EXPLICIT ALTERNATIVE TESTING 4
Explicit Alternative Testing
In the criminal as well as the civil cases, the use of a single test to determine whether a person is cheating or he or she is telling the truth is not recommended and there should be at least six tests that should be involved. The reason behind choosing six binary tests is that chances are 0.56 which translates to 0.016 and which is within the brackets of the 95% confidence interval which is the accepted probability level of a two-tailed test.
Explicit Alternative Testing (EAT) can be used in the forced choice where the sensory inducements are provided to the person under the test. The person is therefore required to either accept or refute that he can recognize or remember. In this test, the objects displayed cannot be missed by a person without impairment not unless the individual is falsifying where their reporting will indicate a chance reporting that is lower. More report is given by an individual who is genuine in the answers that he or she is presenting. The lesser the reporting, the more the error in the probability testing making it fall outside the accepted 95% confidence interval, and the more the individuals are faking the report.
In the case of MT, the attempted murder case followed that a woman was stabbed and left unconscious, but later she recovered. She then identified MT as the suspect of the attempted murder where MT was subjected to a forced-choice test where out of the 24 binary responses that only the offender could have known, 13 were correct. The probability of guessing 13 answers correctly of the possible 24 questions translates to 0.4194 which is greater than 0.05, therefore, proving MT’s innocence.
A research study should present an accurate reflection of the real world and generally when respondents are giving their responses in a questionnaire; they do not take into account the importance of their responses. Instead, they respond for the sake of completing the survey hence, giving wrong or cooked answers. Through the use of a series of yes/no responses, there is the potential of identifying responses that do not have consistency and therefore, a need to ignore those responses. The binomial distribution, thus, remains relevant in my study topic.
References
Hall, H. V., & Thompson, J. S. (2007). Explicit Alternative Testing: Applications of the Binomial Probability Distribution to Clinical-Forensic Evaluations. Forensic Examiner, 16(1), 38.