A specimen question is presented below followed by a model answer, followed by an analysis of how the model answer meets top markband requirements. Happy learning!
The Question
Discuss the following study with reference to two or more of the following concepts: bias, causality, measurement, and/or responsibility.
To test whether self-esteem affects attraction, Kiesler and Baral conducted an experiment. They administered a fake IQ test to two groups of men and gave them fictitious scores.
One group was told that they had scored the highest scores ever seen on the test. The second group was told that there must have been some misunderstanding because their scores were so low and that the test markers could not account for the errors. The second group was asked to redo the test two weeks later.
The scores were given to each participant individually, after which the participant was asked to wait in another room to receive his payment for participating in the study. During that time, an attractive woman (who was part of the experiment) walked into the room and sat one seat away from the participant.
The researchers wanted to see if the participant’s self-esteem affected their willingness to engage in discussion with an attractive woman. They found that the men who received high test scores engaged in conversation with the woman more quickly and in more detail than the men who received low test scores.
Model Answer
The experiment conducted by Kiesler and Baral aimed to test whether a man’s level of self-esteem would influence his willingness to engage in conversation with an attractive woman. While the study presents a clear experimental design, a critical examination shows significant methodological concerns, particularly regarding bias and measurement. These issues substantially undermine the validity and generalisability of the study’s conclusions.
The study shows prominent forms of bias, primarily gender and cultural bias. The research uses only male participants and defines the social interaction solely as an encounter with an attractive female confederate. This approach assumes that attraction and social initiative are relevant only to heterosexual men, ignoring the behaviour of women and individuals with other sexual orientations. As a result, the findings cannot be applied to human behaviour in general, but only reflect the responses of a specific group under very particular conditions.
Furthermore, the study carries a strong Western cultural bias. The expectation that men with higher self-esteem will more readily initiate a conversation reflects cultural norms that value extraversion and assertiveness, which are more common in individualistic Western societies. In many other cultural contexts, such behaviour might be considered inappropriate or overly forward. Therefore, the dependent variable – willingness to engage – is not a neutral measure but is deeply influenced by the researchers’ own cultural assumptions about appropriate social behaviour.
Equally problematic are the issues surrounding the measurement of key variables in the study. First, the independent variable, self-esteem, is manipulated by giving participants false feedback on an IQ test. However, there is no verification that this manipulation actually succeeded in creating the intended psychological states of high or low self-esteem in all participants. Individual differences, such as a person’s pre-existing level of self-esteem or their scepticism about the test, likely affected how they reacted to the feedback. This means the study lacks a proper manipulation check, which compromises internal validity because it is unclear whether the independent variable was consistently and effectively altered across the two groups.
Second, the measurement of the dependent variable is subjective and lacks precision. The researchers measured “willingness to engage” by timing how quickly a participant began talking and by judging the detail of the conversation. The concept of “detail” is not objectively defined and would require interpretation by the observers, introducing the potential for subjective bias, especially since those observers were not blinded to the experimental condition. Moreover, this measurement may not be a valid indicator of either attraction or self-esteem. A participant might talk quickly and in detail for reasons unrelated to confidence or attraction, such as nervousness, politeness, or simple curiosity. Conversely, a person with high self-esteem might choose not to engage extensively. Therefore, the chosen measure has poor construct validity, as it fails to isolate the specific psychological trait the study intends to examine.
These problems of bias and measurement are not isolated; they interact and compound each other. The cultural biases inherent in the study’s design directly influence what behaviour is chosen to be measured and how it is interpreted. The act of “engagement” is thus defined through a culturally specific lens, which in turn affects the objectivity of the measurement process itself. This creates a circular problem where the researchers’ assumptions shape both the design and the outcomes.
In conclusion, while Kiesler and Baral’s study attempts to establish a causal link between self-esteem and social behaviour, its methodological flaws are substantial. The pervasive gender and cultural biases limit the relevance of its findings to a very narrow context. At the same time, serious weaknesses in measurement, including an unverified manipulation and a subjective, low-validity dependent variable, cast doubt on the reliability and internal validity of the results. Ultimately, the study highlights the importance of careful design in psychological research and serves as a reminder that clear experimental narratives must be supported by rigorous and unbiased methodological practices.