The study is a mixed theoretical-empirical study regarding bystander intervention. It is theoretical in the sense that it applies known concepts of the bystander effect and helping mechanisms.
It is however dominantly empirical since it bases its findings and method on observations of actual circumstances. It attempts to reproduce actual situations through a controlled experimental environment wherein limited variables are introduced and measured by the researchers.
The experiment presumes to manifest high external validity in the sense that the situation presented could likely occur in actual real-life scenarios. Even the theory behind the research has been observed to happen in actual crimes or situations. In fact, this same research resulted from a crime effected in New York wherein witnesses failed to give aid to the victim.
Thus, if variables were adequately isolated and conclusions logically inferred, the results of such experiment would be highly valid externally. The problem herein lies in the low levels of construct validity. Albeit the generalizability of the theories in question, it is questionable whether the appropriate instruments of measurement were in place.
The results of the experiment discussed the effect of group number, sex, and educational background on bystander intervention. However only one of these factors was adequately represented, that of group number. Bystander intervention was in practice representative only of the presence of unobservable or unperceived bystanders, as differentiated in the study’s literature.
The effect of bystander presence with regard to perceived and observable bystanders could not have been measured by the researchers as all the bystanders they had planted in the experiment with the participant were unobservable, as they were supposedly in different rooms.
The factor sex was discussed whenever the participant’s sex was different from that of the perceived group member or members, particularly the difference when there was a potential male volunteer in the group. However, the research design shows a large inequality in the number of male and female participants. Further, the research shown to back up the discussion had not been previously discussed to ground the relevance of the effect of sex in helping as to that of the actual study.
In other words, there was no previous discussion as to the intent of the researchers to measure such factor, giving the impression that the discussion was inserted only after the results of the experiment had come in – an unreliable and bias-prone practice in research. Further, the factor regarding the medical background of a potential volunteer and other group member was likewise not discussed until the results portion of the paper.
The failure to distribute this factor to the three observed group sizes may also have presented disparity in the data gathered. There was also a lengthy discussion in the results regarding the avoidant-avoidant nature of the conflict that emerged in participants who did not intervene in the emergency presented. T
his is highly speculative in nature and is largely unsupported by self-reports presented in the research paper. In fact, there wasn’t even any previous research presented to ground the arguments made.
There seems to be high internal validity as the research adequately described the means of measuring intervention and gave data and computations regarding the measurements obtained. There was also adequate presentation of the commutations of the participation of the respondents.
Further, there was also a note regarding participants who had been dropped and for what reason this had been done. Despite what would have been expected given the high external validity of the concept being studied, the results of this experiment show low external validity. The situation presented shows a bystander who has previous knowledge regarding the situation of the victim of the emergency.
The bystander thus has knowledge of the nature of the emergency and the cause of the same. Furthermore, the previous rounds of discussion that the participants had engaged in regarding their college experiences establishes a semblance of relationship between them. The mere fact that they are in the same research environment for a psychological experiment already creates a level of relation between the participant and the victim.
This may not be and often is not the case with actual situations wherein the bystander effect is observed. In many of these situations the participant is not related to the victim and there is no knowledge regarding the situation unfolding – except for certain circumstances where the extremist nature of the emergency could not be mistaken for any other situation. Overall, the research proves to be high in internal validity, low in construct validity, and low in external validity.
Based on existing literature regarding this topic, the research presented information consistent with findings of other researchers. An increase in group size has been found to have a minimizing effect on bystander intervention.
The findings, such as in the murder case presented as the ratio for the research in question, display this finding in practice. It should be bourne in mind however that in spite of the correct conclusion of the research hypothesis, the other results should not be taken at face value, nor should the procedure employed by the researchers by reproduced without thorough analysis of the accurate representation of variables.
Darley, J. M., & Latane, B. (1968). Bystander intervention in emergencies: Diffusion of responsibility. Journal of Personality and Social Psychology, 8 (4), 377-383.