A study on choice and reaction time.
PhD thesis, University of Glasgow.
Full text available as:
Every day we have to make choices, to solve small and big life problems, and we behave accordingly. The experience becomes crucial and determinant in sport situation. The research intends to propose an analysis, at a first level, of the main factors involved in choice and action. The direction and the conditions proposed by the study could be a framework for future research and for further developments in sport.
Selection among options, choice, focal action and the time of reaction are the centre of the investigation. The main parameters are choice and reaction time. Choice describes the process of weighting and picking the more suitable alternative. Reaction time is the time of preparation and execution, the time during selection, choice, and the performing of the focal action, which ends at the recording of the data.
Reaction time period and choice events were analysed separately. Choice, in particular, was related to a positive prediction, that is to a correct answer. Among the variables involved in the experimental situation, choice, reward, delay and cues were chosen. The main aim was to find the change in reaction time, when a correct response could be predicted.
A special Reaction Time Device was built up, tailored on the specific needs of the experiments. The period between a visual “go” signal and a target touching by an arm, the focal action, was the time of reaction. The feedback given by the Device informed the subject on the correct or not correct choice. A basic protocol was defined, and the experimental plan, built up by 4 steps, related to specific aims, was fixed.
The first step had an introductory role. It tested simple and choice conditions. Random sequences and random delays were presented. Choice reaction time was examined when dominant, non-dominant and either arm were performing. The final evaluations allow detecting correct and incorrect answer factors, and the effect of different delays, as first reference of these components. The data register no difference in reaction time between dominant and non dominant arm performed during choice.
The second step involved reward factor. The aim was to compare reaction time without and with a possibility of reward. The results showed that at those conditions, no significant reaction time difference was registered.
The third step examined the delay factor and tested the reaction time when a variable or a fixed foreperiod was presented, before the “go” signal. It was concluded that there is no significant difference in time of reaction between a constant and a changeable delay.
The contractions of the subject’s focal and postural adjustment representative muscles were simultaneously recorded during the tests through EMG, to check the preparation phase and, eventually, a possible change in muscle contractions during the selection of the alternatives. The results confirmed that choice is a high level process and that, at least at those conditions, muscles are involved only at the last, final stage, after selection.
The former 3-step experiments had the role of premise to test the weight of some relevant variables. The centre of decision-making process is guessing, is anticipating the event. It implies the presence of clues, which can be identified, recognised and can lead the person to predict the next answer. This was the main motive of the fourth step. Three of the four sequences had a pattern, made up by 3 numbers, presented 4 times in the same sequence. The reaction time results included very short and very long values, far from the normal distribution. They were transformed in Log (- 350), to get coefficients suitable to be processed through parametric tests.
The numbers suggested that after the first experiences of the pattern, some subjects, having the feedback at each trial, identified and recognised the regularity and tended to be more correct at the last presentation. Some of them took a longer time to come to a choice, some were quicker. Two among 3 patterns were detected easier. In one of the 3 special sequences, the percentage of pattern correct answers was 48%, beyond the probability limit. It means that nearly half of the answers were positive and nearly half of the people guessed the pattern.
There was a clear difference in percentage of correct answers between the random sequence, which remained within the probability percentage, 33%, and 2 of the 3 sequences. In addition the percentage of patterned correct answer was higher than the total percentage of positive responses, inside the same sequence. Nevertheless in statistical terms the p values were not significant.
The 2 factors tested during the investigation, reaction time and choice, showed that, at the specific conditions of the experiments, there is no clear reciprocal correspondence. Unlike the studies in the field, correct answers were not directly related to lower reaction times, as expected, in patterned trials. Sometimes the subjects took more time to make their choice.
Short and long reactions, within the same subject and among quick and slow volunteers, balanced the data. The results were not significant, nevertheless the differences were evident and in the right direction.
The conclusion was that in a choice situation, when guessing was encouraged by cues to get correct response, the number of patterned correct trials tended to increase, in particular at the last repetition of the pattern.
The analysis on reaction time could not confirm the expected relationship between pattern correct trials, a sign of guessed cue, and a decreasing time of reaction.
Actions (login required)