se, it is necessary to use the following steps in order to determine whether the bias in recommending for promotion was a change variable or was due to discrimination. The students could work in groups and brainstorm regarding the possibility of discrimination or the absence of it.
It is next necessary to work out a two way mathematical table showing the two variables- recommended or not recommended for promotion, and gender- male / females. It is now necessary to consider the three possible outcomes that could arise:
From the above, it is well discerned that there is no element of discrimination present. In the event, where there was perceived discrimination against women, the figures for recommendation would not be as figured above.
The use of the simulation technique is indeed a robust tool for repeating the exercise several times to determine its probability or the absence of it. In this case however, it is seen that the value of probability of occurrence of discrimination is just 4%. Thus in most cases, the theory of discrimination may not hold water and could be rejected after ruling out the possibilities of Type 1 or Type II errors.
Typically, a Type I error is when we could make an error in asserting that discrimination against women does play a part, when actually it has not; and secondly Type II error, when in reality, there being evidence of discrimination against women, but it was not accepted.
Navigating through data analysis: Chapter 2: Making decisions with categorical data. (2009). Navigations Series, p.1. Retrieved June 1, 2010, from