Key study: Boroditsky (2001)

English mostly describes time horizontally ('looking forward to the weekend', 'that's behind us'). Mandarin also often uses vertical terms: earlier events can be described as 'up' and later ones as 'down'.
Aim: to see whether these language differences affect how people think about time, even when using English.
Participants: 26 native English speakers and 20 native Mandarin speakers (who also spoke English) at Stanford University. All were tested in English.
Procedure: on a computer, participants first saw a spatial prime: either a horizontal picture (for example, a black worm ahead of a white worm) or a vertical one (a black ball above a white ball), with a true/false question about it. Then they answered a true/false time question such as 'March comes earlier than April'. Their reaction time was measured.

Results: English speakers answered time questions faster after horizontal primes. Mandarin speakers answered faster after vertical primes, even though they were tested in English.
In a further experiment, English speakers were trained to talk about time using vertical terms; afterwards they too answered faster after vertical primes.
Conclusion: the language we speak can shape habitual ways of thinking about abstract ideas such as time. This supports linguistic relativism (language influences thought), but not strong determinism, because English speakers could learn a new way of thinking.

Strengths: a laboratory experiment with standardised computer tasks and objective reaction-time data, so it is easy to replicate. The training experiment suggests a cause and effect link with language.
Weaknesses: small samples, all university students, so it may not generalise. The tasks are artificial, so they may not reflect how people think about time in everyday life. Some later researchers failed to replicate the findings (for example January and Kako, 2007). Differences could be due to culture as well as language.