A reliable test is universally consistent and equitable. If it is a predictor for admissions, it should have enough equitable content on the test to not directly single out a particular group of individuals. According to Cohen & Swerdlik (2018), the item-validity index should provide a statistic that shows the test is measuring what it was created to test. If the test had a fair amount of participants not secluded to a particular group scoring low, we could assume that the test is ethical. However, the scenario is stating that members of a particular group score low. This could either be because of an achievement gap or an opportunity gap in this particular group. Furthermore, predictive validity shows scores based on criteria. If the criteria are based on a standard foreign to a particular group, then the college's well-established predictive validity is a rigged system. Testing is a tool, not a weapon.
Consider the term "well-established" as a negative use of words regarding an admissions test. Well-established could mean old or predated to a time when the particular group wasn't part of the test development phase. The predictive validity is useful for picking the good student or the desired student (Cohen & Swerdlik, 2018). It is unethical to use a test to disqualify potential admission when the criterion wasn't inclusive. Since there is a suspicion of foul play, I would constitute this test as an ethical dilemma and would no longer administer it until it is updated. Lastly, only if the test is sensitive to legal mandates and is only influenced by school-related variables can the school use it to disqualify(Cohen & Swerdlik, 2018). No matter how shady it looks.
One serious factor is the issue of test-takers variables. Cohen & Swerdlik (2018) mention different variables such as emotional problems, lack of sleep and stress, and drugs and medication effects on certain test-takers. Although, these factors may be universal and not racial, class, or gender-specific. There are definitely individuals where these factors are predominately cognitive impediments and strong sources of error variance to challenge the reliability.
Another serious factor is the concept of the cultural relevancy of the test. If a foreign exchange student attends a Southern Georgia school, the content likely taught will not be culturally relevant to the individual. Therefore, there is an achievement gap in learning as many ethnicities are forced to learn irrelevant or uncommon subjects. We teach the good parts of history while sharing other countries' bad parts, such as the Nazi Germany Reich. Cohen & Swerdlik (2018) mention World War II hero Yeiichi "Kelly" Kuwayama as a great hero on a history test in America, but this may not be valid in another classroom, in another place at another time.
For African-Americans, is there a consideration of cultural sensitivity when teaching American history? Andrew Johnson and other presidents are not my heroes. Mr. Johnson stated, "African Americans were incapable of self-government and relapsed into barbarism if they weren’t closely supervised." For a minority group, they are our oppressors. No history test should consider perceptions and judgments as gradable criteria to ensure validity. It forces me to accept the mistreat, hatred, lacks of cultural sensitivity for my race.
Another Article: https://socialism.com/fs-article/standardized-testing-a-scam-and-a-shame/
Cohen, R.J. & Swerdlik, M.E. (2018). Psychological testing and assessment (9th ed.). Boston, MA: McGraw-Hill Custom. ISBN: 978-1-260-36533-7
Comments