Telling yourself I can do better, can really make you do better at a given task, a study published in Frontiers in Psychology has found.

Over 44,000 people took part in an experiment to discover what motivational techniques really worked. In conjunction with BBC Lab UK, Professor Andrew Lane and his colleagues tested which physiological skills would help people improve their scores in an online game.

This complex study examined if one motivational method would be more effective for any specific aspect of a task. The methods tested were self-talk, imagery, and if-then planning. Each of these psychological skills was applied to one of four parts of a competitive task: process, outcome, arousal-control, and instruction.

People using self-talk, for example telling yourself "I can do better next time" - performed better than the control group in every portion of the task.

The greatest improvements were seen in self-talk-outcome (telling yourself, "I can beat my best score"), self-talk-process (telling yourself, "I can react quicker this time"), imagery-outcome (imagining yourself playing the game and beating your best score), and imagery-process (imagining yourself playing and reacting quicker than last time).

They also found a short motivational video could improve performance. Participants watched a short video before playing the online game. The coach for these videos was, none other than, four-time Olympic gold medalist Michael Johnson, an athlete known for advocating mental preparedness in addition to physical training.

If-then planning was found to be one of the least successful of this study, despite being an effective tool in weight management and other real life challenges.

Professor Lane said: "Working on, 'Can You Compete?' was inspirational and educational; since we have been developing online interventions to help people manage their emotions and doing this across a range of specific contexts from delivering a speech to fighting in a boxing ring, from taking an exam to going into dangerous places."

Over 44,000 people participated in the study, an astounding number considering that the majority of psychological experiments have fewer than 300 participants. The participants were divided into 12 experimental groups and one control group, also impressive, because most studies have two or three experimental groups.