aimed at improving critical thinking across the whole curriculum don’t necessarily add long-term gains on top of that (Huber & Kuncel, 2016).
• Counting on intelligence or expertise to protect you. Higher cognitive ability doesn’t reduce the bias blind spot, the tendency to see biases in others more than in yourself (West et al., 2012). Experts also reason less well outside their own field (Willingham, 2019).
• Simply resolving to be fair. Instructions to be fair and unbiased corrected bias less than a concrete strategy did (Lord et al., 1984).

Where the sources disagree: general skill or subject-specific?

Researchers don’t fully agree on whether critical thinking is one general skill or many specific ones:

• The Delphi panel (1990) held that critical thinking skills are useful across almost every area of life. It argued against narrowing critical thinking to subject content, while agreeing that one of the best ways to learn it is within a subject (Facione, 1990).
• Halpern (1998) wrote that numerous studies show critical thinking can be learned in ways that transfer to new contexts (Halpern, 1998).
• Willingham (2019) finds the research on transfer “decidedly mixed.” He says the existing data favor the view that critical thinking is mostly made up of specific skills, and that there’s no proven way to teach general thinking skills directly (Willingham, 2019).

The Delphi report records expert opinion, while Willingham’s review and the 2015 meta-analysis are newer and rest on experimental studies, so they carry more weight on how skills are learned. In practice the positions fit together. The meta-analysis found effects for both general and subject-specific skills (Abrami et al., 2015), and its strongest results came from teaching general principles explicitly and then applying them to real topics (Clearing House Unterricht, TUM). For anyone working on their own thinking, that means learning the general ideas and practicing them inside subjects you know well.