It has been established that for choice of alternatives not only the information characterising such alternatives is important, but the manner of its presentation as well. Therefore, the question is whether a similar anchoring effect occurs in other fields of estimation. Since in the course of research it has been established that the precision of estimation of event occurrence time varies due to minor changes in inquiry methods, e.g. increase or reduction of the frequency of forward or backward telescoping, therefore the anchoring effect is presumable in the field of time estimation. Sixty undergraduate students participated in the study. The objective of this study is to determine whether the manner of time-span description influences estimations of the people involved in the experiment when they are to reply whether an event occurred within a certain interval of time. The study tested the hypothesis stating that when an instruction indicates a date of the lower limit of an interval, the subjects will be more precise in estimation whether the event occurred within the certain time-span than that specifying no date whatsoever – just putting a question whether the event took place within a certain interval of time, suppose, within the last halfyear.
The experiment carried out was a between-group one. The first group was given the lower limit of an interval, while the second one was given none. Both groups were given time-spans of the same duration.
The study data proved the above-mentioned hypothesis. The group that was given the date of the lower limit of the interval made less mistakes (t58 = 4.084; p = 0.000). The study analysed the mistakes occurring due to the forward telescoping only. The impact of the intervals characterisation manner upon the number of mistakes is average (rpb = 0.42). From the point of view of classical rationality, it should be expected that both groups will make approximately the same number of mistakes, however, it has not happened, therefore it will be observed that the principle of invariance has been violated. The previous study (Gaskel et al., 2000; Loftus and Marburger, 1983) has established that when people are to estimate whether the event occurred before or after an event relevant for them, the number of mistakes is less than in the cases, where such relevant event is not indicated. The relevant events become landmarks, the time of occurrence whereof is the point of comparison to the time of other events occurrence. Further to the data of our study, undistinguished dates become the landmarks.
Please read the Copyright Notice in Journal Policy.