In this study we generated number sequences at two different speeds, which relatively are one number every second (1Hz) and one number every four seconds (4Hz) in order to test varying conditions in which individuals might act randomly.

Authors Avatar

Randomness            

Running head: RANDOMNESS AND RANDOM IDENTIFICATION

Determining Randomness and Random Identification

Li Howe Tan (SID: 308142659)

University of Sydney

 Tutorial: Tuesday (4.00pm – 5.00pm)

Abstract

In this study we generated number sequences at two different speeds, which relatively are one number every second (1Hz) and one number every four seconds (4Hz) in order to test varying conditions in which individuals might act randomly. Specifically we are interested the extent at which individuals were to make stereotypical errors in generating numbers at different speed, and how they would take into account of their previous responses of their decisions in order to formulate a new sequence. Also, to confirm if individuals are capable of being random under stipulative conditions. The result obtained was highly correlated, however we cannot determine on how good individuals are at generating numbers randomly, as ‘random’ is a term which has been stereotyped as being not the same, however could be argued that a sequence is not random if an individual uses effort to think of the following number he or she wants to generate.

Determining Randomness and Random Identification

        It has long been discussed that individuals uneducated in probability theory will often deviate from the statistical notion of randomness (Reichenbach, 1934/1949, as cited in Rapoport & Budescu, 1992). This notion was tested by Bakan (1960), that individuals produce far too many runs suggesting the incapability of humans to produce random sequences without bias from inherent patterns. This notion was contested by Ayton who argued that humans are unable to behave randomly due to instructional bias in asking individuals to behave random (Ayton et al., 1989, as cited in, Rapoport & Budescu, 1992). Rapoport & Budescu (1992) did empirical testing on this notion based on a dyadic interaction model proving that instructing individuals to predict randomness only caused them to deviate further from statistical randomness. Following the notion of instructional bias, we hypothesize that under the Random condition in terms of deviation will be the greatest. If individuals do behave randomly under competitions, their deviation should be the lowest with Sequence condition in the middle (Rapoport & Budescu, 1992). In the first task, we generated number sequences at two different speeds, which relatively are one number every second (1Hz) and one number every four seconds (4Hz) in order to test varying conditions in which individuals might act randomly. From the experiment we hypothesize that an increase in speed would more likely greatly increase stereotypical series of errors of people generating random numbers, as individuals would not ignore the number before and try to make the sequence look random. Therefore it is not random as individuals took the number written before into account, thus not random, and is one stereotypical error. In the second task there were two conditions in this, where half of the classes were shown sequences on the computer screen, and were asked to determine whether or not it was randomly generated, or following a type of pattern (ie. a random sequence, or a patterned sequence). And the other half of the classes were to determine whether the sequences shown were human generated random numbers, or computer generated random numbers instead. With reference to the first task that was carried out, the random string of sequence of numbers generated by students were ranged from a very typical human generated sequence as it mostly had the behavior of having the many sequences in between the numbers generated and less repeats (ie. 1-3-5-4-2) to a very typical perfectly random strings such as 1-1-1-3-2 (ie. less sequences and more repeats). Thus this experiment was conducted to determine how people identify random as being random, ie. identify the string of numbers as random as how they would generate randomly, and not of what is actually random, and also how the environment would affect the idenfication of randomness as well as when individuals are told to look for patterns within sequences while determining whether it is random or not.

Join now!

Method

Participants

        The participants in the study consisted of 398 students of which 211 of them are female participants, from their respective tutorials who were taking Psychology 1002 in the University of Sydney for a mandatory experiment. Their mean age was 19.1 (S.D. 4.3)

Materials/Procedure

        The randomness and random determination experiment was constructed to measure the validity of what people classify of random as well as how people get affected by time in generating random numbers in a task at different speeds. Computers were used to recognize as to how students were required to choose whether a particular sequence were ...

This is a preview of the whole essay